From joconnor at cybermesa.com  Thu Dec 21 13:42:20 2000
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Thu, 21 Dec 2000 18:42:20 +0000
Subject: while (thread.alive()):
 
Message-ID: <3A424F0C.A57C155@cybermesa.com>

>> I'm continuously amused that people try to take language design
>> forward and still drag obsolete ideas along for the ride.

>Like having a print command that doesn't, by default, send output
>to the printer...  :o)

Good example

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

"God himself plays the bass strings first when He tunes the soul"



From syver at NOSPAMcyberwatcher.com  Wed Dec 13 14:22:51 2000
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Wed, 13 Dec 2000 20:22:51 +0100
Subject: Source Code Management
References: <fb4f3tsgffnmvme64aol1cn397te7ln349@4ax.com> <918711$meh$1@reader1.imaginet.fr> <prff3tc59qtv0qr9afj16dali3jv789c7u@4ax.com>
Message-ID: <918ifa$rg8$1@troll.powertech.no>

"Dale Strickland-Clark" <dale at out-think.NOSPAMco.uk> wrote in message
news:prff3tc59qtv0qr9afj16dali3jv789c7u at 4ax.com...
> I've just dug CVS out and downloaded it but I discover that the NT version
appears to require a
> mounted volume.
>
> That's a disappointment.
Nah, you just have to set up the environment properly. set the cvs root or
something.




From hs at paradise.nirvananet  Mon Dec 25 10:18:25 2000
From: hs at paradise.nirvananet (Hartmann Schaffer)
Date: 25 Dec 2000 10:18:25 -0500
Subject: One Python 2.1 idea
References: <mailman.977590932.19335.python-list@python.org> <925baq$5sj$1@nnrp1.deja.com> <925tak01tpu@news1.newsguy.com> <926ici$vi9$1@nnrp1.deja.com>
Message-ID: <927og1$66v$1@paradise.nirvananet>

In article <926ici$vi9$1 at nnrp1.deja.com>,  <rturpin at my-deja.com> wrote:
> ...
>responsibilities. Application programmers should use the
>abstractions that best solve their problem, functionally,
>assuming that these are implemented well. Language
>implementors should strive to make sure that the
>implementation for an abstraction performs well in all
>the ways it is used. This route provides significant gains
>in productivity. But it only works if BOTH groups take up
>their mantle.

unfortunately, that simply doesn't work in practice.  programmer tend
to develop a style that is dictated by the implementation they are
using.  if a particularly abstraction performs lousy in the
implementation i'm using and my users are screaming that they are
falling asleep while waiting for a particular result, telling them
that it might get better with the next release of the compiler won't
cut it.  if i expect something like that to happen, i won't wait for
the complaints to come in, but write the code so that they won't,
i.e. avoid the abstraction

> ...
>I think this approach is misguided. A listBuffer, as
>far as I can tell, is functionally identical to a
>sequence. We've cluttered the language with another
>abstraction, whose only purpose is to help the compiler
>decide an implementation. Now, everyone learning Python
>has to learn about listBuffers, in order to read your
>code. This same mistake has been made already with
>xrange(). Given enough time, we'll have to teach
>programmers about a dozen different things that all
>behave like Python sequences. Blech!
>
>In my mind, the ideal is very different. The Python
>implementation should have a dozen or more ways of
>implementing sequences, depending on how they are
>constructed, what kinds of objects are put into them,
>the contexts in which they are indexed, whether they
>are used for iteration, etc. The vast majority of the
>time, these things can be determined at compile time.
>There shouldn't be a different language abstraction
>for sequences that are not stored. xrange() is an
>abomination. The compiler should simply use this
>implementation when it fits.
>
>ONLY ONE SEQUENCE ABSTRACTION SHOULD BE EXPOSED TO THE
>APPLICATION PROGRAMMER. It would be a mistake -- a BIG
>mistake -- to continue the process of cluttering up
>the language with new abstractions, merely to give
>implementation directive to the compiler. I vote we

noble thought, but i see at least two objections here:
   
1. the compiler can only find the best way to implement a general
   abstraction iff it can see all uses of a particular instance.  in a
   language that has seperate modules that come together at execution
   time this often is not the case

2. the programmer should be able to communicate to the compiler all
   the relevant information about the particular instance of an
   abstraction he uses that s/he is aware of.  whether you use extra
   language features or annotations is irrelevant, and i would argue
   that for all practical purposes the annotations are as much part of
   the language as some special construct, only that it uses a
   different syntax.  if the implementation improves, the compiler is
   as much free to map different source constructs into identical
   internal constructs, as it is free to ignore annotations

> ...

hs


From sab at NOSPAM.freeuk.com  Thu Dec 21 15:45:18 2000
From: sab at NOSPAM.freeuk.com (G. Willoughby)
Date: Thu, 21 Dec 2000 20:45:18 -0000
Subject: 2 question from a python newbie!
References: <91tl43$rko$1@neptunium.btinternet.com> <slrn944q0t.919.scarblac@pino.selwerd.nl>
Message-ID: <91tq58$cq6$1@uranium.btinternet.com>

Thanks mate,
                i'm not a total programming newbie just a python newbie  :)
ill check out that site and them, thar books.

ta

G. Willoughby

P.S. so python is really used as a scripting language and nothing more?


"Remco Gerlich" <scarblac at pino.selwerd.nl> wrote in message
news:slrn944q0t.919.scarblac at pino.selwerd.nl...
> G. Willoughby <sab at NOSPAM.freeuk.com> wrote in comp.lang.python:
> > 1). Can python scripts be compiled to 'stand alone' .exe files?
>
> Sort of. There is a utility ('Freeze') that packs your program and Python
> itself into one big .exe, and it's stand alone then. But that's not the
> usual way to use Python.
>
> > 2). What's the best book/s available for learning python from a total
> > beginers point of view?
>
> There are several options. Alan Gauld has a nice website for total
> programming newbies at http://www.crosswinds.net/~agauld/ . He's also
> often helpful at the tutor mailing list. He has a book out now, called
> "Learn to program using Python". It's brand new and I haven't seen any
> reviews yet, but you can try his page first and buy the book if you want
> to have an extended version printed on dead trees.
>
> Other options are for instance "Teach Yourself Python in 24 Hours" and
> "The Quick Python Book". There are also more tutorials online, see the
> Documentation link on http://www.python.org .
>
> --
> Remco Gerlich




From sholden at holdenweb.com  Thu Dec 14 10:38:21 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 14 Dec 2000 10:38:21 -0500
Subject: MS + Python = help???
References: <91an5e$qfm$1@nnrp1.deja.com>
Message-ID: <uS5_5.7682$4y4.198409@e420r-atl1.usenetserver.com>

Arint? <jamarijr at my-deja.com> wrote in message
news:91an5e$qfm$1 at nnrp1.deja.com...
> What is the address of the webpage that had the python help in the form
> of one of the new micrsoft html help format things.
>
> tia
>
Go to

    http://alldunn.com/python/

and use the py2docs.zip link at the top of the page.

regards
 Steve






From sh at ttsoftware.co.uk  Fri Dec 15 10:41:05 2000
From: sh at ttsoftware.co.uk (Steve Horne)
Date: Fri, 15 Dec 2000 15:41:05 +0000
Subject: what is a closure?
References: <3a39033b@news.xtml.co.uk> <S18_5.3929$Kd1.455131@newsb.telia.net>
Message-ID: <8nck3tocsg8kp1u2g1du6nd3lt9gfeckr2@4ax.com>

On Thu, 14 Dec 2000 18:09:54 GMT, "Fredrik Lundh" <fredrik at effbot.org>
wrote:

>In the following example, the "a=a" default argument binding
>binds add's local name "a" to the value of the "a" argument
>from make_adder.
>
>    def make_adder(a):
>        def add(b, a=a):
>            return a + b
>        return add
>
>    >>> m = make_adder(10)
>    >>> m(20)
>    30

Are you sure that's a closure? - It looks closer to currying to me.
And the definition seemed pretty close to the definition of an object
- a collection of related data and methods.

Currying, in functional languages, is the process of specialising a
function by specifying one or more of its parameters to produce a
function with fewer parameters than the original. The notation in
Python is OK, but not very convenient...

If fn_name has two parameters, you could apply currying to derive a
single-parameter function as follows...

  lambda p1 : fn_name (p1, 10)  #  only works for literals
  lambda p1, p2=whatever : fn_name (p1, p2)

Because the second form is a bit wordy, sometimes it is easier to have
a currying function such as...

  def curry_p2_of_2 (fn, p2) :
    return lambda p1, fn=fn, p2=p2 : fn (p1, p2)

This is clearly similar to your example.


I was reading a book about developing compilers recently (Modern
Compiler Design, by Dick Grune, Henri Bal, Ceriel Jacobs and Koen
Langendoen), and something that was presented in it was a 'closure
algorithm'. This took a collection of data and some related
expressions, and repeatedly evaluated the expressions using both
original and previously derived data until all the data it was
possible to derive was fully derived.

The thing that made it a closure algorithm seemed to be the fact that
the expressions were applied repeatedly on an accumulating set of data
until no further data could be derived - the closure being the
complete collection of data.

An example would be that given the expressions A = int + (int + int),
you could represent it as two simple expressions...

A   = int + <b>
<b> = int + int

And then, given an expression to determine result types from the
parameter types for the '+' operator, the closure algorithm would work
as follows...

Step 1 - it is not possible to determine A from A = int + <b>B, as <b>
is not yet known.

Step 2 - <b> = int + int, therefore <b> = int

Step 3 - A = int + <b>, therefore A = int + int

Step 4 - A = int + int, therefore A = int.

Step 5 - Unable to find any expressions to apply for which the inputs
         are known, yet the output is unknown, so terminate.

I don't claim to be an expert on this. I would simply appreciate it if
someone could clear up my confusion!

-- 
Steve Horne
Home : steve at lurking.demon.co.uk
Work : sh at ttsoftware.co.uk


From ben.hutchings at roundpoint.com  Tue Dec  5 11:05:28 2000
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 05 Dec 2000 16:05:28 +0000
Subject: os.rmdir() and os.removedirs() only work on empty directories?
References: <8v1p27$7d7$1@nnrp1.deja.com>
Message-ID: <uhf4ilupj.fsf@roundpoint.com>

jschmitt at vmlabs.com writes:

> I was a bit surprised to find that os.rmdir() and os.removedirs() both
> complained if I tried to delete a directory that was not empty.  Is
> that expected?

POSIX says the rmdir call should only remove an empty directory, so it
doesn't surprise me that os.rmdir behaves this way.  I have yet to see
what os.removedirs is for.

> I guess shutil.rmtree() is the tool to use if I really want to
> recursively delete a directory hierarchy that is non-empty.

I hadn't seen that one - I wrote this function myself when I wanted
recursive deletion:

def rmdir_recursive(dir):
    """Remove a directory, and all its contents if it is not already empty."""
    for name in os.listdir(dir):
        full_name = os.path.join(dir, name)
        # on Windows, if we don't have write permission we can't remove
        # the file/directory either, so turn that on
        if not os.access(full_name, os.W_OK):
            os.chmod(full_name, 0600)
        if os.path.isdir(full_name):
            rmdir_recursive(full_name)
        else:
            os.remove(full_name)
    os.rmdir(dir)

> I don't quite understand this part of the documentation os.removedirs():
> '...if the leaf directory is successfully removed, directories
> corresponding to rightmost path segments will be pruned way until
> either the whole path is consumed...'

I believe that this means it will remove successive parent directories
referred to in the given path if they have become empty - so if you have
the following directory structure:

    . (current directory)
    |
    +--a/
    |  |
    |  +--b/
    |  |  |
    |  |  +--c/
    |  |     |
    |  |     +--d/
    |  |
    |  +--e/
    |
    +--f

and you call os.removedirs('a/b/c/d'), it will remove directories
a/b/c/d, a/b/c and a/b (assuming the process is permitted to do so).
It will attempt to remove a, fail because a is not empty, and then
stop.

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From fredp at mygale.org.nospam  Wed Dec 13 03:35:30 2000
From: fredp at mygale.org.nospam (Fred Pacquier)
Date: 13 Dec 2000 08:35:30 GMT
Subject: Thank you, Fredrik. (was Re: Label-Value)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <WntZ5.3566$Kd1.414873@newsb.telia.net> <Pine.GSO.4.21.0012121550530.19682-100000@y.glue.umd.edu>
Message-ID: <9009628F0PaCmAnRDLM@161.48.76.20>

katz at Glue.umd.edu (Roy Katz) said :

>On Tue, 12 Dec 2000, Fredrik Lundh wrote:
>> Sigh.  The problem is that you keep using terms that mean
>> something for you, whether or not they apply to the current
>> situation.  [snip]
>> If you want to learn something new, you have to stop doing
>> that.  [snip]
>> Here's how it's done in Python:
>> [wonderful explanation snipped]
>
>Thank you very much, I understand now more or less how this works. 
>Furthermore I appreciate your thoughtful, measured, and flame-free tone.  

Seconded. I'm sure it's been useful to many other people who have been 
"doing" Python for as long as you have (including me) without an 
understanding  that deep, whether they will admit to it or not.
As Fredrik says, this info certainly exists elsewhere and could be pieced 
together from existing material. But having it summed up clearly and in few 
words in one place can be a revelation, so this definitely needs to go in 
python-snippets or FAQTS or somewhere... Thanks /F !

-- 
YAFAP : http://www.multimania.com/fredp/


From paulp at ActiveState.com  Fri Dec  8 11:52:34 2000
From: paulp at ActiveState.com (Paul Prescod)
Date: Fri, 08 Dec 2000 08:52:34 -0800
Subject: xml.dom
References: <90q292$k9k$1@merlin.rembrandtstr.bocholt.de>
Message-ID: <3A3111D2.5C236D22@ActiveState.com>

Magnus Heino wrote:
> 
> Hi.
> 
> How do I build an xmldocument using something in xml.dom?

I'm not sure what you are asking. Here's a potential answer if the
question is: "How do I parse an XML document into a DOM using the DOM
implementation that ships with Python 2.0."

>>> from xml.dom import minidom
>>> dom = minidom.Parse("c:\\temp\\test.xml")

 Paul Prescod



From lvirden at cas.org  Sat Dec 30 07:22:17 2000
From: lvirden at cas.org (lvirden at cas.org)
Date: 30 Dec 2000 12:22:17 GMT
Subject: Python legalities
References: <3A2FC63D.652101A@accessone.com>
Message-ID: <92kk1p$c65$1@srv38.cas.org>

According to Helen Dawson  <helend at accessone.com>:
:The legal department at my company is a bit nervous ...
:They are concerned about the
:risks of using software that may inadvertently contain fragments
:of patented or copyrighted code, without anyone who can indemnify
:us against this risk.


It's good to worry about this.  However, how do you handle the situation
when dealing with software you've written on site?  I would think that
passing the free software through the same reviews as the inhouse procedures
should be sufficient...

-- 
-- 
"See, he's not just anyone ... he's my son."  Mark Schultz
<URL: mailto:lvirden at cas.org> <URL: http://www.purl.org/NET/lvirden/>
Even if explicitly stated to the contrary, nothing in this posting


From fredrik at effbot.org  Fri Dec  8 05:51:26 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 08 Dec 2000 10:51:26 GMT
Subject: "**" operator in method/function interfaces ??
References: <90qdi3$ar6$1@reader1.imaginet.fr>
Message-ID: <O23Y5.2167$Kd1.313338@newsb.telia.net>

Gilles Lenfant wrote:
> What's the use of "**" in such a construct ?
> <sample>
> ...
> class MyClass:
>     def __init__(self, adata, **anotherdata)
> ...
> </sample>
>
> I found such construct in some examples of the Tkinter tutorial at
> pythonware, and in some other documents.
> I have gone around all books and web language tutorial I know and did not
> find anything about it.

    http://www.python.org/doc/current/ref/function.html

    If the form "*identifier'' is present, it is initialized to
    a tuple receiving any excess positional parameters,
    defaulting to the empty tuple. If the form "**identifier''
    is present, it is initialized to a new dictionary receiving
    any excess keyword arguments, defaulting to a new
    empty dictionary.

or, in other words:

    >>> def func(*args, **kwargs):
    ...     print args, kwargs
    ...
    >>> func(1, 2, 3, a=4, b=5, c=6)
    (1, 2, 3) {'b': 5, 'c': 6, 'a': 4}

</F>




From aleaxit at yahoo.com  Tue Dec 12 04:12:56 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 12 Dec 2000 10:12:56 +0100
Subject: 3rd argument to os.path.walk(): Why?
References: <912gdb$1oqr$1@news.idiom.com> <912nj60ccr@news2.newsguy.com> <m2vgsqw1k2.fsf@hooft.net>
Message-ID: <914q6v02l3p@news2.newsguy.com>

"Rob Hooft" <rob at hooft.net> wrote in message
news:m2vgsqw1k2.fsf at hooft.net...
>
>  AM> E.g., say you want to _count_ the files:
>
>  AM> import os
>
>  AM> class Counter:
>  AM>   def __init__(self):
>  AM>     self.count = 0
>  AM>   def inc(self, n=1):
>  AM>     self.count += n
>
>  AM> def count_names(count, dir, flist):
>  AM>   count.inc(len(flist))
>
>  AM> count = Counter()
>  AM> os.path.walk("/", count_names, count)
>
>  AM> print count.count,"files in all"
>
> If you'd do this, why not make "count_names" a method of the Counter
> class?

As things stand here, "Counter" is a completely general-purpose
class, which has no knowledge whatsoever of files, their names,
and path-walking -- it's completely decoupled from all of these
concerns, a low-level "utility class" that is reusable in all
sorts of contexts.

Strictly coupling a generic counter-class to specific things it
could be counting doesn't seem like an elegant design at all.

If anything, one might *derive* from Counter a NamesCounter
subclass, or *contain and delegate-to* a Counter from a wrapper
class -- assuming os.path.walk takes any callable (specifically,
a bound method is what we would pass here) even though the docs
specifically say it takes a *function* instead (the docs are
often too-restrictive regarding object-types expected by core
Python functions...).

Once that is feasible, then both idioms for setting callbacks
are equally elegant:
1. the classical procedural set-callback idiom, where you
    provide a function (typically not carrying state) and
    the state to be passed to it as an argument when the
    calling-back does happen;
2. the strongly object-oriented set-callback idiom, where
    you provide a callable object which entirely takes care
    of its own state issues.

The tradeoffs between these two design approaches are,
more or less, 'the usual suspects'.  One way to frame [1]
is strictly object-oriented terms is as an application
of the "Memento" design-pattern -- the 'function' might
still be a more-general 'callable', but some part of its
state is objectified as a separate 'memento' object that
is explicitly handled back to it as needed.

"Memento" is one of many important ways in which issues
of object-identity and object-state can be decoupled, and
thereby handled more explicitly, with finer-grained
control.  (One doesn't always _want_ to control things
on a fine-grained level, which is why it's nice to have
different idioms available in the language/framework:-).


Alex





From aleaxit at yahoo.com  Wed Dec 27 15:53:34 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 27 Dec 2000 21:53:34 +0100
Subject: Python and Napster
References: <000c01c06fe3$efc29520$102b2bc1@cadlab.it>, <92ap0g12b6r@news1.newsguy.com>, <sx226.18970$RC1.511898@newsread1.prod.itd.earthlink.net> <20001227164900.DF7E6A84F@darjeeling.zadka.site.co.il> <mailman.977910196.13387.python-list@python.org>
Message-ID: <92dptl11vqa@news2.newsguy.com>

"Moshe Zadka" <moshez at zadka.site.co.il> wrote in message
news:mailman.977910196.13387.python-list at python.org...
> On Wed, 27 Dec 2000, "Alex Martelli" <aleaxit at yahoo.com> wrote:
>
> > HTTP (and the Python tools around it) is great for file transfer (and
> > suchlike) but it's not oriented to peer-to-peer at all.
>
> Well, yeah, you'd have to implement the low-level part of peer-to-peer.
> But the HTTP infrastructure is usually enough for that: broadcast by
> performing a GET /?file=name to say you have a file called "name"
> and the nameing service turns out to be composed of 302 redirects.

I'll have to trust you on that.  A '302 redirect' is somehow broadcast
to...?


> > I think that actually transferring a file's data and metadata, once
> > you have a proxy for a remote object, is no big issue.  It seems to
> > me that naming, exceptions, resuming transfer (not supported by HTTP,
> > is it?)
>
> You have insulted the prophet: your punishment shall be cruel indeed!
> Of course HTTP supports resuming transfer. SimpleHTTPServer doesn't
> though. I don't think it should be hard to add, though.

I guess I'm stuck back in HTTP 1.0 times.  What's the transfer-resumption
transaction, then?


> An advantage of HTTP is that it will be possible to use the protocol
> with clients not written in Python. If what he wants is simple file
> exchange, there is no reason to be limited to Python at the protocol
> level.

If cross-language and cross-platform operation is desired, Corba
would seem most appropriate to me.  I saw no indication of such
desire in the original post, though -- cross-platform, yes, but the
idea seemed to be to have a single Python app that would handle
it all.


Alex





From fidelman at world.std.com  Thu Dec  7 07:36:21 2000
From: fidelman at world.std.com (Miles R. Fidelman)
Date: Thu, 7 Dec 2000 12:36:21 GMT
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu> <976153091.257544@clam-ext>
Message-ID: <G5770L.IGx@world.std.com>

Mark,

Mark Hadfield (m.hadfield at niwa.cri.nz) wrote:
: "Nathan Gray" <nospam at caltech.edu> wrote in message
: > What a rude response to a polite and legitimate question.
: Well I think a more legitimate question would have been:

: "I want to know about Pythons origins. I've looked in the obvious places,
: i.e. www.python.org, and read the stuff there but I want to know more about
: this: ..."

What a great way of summarizing things - this says it all.

Miles Fidelman


From sdm7g at virginia.edu  Thu Dec 21 13:26:29 2000
From: sdm7g at virginia.edu (Steven D. Majewski)
Date: Thu, 21 Dec 2000 13:26:29 -0500 (EST)
Subject: Python in "math through programming" curriculum
In-Reply-To: <slrn942nln.mjs.neelk@alum.mit.edu>
References: <2g2t3t06r2agrrjmfnr41eu4lc2u39a8qi@4ax.com> <3A3EFF47.67DE69CF@gte.net> <91n9i30pr7@news2.newsguy.com> <slrn9402bt.fa6.neelk@alum.mit.edu> <91qa3r01m78@news2.newsguy.com> <slrn942nln.mjs.neelk@alum.mit.edu>
Message-ID: <Pine.NXT.4.21.0012211311440.565-100000@localhost.virginia.edu>

On 21 Dec 2000, Neelakantan Krishnaswami wrote:

> 
> Yes -- the reference to Knuth's _Concrete Mathematics_ made me think
> that an introduction to CS was envisioned. For things like graphing
> and statistics, there's no reason not to use Python, since it has good
> graphing and statistical libraries. Static typechecking doesn't help
> here, I don't think, and in Haskell's case it becomes a positive
> nuisance (because of all the IO).
> 

XlispStat -- a free Lisp variant with extensions for statictics, vector
math and object-oriented interactive graphics -- is another good choice.

There's a good textbook by Luke Tierney available.

The graphics and statistics support is quite a bit more mature than
what's available in Python, and the graphics is better integrated and
easier to use. ( That maturity and integration, however, is one of 
the current problems: for example, the 3-D graphics predates the wide
availability of OpenGL libraries -- it would be nice to be able to
extend some of it's graphics abilities but the current graphics model
is pretty tightly tied. That's where Python's not being tied to any
one graphics model is an advantage. ) 


See: <http://www.xlispstat.org/> 


You might also check out StarLogo from MIT's media lab -- 
a special version of Logo designed to handle thousands of 
independent "turtle" agents in parallel:

<http://www.media.mit.edu/starlogo/> 


| StarLogo is a programmable modeling environment for exploring the
| workings of decentralized systems -- systems that are organized without
| an organizer, coordinated without a coordinator. With StarLogo, you can
| model (and gain insights into) many real-life phenomena, such as bird
| flocks, traffic jams, ant colonies, and market economies.


-- Steve Majewski <sdm7g at Virginia.EDU> 




From aleaxit at yahoo.com  Tue Dec 19 04:59:56 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 19 Dec 2000 10:59:56 +0100
Subject: Need Help sorting alphabetically.
References: <3A3F1A91.A7ED9182@araidesign.com>
Message-ID: <91nbjl0rj3@news2.newsguy.com>

"Chris Arai" <chris at araidesign.com> wrote in message
news:3A3F1A91.A7ED9182 at araidesign.com...
> Hi,
>
> This seems like it should be obvious, but it is evading me!
>
> I would like to sort alphabetically (I'm not sure that is the correct
> term) so that strings of alpha numerics sort alphabetically instead of
> by ASCII order.
>
> eg:
> >>> m=['Pg95_100.GIF',
> 'Pg9_100.GIF','Pg95A_100.gif','Pg9A_100.gif','Pg10_100.gif']
> >>> m.sort()
> >>> m
> ['Pg10_100.gif', 'Pg95A_100.gif', 'Pg95_100.GIF', 'Pg9A_100.gif',
> 'Pg9_100.GIF']
>
>
> I would like the order to be
> ['Pg9_100.GIF', 'Pg9A_100.gif', 'Pg10_100.gif, 'Pg95_100.GIF',
> Pg95A_100.gif' ]
>
> What is the obvious answer??

First: you have to *precisely* define, for each string, what
'data equivalent to it' you ARE comparing for your sorting
purposes.  This is no doubt the major task you face.

For your specific example, e.g., it would seem that the
'equivalent data' to each string is an integer made up
from the digits between a leading 'Pg' prefix and some
following non-digits, then the following non-digits if
any up to but not including a '_', then the rest - but
your criteria might well be very different.  First of all,
nail down those criteria.

Extraction of 'equivalent data' can be done by using
regular expressions, or in other ways.  RE's can be
very compact and pretty fast, but there may be other
simpler ways, depending on what exactly you're doing.

Assuming RE's, then, for example:

import re
cre = re.compile('(\D*)(\d+)([^_]*)_(.*)')

We match 4 parenthesized groups:
    0 or more non-digits (the prefix),
    1 or more digits (the number),
    0 or more non-underlines,
    whatever comes after the underling
Of these, the second field needs to be
converted to an int, the other will
remain strings.

def eqv1(astring):
    mo = cre.match(astring)
    if not mo: raise ValueError, "no match: %r"%astring
    groups = mo.groups()
    return (groups[0], int(groups[1])) + groups[2:]

if __name__=='__main__':
    for s in m:
        print s, eqv1(s)

Running this will confirm we're extracting the right
'equivalent' data:

D:\PySym>python ccc.py
Pg95_100.GIF ('Pg', 95, '', '100.GIF')
Pg9_100.GIF ('Pg', 9, '', '100.GIF')
Pg95A_100.gif ('Pg', 95, 'A', '100.gif')
Pg9A_100.gif ('Pg', 9, 'A', '100.gif')
Pg10_100.gif ('Pg', 10, '', '100.gif')

D:\PySym>

Actually, you'll probably want to lowercase some of
this, too, always for comparison purposes -- I suspect
you want it to be irrelevant to the sort whether a
string ends in '.GIF' or '.gif', for example, or
whether the prefix is spelled mixed, upper, lowercase.

So, you could change the last line in eqv1 (after
also changing the import to
    import re, string
of course), to:

    return (groups[0].lower(), int(groups[1])
        ) + tuple(map(string.lower,groups[2:]))

and now the test would emit:

D:\PySym>python ccc.py
Pg95_100.GIF ('pg', 95, '', '100.gif')
Pg9_100.GIF ('pg', 9, '', '100.gif')
Pg95A_100.gif ('pg', 95, 'a', '100.gif')
Pg9A_100.gif ('pg', 9, 'a', '100.gif')
Pg10_100.gif ('pg', 10, '', '100.gif')

D:\PySym>

more likely to be right.

Anyway, use this kind of exploratory approach to
make sure you have the right 'equivalence extraction'
function for your purposes; be sure to test quite
extensively on data which IS representative of the
scenarios you will want to handle!


Once you *DO* have a good 'eqv' function for your
purposes -- one that, applied to any string you
may be sorting, returns the correct 'equivalent'
tuple you want to use for comparison -- there are
basically two approaches to using it.  One involves
passing a comparison-function to the sort method,
and I would not advise it -- if the extraction
function does anything 'heavy', sorting lists of
substantial length can be quite slow this way.

A better approach (which I _think_ should be
called 'Schwartzian transform', from the name of
its inventor, but there is anything but consensus
on this group regarding the name, or the priority
of Randall Schwartz' "invention" of this...!):

def richSort(aList, eqv):
    auxList = [ (eqv(x), x) for x in aList ]
    auxList.sort()
    return [ x for (junk, x) in auxList ]


Changing the unit-test at the end of the module to:

if __name__=='__main__':
    for s in m:
        print s, eqv1(s)
    x = richSort(m, eqv1)
    print x

we now have the following test-output when just
running the module:

D:\PySym>python ccc.py
Pg95_100.GIF ('pg', 95, '', '100.gif')
Pg9_100.GIF ('pg', 9, '', '100.gif')
Pg95A_100.gif ('pg', 95, 'a', '100.gif')
Pg9A_100.gif ('pg', 9, 'a', '100.gif')
Pg10_100.gif ('pg', 10, '', '100.gif')
['Pg9_100.GIF', 'Pg9A_100.gif', 'Pg10_100.gif', 'Pg95_100.GIF',
'Pg95A_100.gif']

D:\PySym>

which seems a good match for the required functionality.

Function richSort is not complex.  It build an auxiliary
list, where each item's "sorting-equivalent" is prepended
to the item itself; sorts the auxiliary-lists with its
own sort method, which ('lexicographically'...:-) will
consider the comparison of the 'sorting-equivalents' to
have precedence; extract the results by junking all the
sorting-equivalents previously prepended.  It's much
clearer and more concise to express this in Python, of
course, as above done, than in English, at least my
personal, somewhat verbose idiolect of it:-).


If you *had* to do it with comparison-function-argument-
to-sort, it could be, for example:

def richSortAlternate(aList, eqv):
    def compare(a, b, eqv=eqv):
        return cmp(eqv(a), eqv(b))
    aList.sort(compare)
    return aList

and it would operate in-place.  But I would strongly
advise the other approach...!


In any case, developing the 'eqv' data-extraction is
the key step, and also the one depending strongly on
the specific problems you want to solve, while the
ways you apply that function to the sorting tasks are
general-purpose Python patterns/idioms.


Alex





From gregj at pobox.com  Sun Dec 31 00:02:22 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Sun, 31 Dec 2000 05:02:22 GMT
Subject: HTML Parser
References: <mailman.978224102.28797.python-list@python.org> <n1z36.208621$U46.6407059@news1.sttls1.wa.home.com>
Message-ID: <y5z36.208644$U46.6407855@news1.sttls1.wa.home.com>

"Greg Jorgensen" <gregj at pobox.com> wrote:

> rx = re.compile('(<.*?>)', re.MULTILINE)

Oops -- that should be:

    rx = re.compile('(<.*?>)', re.DOTALL)

That makes the . match the newlines. You need that because HTML tags can
span lines.

--
Greg Jorgensen
PDXperts
Portland, Oregon, USA
gregj at pobox.com




From Bjoern.Giesler at gmx.de  Thu Dec 14 10:11:01 2000
From: Bjoern.Giesler at gmx.de (Bjoern Giesler)
Date: 14 Dec 2000 15:11:01 GMT
Subject: Why no shared libpython by default?
Message-ID: <91anu5$bmd$1@news.rz.uni-karlsruhe.de>

Hi,

why isn't libpython built as a shared library by default? Yes, I know how to
set it up as a shared library, and it's not that much of a hassle, but why
not as standard?

Regards,
		--Bj?rn


From michael at stroeder.com  Sun Dec 31 08:36:49 2000
From: michael at stroeder.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Sun, 31 Dec 2000 14:36:49 +0100
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <slrn94smds.150.grey@teleute.rpglink.com> <FOt36.17$vd1.2224@e420r-atl2.usenetserver.com> <slrn94t1vj.150.grey@teleute.rpglink.com> <mqw36.50$9s1.10042@e420r-atl2.usenetserver.com> <slrn94tqf3.150.grey@teleute.rpglink.com>
Message-ID: <3A4F3671.AAA18BFF@stroeder.com>

Steve Lamb wrote:
> 
>     Whatever.  *plonk*

While you're at it: Can you please *plonk* me too? We might disagree
in the future and we can avoid flame wars in advance and save work.
And it's a great honor for me to be in the same kill-file as the
Eff-bot.

Your cooperation by *plonking* me is highly appreciated. Thank you.

Ciao, Michael.


From dsh8290 at rit.edu  Tue Dec 12 19:22:26 2000
From: dsh8290 at rit.edu (D-Man)
Date: Tue, 12 Dec 2000 19:22:26 -0500
Subject: DangerousExceptions (was Re[2]: [Q] Are Exceptions used that much in practice?)
In-Reply-To: <196647430.20001213004339@vega.bg>; from spahievi@vega.bg on Tue, Dec 12, 2000 at 17:43:39 -0500
References: <3A35D546.A4731B5B@enteract.com> <914mfd02ibt@news2.newsguy.com> <Q4vZ5.19457$x6.10649942@news2.rdc2.tx.home.com> <196647430.20001213004339@vega.bg>
Message-ID: <20001212192226.E16853@westsidecnc.rh.rit.edu>

That might be nice sometime.  Java has it.

-D

On Tue, 12 Dec 2000 17:43:39 Niki Spahiev wrote:
 | 12.12.2000, 21:34:08, Rainer Deyke wrote:
 | 
 | RD> "Alex Martelli" <aleaxit at yahoo.com> wrote in message
 | RD> news:914mfd02ibt at news2.newsguy.com...
 | >> (Some might prefer a wider 'except: return 0' clause, but
 | >> most often it's better to be specific in catching exceptions).
 | 
 | RD> I would go further and say that 'except: return 0' should never be used,
 | RD> since it would catch and ignore MemoryError, KeyboardInterrupt,
 | SystemExit,
 | RD> ComputerExplodingError, or any number of other exceptions that are
 | totally
 | RD> unrelated to the issue at hand and require careful handling and/or
 | program
 | RD> termination.
 | 
 | While at this i always wished that Exception has two subclasses:
 | "NormalException" and "DangerousException" so one can use
 | 
 | except NormalException: return 0
 | 
 | and not worry about  MemoryError, KeyboardInterrupt, SystemExit,
 |  ComputerExplodingError :) being improperly handled.
 | 
 | 
 | -- 
 | Best regards,
 |  Niki Spahiev
 | 




From scj at mcom.mcom.fr  Wed Dec  6 06:36:12 2000
From: scj at mcom.mcom.fr (Alfred)
Date: Wed, 06 Dec 2000 11:36:12 GMT
Subject: Sort
References: <NlnX5.1444$yv1.2700363@nnrp5.proxad.net> <oa8s2tsls5foll62r8fpa6nlcpisb6v1ek@4ax.com>
Message-ID: <MwpX5.1482$0S1.2947186@nnrp5.proxad.net>

Merci beaucoup !!!!
(Thank you !!!!)

Greg




From leejor at my-deja.com  Tue Dec 12 18:19:26 2000
From: leejor at my-deja.com (leejor at my-deja.com)
Date: Tue, 12 Dec 2000 23:19:26 GMT
Subject: Spinning off time intensive procedures to speed CGI
Message-ID: <916bpn$a6i$1@nnrp1.deja.com>

I have a variety of time consuming proceedures that slow
the response time of my CGI scripts. These are
proceedures that do not send any information back to the
web browser. I want to spin off these proceedures to
allow the original program to imediately continue.
Examples of these proceedures include: saving to Log
files, sending emails, or caching database queries.

How do I accomplish this? I have played with fork() and
os.system(), but have not achieved the results I need. In
my searches of the documentation and online forums, I
have only found methods that _wait_ for the spun off
proceedures to return a result.

I don't want to wait.

Thanks!

--
Lee A. Joramo                      ljoramo at nickads.com
The Nickel Want Ads                www.nickads.com
Internet Manager                   970-242-5555




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


From teyc at bigfoot.com  Wed Dec 13 22:28:21 2000
From: teyc at bigfoot.com (Chui Tey)
Date: Thu, 14 Dec 2000 13:28:21 +1000
Subject: Python for REAL Projects ?
References: <3a380d1e.409238@bendigo.news.origin.net.au>
Message-ID: <919erq$tsr$1@bunyip.cc.uq.edu.au>

Delphi is a development environment and you should really compare

Python vs Pascal,
Delphi vs PythonWorks, WingIDE






From loewis at informatik.hu-berlin.de  Fri Dec  8 14:46:00 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 08 Dec 2000 20:46:00 +0100
Subject: SAX2 XML validating?
References: <3A2BF45A.246B757F@marge.hologic.com> <90ih81$hj9$1@reader1.imaginet.fr>
Message-ID: <p6qwvdak87b.fsf@informatik.hu-berlin.de>

"Gilles Lenfant" <glenfant at equod.com.nospam> writes:

> AFAIK, the only one XML validating parser for Python (100% python) is Lars
> Marius Garshol's xmlproc
> http://www.stud.ifi.uio.no/~lmariusg/download/python/xml/

Indeed. It is also available as part of PyXML,
http://sourceforge.net/projects/pyxml, which can be safely installed
on top of Python 2.

Regards,
Martin


From krussell4 at home.com  Tue Dec  5 22:33:25 2000
From: krussell4 at home.com (Kevin Russell)
Date: Wed, 06 Dec 2000 03:33:25 GMT
Subject: mp3 module
References: <3A2BB89B.D4C30300@ira.uka.de>
Message-ID: <3A2DB330.64D341AD@home.com>

Markus von Ehr wrote:
> 
> Hi,
> 
> is there a module to play mp3 files with python?
> 
> Thanks for answers,
> 

There are a few different modules that can read and play MP3 files
(I'd recommend Snack, but I'm biased) or that offer very nice
front-ends to existing MP3 players on various platforms.  Check
out the Sound/Audio section of the Vaults of Parnassus:

	http://www.vex.net/parnassus/

-- Kevin


From jakobsen at home.com  Fri Dec 29 23:54:48 2000
From: jakobsen at home.com (Peder Jakobsen)
Date: Sat, 30 Dec 2000 04:54:48 GMT
Subject: Is 100,000 entries big for a dictionary?
References: <92iug4$vqb$1@nnrp1.deja.com> <92j54s$t0i$1@reader1.imaginet.fr>
Message-ID: <sUd36.46257$59.13186668@news3.rdc1.on.home.com>

Now try

print megadict

and see how long that takes ;)

"Gilles Lenfant" <glenfant at equod.com.nospam> wrote in message
news:92j54s$t0i$1 at reader1.imaginet.fr...
> Not for my desktop Sun 4500 with 16 Gb RAM 8))
> For your machine:
>
> megadict = {}
> for x in xrange(0, 100000):
>   megadict[str(x)] = 'depends on the stuff you put in the dict'
>
> So what's the result ?
>
> <rturpin at my-deja.com> a ?crit dans le message news:
> 92iug4$vqb$1 at nnrp1.deja.com...
> > I'm in the process of architecting an application with medium sized data
> > sets. Python dictionaries are tempting as part of the representation
> > mechanism. How do they perform when they have 100,000 entries? Can you
> > offer any other experience with this size data set in Python?
> >
> >
> > Sent via Deja.com
> > http://www.deja.com/
>




From hzzhang at hotmail.com  Fri Dec  8 11:56:55 2000
From: hzzhang at hotmail.com (Hongzhen Zhang)
Date: Fri, 08 Dec 2000 11:56:55 -0500
Subject: Jpython not work in Win2000
Message-ID: <3A3112D7.20504@hotmail.com>

Hi,
I just installed Jpython on Windows 2000. After successful
installation, under dos prompt, I type:
jpython
it complains that 'jpython' is not recognized as an internal
or external command, operable program or batch file.
The jpython is a executable jar file, according to the file
property. How can that happen? But it works in Windows NT.
Any suggestions and recommendations are highly welcome.

zhang



From peter at engcorp.com  Tue Dec 12 23:49:54 2000
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 12 Dec 2000 23:49:54 -0500
Subject: How can I use printer port in python?...
References: <90vnol$l5p$1@news1-2.kornet.net> <3A33C429.FE4EF79C@engcorp.com> <912l22$gv$1@news1-2.kornet.net>
Message-ID: <3A36FFF2.CDED9271@engcorp.com>

Seo ji won wrote:
> 
> "Peter Hansen" <peter at engcorp.com> wrote:
> > We could assume you mean a parallel port, and that you are on Windows,
> > and that you don't actually plan to talk to a printer here, but then
> > again we could be wrong.  Help us out!
>
> Oh, I'm using Win2k(server..) and using python 2.0,
> and want to use parallel port...
> 
> I looked at those archives which you suggested, but there's no direct
> answer..

Quite true, but it gave us the needed shared reference for a discussion.
:)

> -- or  I  don't know if its answer :)
> 
> Do I have to use c code to use parallel port on windows?..
> 
> If I have to use c, just give me brief sample for it... ;-)
> 
> if not, tell me how can I deal with parallel port with python...

Unfortunately, you haven't quite answered the question fully: do you
plan to talk to a printer or some other kind of device where you want
directly to control the pins in the parallel port?  

If the latter, what kind of control do you need?  Just the eight data
pins?  Any of the handshaking pins?  Are you transferring data in and
out (bidirectional) or just controlling some device without feedback? 
All these things are rather relevant in providing a correct answer to
the question.

In any case, your use of W2K is probably a big problem if you want the
latter.  As with NT (and I'm assuming here, since I've never used W2K
and hope not to), you cannot directly control the hardware.  You would
have to go through the operating system's API or use a driver and wrap
that with appropriate code (quite possibly with C code, depending on
what you really want here).  All these areas quickly go beyond any
experience I've had, so I doubt I can "help" further...


From erno-news at erno.iki.fi  Tue Dec 26 13:02:48 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 26 Dec 2000 20:02:48 +0200
Subject: python telnet script for openvms host
References: <92ainp$jab$1@nnrp1.deja.com>
Message-ID: <kuito7rrg7.fsf@lasipalatsi.fi>

Ennis Rosamond <ennisr at my-deja.com> writes:

| tn.read_until('[c')    #last char of session hearder
| tn.write('$ dir\r')

| #following this, I receive back the sequence:
| #recv '\377\373\001'
| #IAC WILL 251
| #recv '\377\373\003'
| #IAC WILL 251
| #recv '\003\\\033Z'
| #
| #which I don't know how to interpret.

that looks like telnet option negotiation, telnetlib should
take care of all that automaticalyl for you. i suspect
the above is just some debug output from telnetlib.

are you sure you are seeing it after your last .write call,
and not at the start of the telnet connection? 

  -- erno


From josh at open.com  Tue Dec  5 01:40:17 2000
From: josh at open.com (Joshua Muskovitz)
Date: Tue, 5 Dec 2000 01:40:17 -0500
Subject: Calculating factorial
References: <3a2a9496.647995@uutiset.nic.fi>
Message-ID: <3a2c8c2a_3@corp.newsfeeds.com>

What about using xrange() instead of range() to support *really* large
numbers?

-- josh




-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----


From rodsenra at correionet.com.br  Sat Dec  9 08:40:16 2000
From: rodsenra at correionet.com.br (Rodrigo Senra)
Date: Sat, 09 Dec 2000 11:40:16 -0200
Subject: About Search engines
References: <udn1eluhzo.fsf@yogi.pc1.uni-duesseldorf.de> <roy-9B2D39.19561027112000@news.panix.com> <tcmay-98F57B.17064527112000@news.got.net> <udvgt8xr1t.fsf@yogi.pc1.uni-duesseldorf.de> <tcmay-849C4E.19535003122000@news.got.net> <udwvddompp.fsf@box.home.de> <tcmay-BFA4BD.22543408122000@news.got.net>
Message-ID: <3A323640.65FBBE3B@correionet.com.br>

Tim May wrote:
> 
> Don't be misled by this claim. While it is claimed that Metacrawler
> searches N different other search engines, it often (usually) returns
> fewer results than Google does.

> In any case, the moral is to try many different engines...

I recommend trying HuskySearch (http://huskysearch.cs.washington.edu) 

It uses different search engines
(including google) with a smart result combination strategy.
HTH
Rod


BTW, I know this is not related to Pyhton. But since the issue
came up in the thread 'simmulated annealing' in this list, I
thought this info would benefit the majority.

-- 
Rodrigo Senra         
Computer Engineer   (GPr Sistemas Ltda)  rodsenra at correionet.com.br 
MSc Student of Reflection (IC- UNICAMP) Rodrigo.Senra at ic.unicamp.br
http://www.ic.unicamp.br/~921234 (see also   http://www.gpr.com.br)




From cjensen at bioeng.ucsd.edu  Fri Dec  8 16:26:16 2000
From: cjensen at bioeng.ucsd.edu (Curtis Jensen)
Date: Fri, 08 Dec 2000 13:26:16 -0800
Subject: Parallel Computing
References: <3A300B1D.223630F3@bioeng.ucsd.edu> <m3ofynqilc.fsf@chinon.cnrs-orleans.fr>
Message-ID: <3A3151F8.7149AE05@bioeng.ucsd.edu>

Konrad Hinsen wrote:
> 
> Curtis Jensen <cjensen at bioeng.ucsd.edu> writes:
> 
> > I haven't had much luck on finding anything about parallel computing via
> > Python.  Does Python support this?  Or do we just need to write a C
> > module to run parallel code?  Thanks.
> 
> It depends on what you mean exactly by "parallel code", but I suppose
> that for most interpretations the answer is that you can do it in
> Python using existing modules.
> 
> I'll limit myself to the kind of parallel computing that I do myself,
> which is algorithm parallelization implemented via MPI. Version 2.1 of
> ScientificPython (available at
> ftp://dirac.cnrs-orleans.fr/ScientificPython/) contains an interface
> to the most commonly used MPI functions plus the necessary
> infrastructure to make Python run as an MPI application, and to permit
> C extension modules to use MPI as well without linking to it
> (important if you use dynamic libraries). With this package, you can
> use MPI in Python in exactly the same way you would do in C or
> Fortran. The data objects that can be passed between processes are
> NumPy arrays and strings.
> 
> However, most real-life parallel code will also contain C code for
> speed; you wouldn't use a parallel machine if speed weren't important.
> But you can end up with an efficient program in which most code is
> Python. For an example, see the development releases (2.1.x) of the
> Molecular Modelling Toolkit (available at
> ftp://dirac.cnrs-orleans.fr/MMTK/).
> 

Sorry I suppose I should have been more specific in what I was asking. 
I was looking for something like what Mr. Hinsen suggested.  Though, I
am also interested in RPC;  I didn't expect to get responces about
that.  Glad I did.

bye the way, the above links are a bit off. I believe they should be:
ftp://dirac.cnrs-orleans.fr/pub/ScientificPython/
and
ftp://dirac.cnrs-orleans.fr/pub/MMTK/

Thanks.

-- 
Curtis Jensen
cjensen at bioeng.ucsd.edu
http://www-bioeng.ucsd.edu/~cjensen/
FAX (425) 740-1451


From thomas at xs4all.net  Mon Dec  4 12:03:54 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Mon, 4 Dec 2000 18:03:54 +0100
Subject: Python mailing list/usenet bridge.
In-Reply-To: <200012041005.EAA23178@x56.deja.com>; from sbrunning@bigfoot.com on Mon, Dec 04, 2000 at 10:05:53AM +0000
References: <8vt9t4$r0f$1@nnrp1.deja.com> <900365$fp6$1@nnrp1.deja.com> <905t5s$6ef$1@nnrp1.deja.com> <200012041005.EAA23178@x56.deja.com>
Message-ID: <20001204180354.C4396@xs4all.nl>

On Mon, Dec 04, 2000 at 10:05:53AM +0000, Simon B. wrote:

> *Still* nothing. This is getting a bit irritating - deja is horrid.
> *Please*, anyone know who I can nag to get this sorted?

Noone. The only people we can nag to have been nagged to already, and they
can't fix it (except in the long term, which is what Barry *is* fixing.) If
you want read-access, though, you can use 'tyrannewsaurus.xs4all.nl' as your
newsserver. It's the publically accessible newsserver my company hosts, as a
test-bed. We don't charge, we don't expect you to fill in forms, and we
don't use your ipadress for anything. My first posting that said this
attracted 500+ concurrent connections, so I'm wondering howmuch more this
one is going to attract :-)


-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From aahz at panix.com  Thu Dec 28 12:40:27 2000
From: aahz at panix.com (Aahz Maruch)
Date: 28 Dec 2000 09:40:27 -0800
Subject: Q: Is there any good documentation on MAPI using Python
References: <wpm06.791$AH6.82177@newsc.telia.net> <zIK26.3737$4c.280470@ruti.visi.com>
Message-ID: <92ftub$lev$1@panix3.panix.com>

In article <zIK26.3737$4c.280470 at ruti.visi.com>,
Grant Edwards <grante at visi.com> wrote:
>
>I'm using a Python program to fetch e-mail messages (and
>associated attachements) via MAPI.  You're welcome to it:
>
>  ftp://ftp.visi.com/users/grante/python/mfetch.py

<snicker>  So how's the BOFH these days?

(For those walking in late, Grant wrote that program when his BOFH shut
down POP3 access to the Exchange Server and then shut down HTTP access
(or was it the other way around?).)
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

Ninth Virtual Anniversary: 3 days and counting


From schliep at Octopussy.MI.Uni-Koeln.DE  Sun Dec 10 08:59:41 2000
From: schliep at Octopussy.MI.Uni-Koeln.DE (Alexander Schliep)
Date: 10 Dec 2000 14:59:41 +0100
Subject: Dijkstra's Shortest Path algorithm
References: <Pine.GSO.4.21.0012090042370.15110-100000@y.glue.umd.edu>
Message-ID: <tq91yvgjs1e.fsf@Octopussy.MI.Uni-Koeln.DE>

Roy Katz <katz at Glue.umd.edu> writes:

> I was wondering if Python (or Numerical Python) has anything like this
> (a Graph module??).  If there isn't, perhaps I should post my class at

<shameless plug> 
There is a Graph class among a lot of other things in Gato (graph
animation toolbox), see <http://www.zpr.uni-koeln.de/~gato>.

Gato allows you to implement algorithms and have their action
visualized. BFS/DFS algorithms are bundled. You can easily code 
your own algorithms complete with visualization.

Next year there will be a Springer textbook bundled with Gato,
although Gato itself is released freely under the LGPL.  
</shameless plug>

Yours,
Alexander

PS: I am the principal author of Gato

-- 
Alexander Schliep                    schliep at zpr.uni-koeln.de
ZPR/ZAIK                             Tel: +49-221-470-6040 (w) 
University of Cologne                FAX: +49-221-470-5160
Weyertal 80                          http://www.zpr.uni-koeln.de/~schliep
50931 Cologne, Germany               


From  at home.com  Sat Dec 23 00:58:10 2000
From:  at home.com (John Q. Public)
Date: Sat, 23 Dec 2000 05:58:10 GMT
Subject: PythonWin Question
Message-ID: <3A443EF1.125426ED@home.com>

While trying to add a journaling mechanism to python, I came across an
odd behavior from PythonWin when I tried to override sys.stdout.  For
both the console and IDLE versions of the interpreter, I can override
sys.stdout, but not for PythonWin.

IDLE:

>>> import journal
>>> sys.stdout
<PyShell.PseudoFile instance at 01028F9C>
>>> journal.Open("My_Journal.txt")
>>> sys.stdout
<journal.Journal instance at 010388CC>
>>> journal.Close()
>>> sys.stdout
<PyShell.PseudoFile instance at 01028F9C>

console:

>>> import journal
>>> sys.stdout
<open file '<stdout>', mode 'w' at 00762C80>
>>> journal.Open("My_Journal.txt")
>>> sys.stdout
<journal.Journal instance at 0079285C>
>>> journal.Close()
>>> sys.stdout
<open file '<stdout>', mode 'w' at 00762C80>

PythonWin:

>>> import journal
>>> sys.stdout
<pywin.framework.interact.InteractiveView instance at 0114E01C>
>>> journal.Open("My_Journal.txt")
>>> sys.stdout
<pywin.framework.interact.InteractiveView instance at 0114E01C>
>>> journal.Close()
No open journals!
>>> sys.stdout
<pywin.framework.interact.InteractiveView instance at 0114E01C>

Any thoughts on how I can get PythonWin to "behave?"

Thanks!

Jonathan Polley

jwpolley(a)collins(d)rockwell(d)com




From pbjorn at uswest.net  Fri Dec 22 13:40:06 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Fri, 22 Dec 2000 11:40:06 -0700
Subject: ANNOUNCE: soundex.py
References: <mailman.977338109.32467.python-list@python.org> <3A4119C3.2E5E54E1@uswest.net> <mailman.977501169.22862.python-list@python.org> <slrn94787k.1q7.amk@207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com>
Message-ID: <3A43A006.F2CCC36@uswest.net>

If we could get the soundex.py as part of the standard library, this
would definitely get my vote.

-- bjorn

"A.M. Kuchling" wrote:

> Does this mean that Modules/soundex.c can be dropped from the Python
> CVS tree?
>
> --amk



From erno-news at erno.iki.fi  Sun Dec 24 02:26:36 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 24 Dec 2000 09:26:36 +0200
Subject: Some Python 2.1 ideas
References: <mailman.977590932.19335.python-list@python.org>
Message-ID: <kuelyyi8k3.fsf@lasipalatsi.fi>

>>>>> "Bob" == Bob Alexander <bobalex at home.com> writes:

[snip]

| ---- String method to remove trailing newline

i think it would be prettier and more useful to have methods
to remove a given prefix/suffix from the beginning or the end of a
string. maybe add optional arguments for lstrip/rstrip, or
introduce new methods rtrim/ltrim?

   -- erno


From moshez at zadka.site.co.il  Tue Dec 12 10:17:32 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Tue, 12 Dec 2000 17:17:32 +0200 (IST)
Subject: Inheriting the @ sign from Ruby
Message-ID: <20001212151732.5D5A2A82E@darjeeling.zadka.site.co.il>

On Tue, 12 Dec 2000, Roy Katz <katz at Glue.umd.edu> wrote:

> Why not bind @ to the first argument of method definitions? Consider:
> 
> 
>   class X:
> 
>     def __init__( self, x ):
> 
>       @bind(x)     # binds to 'self', because 
>                    # 'self' is the first argument to __init__
> 
>     def bind(myself, x):
>       @x = x   # equivalent to myself.x

If you're requiring the "@" anyway, why not write the self. and be 
maximally obvious?

Explicit is better then implicit.
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From rjroy at takingcontrol.com  Mon Dec  4 17:39:17 2000
From: rjroy at takingcontrol.com (Robert Roy)
Date: Mon, 04 Dec 2000 22:39:17 GMT
Subject: Dividing the screen in text mode?
References: <90eh84$3u0$1@bob.news.rcn.net> <90ehc9$50p$1@bob.news.rcn.net>
Message-ID: <3a2c1cc3.417918375@news1.on.sympatico.ca>

On Sun, 3 Dec 2000 17:27:24 -0500, "Z 3 Penguin"
<z3penguin at penguinpowered.com> wrote:

>oh yeah, i'm using py2.0 in win32.
>

>From an annoucement by Fredrik a couple of weeks ago:

The Console module is a simple console driver, currently available
for Windows 9X/NT/2000.  This module provides cursor-addressable
text output, plus full support for mouse and keyboard input.

The 001114 release comes with source code, prebuilt binaries for
Python 1.5 and 2.0, documentation, and a Python-style license.

Get your copy from:

    http://effbot.org/efflib/console

enjoy /F



From MarkH at ActiveState.com  Tue Dec  5 00:05:50 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Tue, 05 Dec 2000 05:05:50 GMT
Subject: Python And MS Word Com Server
References: <90bc4c$rhr$02$1@news.t-online.com> <90g54c$chj$1@news.inet.tele.dk>
Message-ID: <3A2C7661.7010404@ActiveState.com>

Jesper Hertel wrote:

> I have exactly the same problem, but no solution, unfortunately.
> 
At the top of the generated file makepy creates, you will find:

# The following 3 lines may need tweaking for the particular server
# Candidates are pythoncom.Missing and pythoncom.Empty
defaultNamedOptArg=pythoncom.Missing
defaultNamedNotOptArg=pythoncom.Missing
defaultUnnamedArg=pythoncom.Missing


Try changing those 3 lines to "Empty"

Mark.



From aleaxit at yahoo.com  Tue Dec 12 07:05:37 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 12 Dec 2000 13:05:37 +0100
Subject: Print whithout \n
References: <aboZ5.3538$Kd1.410322@newsb.telia.net>
Message-ID: <9154aq02st0@news2.newsguy.com>

"Henrik M?rtensson" <henrik.martensson at frontconsulting.se> wrote in message
news:aboZ5.3538$Kd1.410322 at newsb.telia.net...
> Hi
>
> How do you print a string whithout a linefeed in the end.

With a trailing comma on the print statement:

    print "no linefeed here",
    print "nor here","here neither",


Alex





From cbarker at jps.net  Tue Dec 26 19:11:00 2000
From: cbarker at jps.net (cbarker at jps.net)
Date: Wed, 27 Dec 2000 00:11:00 GMT
Subject: Homogeneous Sequences in Python?
Message-ID: <92bc2i$72n$1@nnrp1.deja.com>

Homogeneous Sequences in Python?


It seems that we have recently has yet another discussion about what
kind of performance improvements might be made to the Python
interpreter. It seems that there is some consensus that there are
perhaps a lot of little things that could be done for special cases, but
the big "bang for the buck" stuff has been done. For the most part, I
find that Python is fast enough as it stands for most of my needs, but
that it does fall down flat for a lot of what I do: numerical computing.

A lot of the performance hit comes from what it know to be a limitation
of virtually all interpreted languages: slow looping. While this is
true, I think Python suffers more because of it's extreme dynamicism.

The classic answer to why Python cannot be easily made faster, even if
it were compiled, is that because of Python's dynamic nature, an
expression like "x = a + b" cannot be compiled, because it is not known
until runtime what types a and b are, and what it means to __add__
them. I am a BIG fan of Python's dynamic nature, so I would not like to
see any of that flexibility go away.

I have seen proposals for "optional static typing" that would address
this issue (I believe Guido is talking about it for Py3k). That would
certainly help, but it could be somewhat ugly as well. What occurs to me
are two observations:

1) the occasional staticly typed integer is not going to make much
difference. What would make a difference is a large collection of
integers, for example

2) More often than not, when I build a large sequence, it is
homogeneous, that is a large list (or tuple, whatever) of the same type:
maybe a list of strings ( file.readlines() ), a list of integers
(range(n) ), or even a list of lists.

So, what we need to speed things up is to have a way for Python to
process large sequences of the same type without having to type check
each item in the sequence.

The simple way to implement this is to have an optionally static type
declaration for sequence: list.is_list_of_integers, or something like
that. Then when this list is processed, the interpreter would know
what it was dealing with. It then struck me that it could certainly be
done automatically, on-the-fly: Every time an item was added to a
sequence, it's type would be checked, and some kind of "homogeneous
flag" would be set:

a = [] # type unknown

a.append(4) # homogeneous list of integers

a.append(56) # still a homogeneous list of integers

a.append(4.5) # Not homogeneous

I have not idea how much this would slow append operations, but I
suspect much less than the potential speed up of know that a list is
homogeneous when it is. A plus of doing this automatically, is that it
would take NO changes to existing code to take advantage of the new
feature.

Would this really speed up Python code much, even for large homogeneous
sequences? I have no idea. I am in no way qualified to make that
judgment, I'd like to hear what folks think.

I suspect that it won't speed up straight, run-through-the-interpreter
code enough to make it worth it, but it could have some real benefits in
three places:

1) some of the sequence oriented operations that are being introduced:
list comprehensions, and the possible future Elementwise/Objectwise
Operators (see PEP # 225)

2) Py2c, the Python to C translator. I think a really important element
of Python's future may be Py2C (and perhaps eventually a Python compiler
or JIT compiler). It seems that has been the direction of number of
interpreted languages. With optional static typing and compilation,
LISP can be as fast (or faster) that C++ for numerics. MATLAB now has a
translator to C++ that can create faster code and fully compiled stand
alone applications. That has almost eliminated the need to write
external functions in C. Py2C can only be so fast without some kind of
static typing. I think homogeneous sequences could do it. Right now Py2C
can convert "for i in range(1000)" into a fast C loop, but can't do
anything with "for i in list_of_integers", If it had a way of knowing
that it was a list of integers, it could be highly optimized.

3) Extensions that you have to pass a lot of data to. An example is
wxPython. In order to draw a polyline with a couple of thousand points,
for example, you have to pass a list of 2-tuples of integers of the
coordinates of the points. The extension has to do a type check on every
one of the numbers in every one of the tuples in the list. Frankly,
this is still pretty fast, but not as fast as it could be.


My goal is to be able to write applications that contain some
numerically intensive code (or other repetitious simple processing) and
not to have to use anything but Python. I'm going to have to keep
writing C for a large amount of the core code if something isn't done a
little differently.


What about Numeric?

NumPy provides the Multiarray object, which does provide what I am
talking about, a homogeneous sequence ( at least of numbers, which is
mostly what I need). As a result, it can implement very fast array
oriented operations that eliminate the need for a large amount of hand
coded C extensions. The problem is that core Python has no idea that it
is a homogeneous sequence, so when you have to do some looping, things
slow WAY down.

As an example, I recently wrote a function that had to loop through a
NumPy array of 1000 items, a list of 200 Numpy arrays, and each row in
each of the Arrays in that list. The result was a triple nested loop,
and a very slow run time. I probably could have avoided one of those
loops using array oriented computation, but there was no way to get rid
of all of them. I wrote the functions in the inner part of the loop in
C, and got about a 25% speed up. I then wrote the entire thing in C, and
got a 100X speed up. what had taken 100 seconds, now took 1 second. This
was going to get done a lot, so I needed to do that. Having been working
almost entirely in Python lately, I really didn't enjoy writing the C !


How this could be implemented:

I am not qualified to talk about the actual code in the Python
interpreter, but what I am envisioning is that the concept of the
homogeneous sequence would be get into the Python code, and hopefully
some optimizations of list comprehensions and the like could be done. In
the short term, I would imagine little other benefit, but once it was in
there, it could be used to great advantage in Py2C, and used by
extension writers to really make things easier. In the long term,
homogeneous sequences might provide a basis for the high performance
Python compiler of the future.


My questions:

Is this even a remotely good idea, or have I just gotten carried away?

Could it be useful to dynamically test for homogeneous sequences, or
would they have to be statically defined?

Looking at some of the stuff on the Web about Guido's ideas about static
typing, I see a way to do homogeneous sequences, but no mention of the
concept itself. Is it there somewhere?

What are all the other flaws in my ideas that I haven't come up with
yet?




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


From jp at ulgo.koti.com.pl  Tue Dec 26 10:08:35 2000
From: jp at ulgo.koti.com.pl (Jacek =?iso-8859-2?Q?Pop=B3awski?=)
Date: 26 Dec 2000 15:08:35 GMT
Subject: ~Python ?
References: <3a4833f8.346848@news.t-online.de>
Message-ID: <slrn94hd3j.1br.jp@localhost.localdomain>

Gerson Kurz wrote:
>You can have real fun if you try to deleting an
>actual directory named ~ on Linux, boy. 

hmm... without -rf it's not dangerous... is it?
anyway - you can use \~ or file manager (mc for example)

-- 
They'll take your diamonds, and then give you steel     
You'll be caught in the middle of the madness     
Just lost like them                                 "All The Fools Sailed Away" 
And part of all the pain that they feel                     - Ronnie James Dio


From SBrunning at trisystems.co.uk  Wed Dec  6 12:48:27 2000
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Wed, 6 Dec 2000 17:48:27 -0000 
Subject: parsing and searching big text files
Message-ID: <31575A892FF6D1118F5800600846864D5B155D@intrepid>

> From:	Gabor Gludovatz [SMTP:ggabor at sopron.hu]
> I have a big text file which contains 3 variable lenght columns with
> names. I have to search this file for names in either column and have to
> show to other 2 names, and I have to do this very fast!
> 
> The text file is about 3 megs long. Does someone know a method to do this
> quickly?
> 
> Which is the faster, should I read to whole text file into the memory and
> parse and search it there, or should I read it from the disk line by line
> and parse it..?
> the last seemed to be very slow.
> 
> Here is an example line from this text file:
> Foo Bar:John Doe:Bill
> 
> If I look for, for example, John Doe, the function should return Foo Bar
> and Bill and all the other records which contain John Doe.
> 
> Again: I have to do it very fast.
 
Try this (untested):

def scanGaborsFile(file, scanFor='John Doe', seperator=':'):
    result = []
    try:
        lines = file.read().splitlines()
    except AttributeError:
        lines = open(file, 'r').read().splitlines()
    for line in lines:
        fields = line.split(seperator)
        if scanFor in fields:
            fields.pop(fields.index(scanFor))
            result.append(fields)
    return result

To answer your specific question, reading in the whole file will be by far
the fastest way of doing it, unless you run out of memory.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From gherman at darwin.in-berlin.de  Tue Dec 19 09:27:21 2000
From: gherman at darwin.in-berlin.de (Dinu C. Gherman)
Date: Tue, 19 Dec 2000 15:27:21 +0100
Subject: UN/EDIFACT related packages
References: <91lg8h$2mo$1@reader1.imaginet.fr>
Message-ID: <3A3F7049.29CCAC24@darwin.in-berlin.de>

Gilles Lenfant wrote:
> 
> I'm looking for packages related to UN/EDIFACT EDI messages.
> Yes we got business partners who did not make the XML jump.
> I'm 100% newbie in the EDIFACT field, so if some of you have
> useful related packages/tooks, thei're welcome.
> I did not find throug usual Python related search engines :(

I know that Cees de Groot did some work in that field,
resulting in a package called "ema". Hopefully he will
listen and speak up. 

Long ago I wrote some toy implementation, which will 
not be very helpful, I guess, as my need for this did
almost immediately evaporate after I started to get
interested in EDIFACT. As you mention, the world goes
for XML now!

Dinu

-- 
Dinu C. Gherman
................................................................
"The only possible values [for quality] are 'excellent' and 'in-
sanely excellent', depending on whether lives are at stake or 
not. Otherwise you don't enjoy your work, you don't work well, 
and the project goes down the drain." 
                    (Kent Beck, "Extreme Programming Explained")


From junaftnoon at nospamplzyahoo.com  Thu Dec 28 12:36:44 2000
From: junaftnoon at nospamplzyahoo.com (June Kim)
Date: Fri, 29 Dec 2000 02:36:44 +0900
Subject: sniffing btw my web browser and a server
References: <92aush$rhc$1@news.nuri.net> <92fs3l$5gg$01$1@news.t-online.com>
Message-ID: <92ftpn$e2j$1@news.nuri.net>

"Eduard Hiti" <KatMouse at gmx.de> wrote :

[snip]
>
> There's also a Python only solution: 'WebDebug'. It also acts as a proxy,
> logging all transfers and generating a nice and comprehensive HTML report.
>
> The URL: http://www.cyberclip.com/webdebug/download.html

This one's what I've been looking for, but the link seems broken. Does
anyone have it or know alternative links?

Best regards,

June



From News at Titanic.co.uk  Wed Dec 27 09:59:13 2000
From: News at Titanic.co.uk (Simon Faulkner)
Date: Wed, 27 Dec 2000 14:59:13 GMT
Subject: Controlling the Parallel Port
References: <mvuj4tkmn22gbth1a6t8fag39902qeeu9l@4ax.com>
Message-ID: <1q0k4t41id4d0nii3sofvbbchm44dllfp9@4ax.com>

Having read the previous messages I note that this has been discussed
on the Win platform but I am interested in the Linux (RH7) platform.

I have the Jan Axelson Parallel Port Complete book but I am not sure
how to write the port from Python.

Any code snippets gratefully stolen...

Simon


On Wed, 27 Dec 2000 14:33:36 GMT, Simon Faulkner <News at Titanic.co.uk>
wrote:

>Is it easy to control (read/write) the Parallel Port in linux using
>Python?
>
>I am hoping to use it to drive some simple electronics
>
>(Usual disclaimer understood - If it smokes then you probably got the
>wiring wrong!)
>
>
>Simon
>Simon Faulkner

Simon Faulkner


From alan at littleford.net  Wed Dec 20 01:06:03 2000
From: alan at littleford.net (Alan Littleford)
Date: Tue, 19 Dec 2000 22:06:03 -0800
Subject: Jython stability and JITs
Message-ID: <3A404C4B.CBCA036D@littleford.net>

I am thinking of using Jython for a project -- I know it is Alpha and do
not  have too much trouble with that. However the FAQ seems to go to
great length implying that Jython and JITs spell trouble (because of
"non-standard" JVM code that Jython emits) -- indeed it basically says
turn off your JIT if you want to sleep easy at night.

Is this indeed the case -- have people had trouble with JITs and Jython,
or is this a left over warning from earlier (JPython) days ?

Thanks
Alanl




From fhorch at ecoaccess.org  Mon Dec 18 13:04:12 2000
From: fhorch at ecoaccess.org (Fred Wilson Horch)
Date: Mon, 18 Dec 2000 13:04:12 -0500
Subject: Volunteer internship for Python programmers
Message-ID: <3A3E519C.37004BB7@ecoaccess.org>

Dear colleague,

If you are

 * interested in Python programming or already a programmer,
 * looking for a volunteer opportunity or an independent study project,
 * eager to publish code and work on a development team, and
 * able to invest a few hours a week between January and May 2001
   reading, writing and documenting open source object-oriented web
   application software (specifically Java, C and Python code for Zope
   running on Linux),

you may be interested in the attached announcement.

Thanks for the bandwidth,

Fred Wilson Horch			mailto:fhorch at ecoaccess.org
Executive Director, EcoAccess		http://ecoaccess.org/
P.O. Box 2823, Durham, NC 27715-2823	phone: 919.419-8354
-------------- next part --------------
Spring 2001 EcoAccess Technology Associate Internship

EcoAccess, a public charity working to enhance public understanding
of our natural environment, is offering a volunteer Technology
Associate internship program for people seeking the opportunity to
enhance their software development skills.

Selected Technology Associates will receive a loaner Linux system to
allow them to develop free open source software for the EcoAccess
database-backed web site.

Designed for Associates who have some programming experience,
the internship runs from January 8 through May 4.  Our software is
published on SourceForge and licensed under the General Public
License.

How to Apply

The internship is open to anyone who has any programming experience
or training, with preference to those residing in the Triangle area of
North Carolina.  Apply by e-mail to

 Fred Wilson Horch, Executive Director
 internship at ecoaccess.org

Please include the following information:

 * Your name
 * Preferred e-mail address
 * A summary of your education and relevant work experience, if any
 * Your experience with open source software, if any
 * Your other time commitments between January and May

If we are interested in working with you, we will request a code
sample before making a decision whether to accept you into the
program.

More About the Internship

Our Technology Associates will learn how to develop for Zope, a
leading Open Source web application server.  Part of the internship
will be devoted to learning and improving Zope itself, including
documenting aspects of its interfaces.  The other part will be devoted
to writing applications in Python, Java and C to support the charitable
EcoAccess project.

Technology Associates must have at least dial-up Internet
connectivity.  Four interns who are in the Triangle area of North Carolina
(Raleigh-Durham-Chapel Hill) can receive fully-configured development
machines for use during the internship.  (Associates may also use
their own Linux machines if they prefer.)  The loaner machines are
best described as tomorrow's software running on yesterday's hardware.

Our Technology Associates will get hands-on experience building
object-oriented transactional web applications.  Specifically, our
Associates will write programs in Java, C and Python that run on Linux
and use PostgreSQL databases.

This internship is made possible through the generosity of several
members of the Triangle Linux Users Group (TriLUG) who donated the
hardware for our loaner machines.  We appreciate the help!

(Announcement revision 1.0 2000-12-14 10:21)

From aleaxit at yahoo.com  Mon Dec  4 16:45:56 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 4 Dec 2000 22:45:56 +0100
Subject: Converting Perl Regex to Python!
References: <90h0a8$jgm$1@nnrp1.deja.com>
Message-ID: <90h3dr0i17@news2.newsguy.com>

<thelocust at my-deja.com> wrote in message news:90h0a8$jgm$1 at nnrp1.deja.com...
> I'm working on porting a program from Perl to Python, and i'm to a
> point where I need to convert their Perl regex stuff over to python.
> Now, I understand what is going on in the Perl if statement below, but
> i haven't seen to many good examples of regex stuff for Python, so i'm
> hurting for an translation
>
> The string i need to grab the info from looks like this:
>
> MSG notify timecode [00:00:00:00]
>
> if ($msg =~ /^MSG notify timecode \[(\d\d:\d\d:\d\d):\d\d\]/) {
>     $cajun->{state_clock}=$1;
>     }
>
> So, how exactly do i convert this to python?

import re

mo = re.match(r'MSG notify timecode \[(\d\d:\d\d:\d\d):\d\d\]', msg)

if mo:
    cajun.state_clock = mo.group(1)


Alex






From joconnor at cybermesa.com  Tue Dec 12 13:24:31 2000
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Tue, 12 Dec 2000 18:24:31 +0000
Subject: ooopss...
Message-ID: <3A366D5F.CDBAC92F@cybermesa.com>

>         while (stream.atEnd()):
>                 element = stream.next()
>                 #process element

Oops...Of course that won't work.  I was translating the normal idiom
from Smalltalk, which uses #whileFalse:  in such situations instead of
#whileTrue:

It should read 
                
        while (! stream.atEnd()):
                element = stream.next()

As an aside, is there a Stream library for Python?  Similar to
Smalltalk's? I miss the idiom...If not, would anyone be interested in
having one? I could probably do it as a side project

It's a lot easier to extend a language by adding libraries than by
changing the syntax
-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

"God himself plays the bass strings first when He tunes the soul"



From fredrik at effbot.org  Sat Dec  9 05:21:57 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sat, 09 Dec 2000 10:21:57 GMT
Subject: PIL binary for Python 2.0 for Windows ?
References: <3a2f472f$0$235@hades.is.co.za> <KZOX5.2015$Kd1.294314@newsb.telia.net> <IlSX5.24763$R77.1876538@nnrp4.clara.net>
Message-ID: <9JnY5.2716$O%.428568@newsc.telia.net>

"Richard" wrote:
> I have downloaded the dll & the PIL tar file Imaging-1.1.1.tar.gz.
>
> Please can you tell me where these file should be copied? (My Python 2.0 is
> istalled under C:\Python20).

put them anywhere on your Python path (if you don't
know how Python searches for modules, look it up in
the Python docs).

also see the updated pil page:

    http://effbot.org/pil

</F>




From mksql at my-deja.com  Wed Dec  6 15:33:35 2000
From: mksql at my-deja.com (Matt)
Date: Wed, 06 Dec 2000 20:33:35 GMT
Subject: mxODBC + MSSQL = code -9 error (again)
Message-ID: <90m7qv$rj2$1@nnrp1.deja.com>

I am just starting to use the mxODBC package, and have come across an
issue that has been posted in the newsgroups before, but I have not yet
found a solution.

I am using ActiveState Python 2.0 "ActivePython 2.0, build 202
(ActiveState Tool Corp.) based on Python 2.0 (#8, Oct 19 2000,
11:30:05) [MSC 32 bit (Intel)] on win32", on a Windows 2000
Professional PC, and attempting to connect to a Microsoft SQL Server
7.0 database, using a simple test script adapted from your
documentation. Thes test script, included below, works fine when
connecting to an Access database, so I am assuming my installation is
valid.

But, when I connect to a SQL Server database, and have a result set
that includes a Unicode character data  (SQLDMO_DTypeUVarchar -9
Variable length) type, I get the "InterfaceError: SQL type (code -9)
not implemented" error. This is avoidable since my tables to not use
Unicode fields, but apparently some of the SQL Server system tables do.

The script below, when run against a MS Access DSN, properly returns a
list of tables. But it appears the same call against the MS SQL DSN
returns Unicode data in the result set (in c.tables).

***

import ODBC
db = ODBC.Windows.Connect('SQLTest','sa')
c = db.cursor()
c.execute('select count(*) from testdata')
c.fetchone()
c.tables(None,None,None,None)
ODBC.print_resultset(c)
c.close()
db.close()

Traceback (most recent call last):
  File "d:\python20\pythonwin\pywin\framework\scriptutils.py", line
301, in RunScript
    exec codeObject in __main__.__dict__
  File "D:\Python20\Lib\ODBC\Misc\sqltest.py", line 7, in ?
    ODBC.print_resultset(c)
  File "d:\python20\lib\ODBC\ODBC.py", line 75, in print_resultset
    lines = apply(format_resultset,(cursor,headers),kws)
  File "d:\python20\lib\ODBC\ODBC.py", line 33, in format_resultset
    rows = cursor.fetchall()
InterfaceError: SQL type (code -9) not implemented


Sent via Deja.com http://www.deja.com/
Before you buy.


From MarkH at ActiveState.com  Sat Dec 30 19:14:47 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sun, 31 Dec 2000 00:14:47 GMT
Subject: msxml3 and IStream interface
References: <QV_26.228$Io4.17256@news.siol.net> <3A4D3814.6090201@ActiveState.com> <uvgs2aayn.fsf@ctwd0143.fitlinxx.com>
Message-ID: <3A4E7956.5060903@ActiveState.com>

David Bolen wrote:

> Mark Hammond <MarkH at ActiveState.com> writes:

> In a similar vein (already known interfaces versus not), I'm beginning
> to look/work more with the server side of things, and was wondering
> how one might go about using win32com to produce a COM server using a
> non-IDispatch interface, where that interface wasn't one of the
> predefined ones in pythoncom.
> 
> Your book doesn't seem to really touch the construction of
> non-IDispatch interfaces - is the most practical approach for me to
> build a typelib via standard means (compiling an IDL), then use makepy
> to define it for win32com, which then would let me declare it in the
> _com_interfaces entry in my class?

This will still only allow Python to implement it as (basically) an 
IDispatch based interface.  To implement a true vtable interface like 
IStream, you really hav 2 choices:
* Implement your own extension DLL, like win32comext\*.pyd
* Dig into "univgw", a tool that is 85% of the way to being able to 
implement arbitary vtable interfaces given a TLB definition.

> And if I'm intending to use these over DCOM and might involve
> non-standard marshaling code, is there a way to make that code
> available to win32com?

I understood that interfaces registered their own marshalling proxies 
explicitly in the registry, and that the COM framework itself would use 
these behind the scenes for you.  Of course, you still need to implement 
that marshalling code.

> Book and/or code pointers are fine (I've saved various COM stuff from
> you and Alex in the past which I'm still also going through).

I'm afraid you are already into "use the source" territory!

Mark.



From aleaxit at yahoo.com  Thu Dec 28 05:01:55 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 28 Dec 2000 11:01:55 +0100
Subject: A Suggestion for Python Colon Syntax
References: <mailman.977974330.23639.python-list@python.org>
Message-ID: <92f35001cbf@news1.newsguy.com>

"Tim Peters" <tim.one at home.com> wrote in message
news:mailman.977974330.23639.python-list at python.org...
    [snip]
> despite not understanding much.  Thoreau went on to make that principle
the
> basis of his form of poetry:

Isn't Thoreau's poetry a rather minor part of his writings, despite
his beautiful "if men read aright, methinks they would never read
anything but poems" ... "after all, man is the great poet, and not
Homer nor Shakespeare; and our language itself, and the common arts
of life, are his work"?  I always think of him mainly as a thinker,
and writer of essays, rather than a poet.

In a way, I think it's thus fortunate that the following snippet:

> # Local Variables:
> # py-indent-offset: 8
> # End:
> # My Captain:  Oh, My Captain!

if, as you say, it was Thoreau's, ended up inspiring one Walt
Whitman instead.  Lincoln well deserved such a memorial, after
all, and, as a versifier, I think Whitman had it all over
Thoreau (which rather appreciated his poems, isn't it?).


Alex





From steve.waterbury at gsfc.nasa.gov  Mon Dec 11 00:31:10 2000
From: steve.waterbury at gsfc.nasa.gov (Steve Waterbury)
Date: Mon, 11 Dec 2000 00:31:10 -0500
Subject: NIST meeting: "Defining Standard, Scriptable Objects for 
 CAD/CAE/CAM/PDM"
Message-ID: <3A34669E.8FA938D8@gsfc.nasa.gov>

"Defining Standard, Scriptable Objects for CAD/CAE/CAM/PDM"

Date:      December 13th 
Time:      10:00 AM - 4:00 PM
Location:  NIST, Building 225, Room A54
           -> for directions and maps to NIST, see:
http://www.nist.gov/public_affairs/maps/nistmaps.html

* If you plan to come, you MUST contact Michael McLay, so that 
we are sure that the meeting room space is adequate:

   mailto:mclay at nist.gov
   - or call 301-975-4099

Background:

Roughly, PDM ("Product  Data  Management")  does for  CAD/CAE/CAM
models and related  objects  what Zope does for the objects  that
need to be managed in a typical Web site (HTML docs, etc.).

Problem Statement

The problem is to find an open  architecture  for integrating and
managing the configuration of  computer-aided  design,  analysis,
and manufacturing (CAD/CAE/CAM) tools, models, and data.

The purpose of the meeting is to  evaluate  technologies,  try to
define an approximate  roadmap for building an open architecture,
and establish a ROM guess of (1) how much work may be required to
implement  the  architecture  components  and  (2)  the  cost  of
developing and maintaining an instance of the architecture for an
engineering organization.

Agenda:

  * Engineering Tool/Data Integration Requirements
    Steve Waterbury/NASA

  * GenCAM: a Standard API for Electronic Manufacturing Data
    Michael McLay/NIST

  * The MEMS Exchange: Python and ZODB for Management of Complex Objects
    Andrew Kuchling/CNRI

  * The CADScript/IScript Architecture (and Demo)
    Doug Cheney/International TechneGroup, Inc.

  * Python and Zope: Architectures and Directions
    Barry Warsaw/Python Labs

  * Group discussion
    - How to incorporate interfaces to Python and Zope components 
      such as ZODB into a standard PDM architecture
    - What enhancements to ZODB, ZCatalog, ZPatterns, etc., can 
      address functionality and interfaces needed for general PDM
    - Which pieces of the architecture make the most sense as open 
      source products, which ones are most logical as commercial 
      products, and what interfaces should be standardized to 
      enable open source and commercial products to interoperate.  

Cheers,
-- Steve Waterbury
   NASA/Goddard Space Flight Center.

                                           oo _\o
                                            \/\ \
                                              /
____________________________________________ oo _________________
"Sometimes you're the windshield; sometimes you're the bug."
- Knopfler

Stephen C. Waterbury                       Component Technologies
Code 562, NASA/GSFC                  and Radiation Effects Branch
Greenbelt, MD 20771           Engineering Web/Database Specialist
Tel: 301-286-7557                              FAX:  301-286-1695        
WWW:          http://misspiggy.gsfc.nasa.gov/people/waterbug.html
_________________________________________________________________



From tcmay at got.net  Sat Dec  9 19:29:45 2000
From: tcmay at got.net (Tim May)
Date: Sat, 09 Dec 2000 16:29:45 -0800
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu> <slrn92tu7t.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com> <tcmay-8BCFAB.23025408122000@news.got.net> <90uddk$31f@gap.cco.caltech.edu>
Message-ID: <tcmay-4AACBE.16294509122000@news.got.net>

In article <90uddk$31f at gap.cco.caltech.edu>, Nathaniel Gray 
<nospam at caltech.edu> wrote:

> Tim May wrote:
> 
> > I did some work on simulated annealing while at Intel, which is why I
> > commented in the SA thread. 
> 
> Wow.  I hadn't even noticed it was the same person.  This one's even 
> worse. 
>  Just out of curiosity, why did you bother to take the time to do a 
>  google 
> search just to castigate the original poster in that thread?

Notice how you're the one fanning the flames in these threads?

Like many fuzzy do-gooders, you preach "why can't we all just get 
along?", but in actuality you spread poison.


> He didn't ask you to do a search.  He asked if anybody had solved the 
> problem and would be willing to share their code.  He was quite polite, 
> and 
> even explicitly stated that he'd already searched dejanews and 
> python.org.
> What exactly did you take issue with?  Did he use the wrong search engine 
> for your tastes?  

He has expressed happiness that he could find 1200 references, many of 
them applicable, to Python and simulated annelaing by using a better 
search engine.

I didn't see _you_ providing any help, either in that thread or in this 
"Python origins" thread. Your role is apparently just to join in later 
and argue for niceness.

You remind me of Alan Dershowitz.


--Tim May

-- 
---------:---------:---------:---------:---------:---------:---------:----
Timothy C. May              | Crypto Anarchy: encryption, digital money,  
ComSec 3DES:   831-728-0152 | anonymous networks, digital pseudonyms, zero
W.A.S.T.E.: Corralitos, CA  | knowledge, reputations, information markets, 
"Cyphernomicon"             | black markets, collapse of governments.


From gerson.kurz at t-online.de  Tue Dec 26 01:03:40 2000
From: gerson.kurz at t-online.de (Gerson Kurz)
Date: Tue, 26 Dec 2000 06:03:40 GMT
Subject: ~Python ?
Message-ID: <3a4833f8.346848@news.t-online.de>

On Linux, ~ usually refers to the users homedir, but not so in Python
(at least, in ActivePython 2.0). I found out the hard way when using a
script that issued os.makedirs() relative to ~ - It created an actual
directory named ~. You can have real fun if you try to deleting an
actual directory named ~ on Linux, boy. 

I understand that ~ is probably handled by the shell, but if Python
doesn't deal with it, its hard to use Python in some kinds of scripts.
(I think Perl, Pythons ugly twin-brother, does)


From quinn at dinar.ugcs.caltech.edu  Sat Dec 16 21:09:26 2000
From: quinn at dinar.ugcs.caltech.edu (Quinn Dunkan)
Date: 17 Dec 2000 02:09:26 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <r0pzoi11lhz.fsf@black.cs.washington.edu> <5jvZ5.3588$Kd1.418770@newsb.telia.net> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com> <kuu28971kp.fsf@lasipalatsi.fi> <3A380E4E.8537A0FB@san.rr.com> <91a1vr02b02@news2.newsguy.com> <Co9_5.25465$x6.13756579@news2.rdc2.tx.home.com> <91bbq102mtk@news1.newsguy.com> <rWd_5.26820$x6.14119694@news2.rdc2.tx.home.com> <91d5el0me7@news1.newsguy.com> <rOr_5.31120$x6.15241856@news2.rdc2.tx.home.com>
Message-ID: <slrn93o829.tu0.quinn@dinar.ugcs.caltech.edu>

On Fri, 15 Dec 2000 16:38:47 GMT, Rainer Deyke <root at rainerdeyke.com> wrote:
>Yes, in-place modification should be supported.  I had something like this
>in mind:
>
>class UserInt:
>  def __init__(self, value):
>    self.data = value
>  def __add__(self, other):
>    if isinstance(other, UserInt):
>      return self.__class__(self.data + other.data)
>    else:
>      return self.__class__(self.data + other)
>  __radd__ = __add__
>  def __iadd__(self, other):
>    if isinstance(other, UserInt):
>      self.data += other.data
>    else:
>      self.data += other
>  # Repeat for all other operations
>
>UserInt as written is more properly UserNumber, since it supports longs ints
>and floats.

Just as a matter of style, would it be superior to add a __coerce__ method
that knows how to coerce ints, longs, and floats into UserInts, or make an
explicit check on each method?  __coerce__ saves a lot of typing, but I'm
generally wary of implicit things like coercion.  When I implemented a class
like this (a vector, actually), I used __coerce__, but after trying to wrap my
brain around the control flow, I was thinking maybe I wouldn't next time, even
though it seems to work fine.

Is this fear unfounded?  Is there a better way?


From tdelaney at avaya.com  Tue Dec 19 18:22:51 2000
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 20 Dec 2000 10:22:51 +1100
Subject: while true: !!!
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B0FBCD6@natasha.auslabs.avaya.com>

> > > >The idiom is actually
> > > >
> > > >while 1:
> > > >
> > > > do something
> > > >
> > > > if condition:
> > > > break
> > > >
> > > > do something else
>     [snip]
> > I go well out of my way to avoid this idiom, as I feel you 
> should always
> > have the termination condition at one end of the loop, 
> preferably the top
> 
> I disagree intensely with this stance: I strongly feel that the
> loop-exit condition should be *in the most natural place for it*,
> which is often the top, but not all that rarely elsewhere.
> 
> Anyway, one doesn't have to go 'well' out of one way to hide
> this clean, general loop structure as if it was otherwise --
> one 'status-variable' will suffice, e.g:
> 
> looping = 1
> while looping:
>     do something
>     if condition:
>         looping = 0
>     else:
>         do something else
>
> I think this is *substantially* less readable than the
> clean, natural form:
> 
> while 1:
>     do something
>     if condition: break
>     do something else

I don't usually do this either, as it is actually *worse* unless there is a
very obvious "thing" which the loop should end on.

Instead, I attempt to create a situation where *all* the decision-making is
encapsulated in the condition. This may well mean that I define a class or
function which does something, supplying me with data and a loop exit
condition. Again, the canonical example is reading a file:

maybe_buffered_stream = MaybeBufferedFileInputStream(file_object)

while maybe_buffered_stream.has_unread_contents:
	line = maybe_buffered_stream.readline()

Tim Delaney
Avaya Australia



From robin at jessikat.fsnet.co.uk  Tue Dec  5 05:15:18 2000
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Tue, 5 Dec 2000 10:15:18 +0000
Subject: PyThreadState_Get error again :(
References: <42+P6LAmqiK6Ewl1@jessikat.fsnet.co.uk>
 <3A2AD10F.8040803@ActiveState.com> <uTVaDtAJ13K6EwUw@jessikat.fsnet.co.uk>
 <3A2C7488.4040005@ActiveState.com>
Message-ID: <SYblhKA2AML6EwLn@jessikat.fsnet.co.uk>

In article <3A2C7488.4040005 at ActiveState.com>, Mark Hammond
<MarkH at ActiveState.com> writes
>Robin Becker wrote:
>
>> Is there any way to get 135 without reinstalling 2.0 or perhaps there
>> has been a fix in that? Is it worth pursuing further?
>
>135 is available stand alone at 
>http://www.activestate.com/Products/ActivePython/win32all.html
>
>I would love to know if it is a problem there, as I can't repro on 98. 
>Obviously, this is not to suggest that 95 has a problem while 98 doesnt. 
> I no longer have 95 installed here.
>
>Mark.
...
well the 135 release certainly fixes that problem on the Win95 OSR2
machine.
-- 
Robin Becker


From rrasche at dspace.de  Thu Dec  7 10:06:28 2000
From: rrasche at dspace.de (Rainer  Rasche)
Date: Thu, 7 Dec 2000 16:06:28 +0100
Subject: UML and Python
Message-ID: <90o91k$eks$1@desperado.vew-telnet.net>

Hi,

is somebody out there, who is interested in developing python applications
with UML-CASE Tools (e. g. ObjectDomain) and wants to exchange some
experiences with me?

Thanks, Rainer

rrasche at dspace.de




From gherman at darwin.in-berlin.de  Mon Dec 18 06:48:58 2000
From: gherman at darwin.in-berlin.de (Dinu C. Gherman)
Date: Mon, 18 Dec 2000 12:48:58 +0100
Subject: Singleton vs Proxies DP (was Re: Solution: Direct access to 
 Printer   I/O lines)
References: <917k9b$9i9$1@nnrp1.deja.com> <slrn93hrjn.2vp.ssthapa@localhost.localdomain> <91cnp3$g59$1@nnrp1.deja.com> <91d23f0jk7@news1.newsguy.com> <91gfnp$7r4$1@animus.fel.iae.nl> <91gkh801e49@news2.newsguy.com> <3A3CB83D.45C8775E@darwin.in-berlin.de> <91iu95013ta@news2.newsguy.com> <3A3D1D49.1A453B51@darwin.in-berlin.de> <91kmfd020hp@news1.newsguy.com>
Message-ID: <3A3DF9AA.74552346@darwin.in-berlin.de>

Alex Martelli wrote:
> 
> [gives well-reasoned explanations]

Following your reasoning I do see there is a place for such
solutions as you describe them. Luckily, patterns can (and,
in fact, should) be tweaked and adapted a lot to the current
needs (and "forces") at hand!

Actually, I'm pretty sure there is already something like a
singleton-like pattern with more of the properties you like.
Maybe you should check Linda Rising's "Pattern Almanac" or
similar sources?

  http://www.amazon.com/exec/obidos/ASIN/0201615673

Best regards,

Dinu

-- 
Dinu C. Gherman
................................................................
"The only possible values [for quality] are 'excellent' and 'in-
sanely excellent', depending on whether lives are at stake or 
not. Otherwise you don't enjoy your work, you don't work well, 
and the project goes down the drain." 
                    (Kent Beck, "Extreme Programming Explained")


From greg at cosc.canterbury.ac.nz  Wed Dec 20 19:49:34 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Thu, 21 Dec 2000 13:49:34 +1300
Subject: while true: !!!
References: <3A4024E5.98843CFC@cosc.canterbury.ac.nz> <91qadb$1di$1@plutonium.compulink.co.uk>
Message-ID: <3A41539E.DF6977D6@cosc.canterbury.ac.nz>

gbreed at cix.compulink.co.uk wrote:
> 
> Take any type, and if
> there's a way of it being nothing, it evaluates false.

Hmmm, let's try that out. Is [[]] true or false? 
It's a list containing a list containing nothing. 
Since [] is false, a list containing nothing must be
nothing. So [[]] is a list containing nothing,
which is nothing. Ah, it must be false! :-)

Nothing-is-weird-stuff,

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


From sholden at holdenweb.com  Tue Dec 19 08:29:46 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 19 Dec 2000 08:29:46 -0500
Subject: while true: !!!
References: <mailman.977180341.10206.python-list@python.org> <slrn93t7bv.le.grey@teleute.rpglink.com> <KTz%5.48644$x6.22165257@news2.rdc2.tx.home.com> <91nc7h0s0p@news2.newsguy.com>
Message-ID: <SqJ%5.3148$Yv6.82262@e420r-atl1.usenetserver.com>

Alex Martelli <aleaxit at yahoo.com> wrote in message
news:91nc7h0s0p at news2.newsguy.com...
> "Rainer Deyke" <root at rainerdeyke.com> wrote in message
> news:KTz%5.48644$x6.22165257 at news2.rdc2.tx.home.com...
> > "Steve Lamb" <grey at despair.rpglink.com> wrote in message
> > news:slrn93t7bv.le.grey at teleute.rpglink.com...
> > > On Mon, 18 Dec 2000 23:32:23 +0100 (MET), Carsten Geckeler
> > <nospam at no.spam>
> > > wrote:
> > > >`while true' would be more readable and clear than `while 1'.  By the
> > way,
> > > >explain a newbie that -1 is true. ;)
> > >
> > >     0 is true, all else is false.  And?  :P
> >
> > 0 is false, None is false, [] is false, () is false, {} is false, and x
is
> > false where x is a class instance and x.__nonzero__() returns false.
> > Everything else is true.  I think.
>
> Another couple of 'false's: a class instance which doesn't define
> __nonzero__ but does define __length__ when the latter returns 0;
> and, C-implemented user objects, similarly to class instances.
>
> I don't _think_ there are any more...
>
>
> Alex
>
Nothing is true.  All is permitted.  Hassan-i-Sabbah

regards
 Steve







From aahz at panix.com  Sun Dec 31 11:05:12 2000
From: aahz at panix.com (Aahz Maruch)
Date: 31 Dec 2000 08:05:12 -0800
Subject: 2001 Enchancement Wishlist
References: <mailman.978265218.29152.python-list@python.org>
Message-ID: <92nlfo$3j5$1@panix3.panix.com>

In article <mailman.978265218.29152.python-list at python.org>,
Tim Peters <tim.one at home.com> wrote:
>
>I suggest you're off track in your modeling, confounding roles that should
>be distinct:  that of Plus as a function and Plus as an object (representing
>an only partially evaluated sum).  In this case I'd probably have a base
>class Expression for all kinds of expresssions (incl. atoms like Symbols),
>make Plus a function mapping zero or more Expressions (or subclasses) to an
>Expression (or subclass), and make Sum a subclass of Expression that
>represents a partially evaluated sum.  That's the way I *think* of the
>domain, so that's the way I'd want to code it (and, no, I'm not bothered by
>having to come up with distinct names for the verb sense (Plus) and the noun
>sense (Sum) -- the distinction between nouns and verbs is usually helpful!).

Actually, to nitpick a bit, I'd suggest "Add" as the verb.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

Ninth Virtual Anniversary: Tomorrow!


From un4e at rzstud1.rz.uni-karlsruhe.de  Mon Dec 11 08:52:05 2000
From: un4e at rzstud1.rz.uni-karlsruhe.de (Bjoern Giesler)
Date: 11 Dec 2000 13:52:05 GMT
Subject: Modifying running code? [HACK ALERT]
Message-ID: <912m65$pu1$1@news.rz.uni-karlsruhe.de>

Hi,

I'm writing a kind of mini runtime environment that lets me debug / correct
buggy Python code "on the fly". This is supposed to work as follows:

1) an exception gets thrown by the buggy code and is caught by a toplevel
exception handler

2) the toplevel handler figures out the function that caused the exception

3) the function source is loaded into an editor and can be corrected by the
user

4) if the code has been saved, it is recompiled and inserted instead of the
original function

5) the program is then resumed at the point where the original function had
been called

Can this be done in Python at all? Can it be done in regular Python or only
in Stackless Python (which supports continuations and would therefore, I
believe, make step 5 easy)? Is the whole thing complete bull doodoo? Any
hints or suggestions?

Thanks in advance for any input,
		--Bj?rn


From nospam at caltech.edu  Thu Dec 14 04:08:04 2000
From: nospam at caltech.edu (Nathan Gray)
Date: Thu, 14 Dec 2000 01:08:04 -0800
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu> <slrn92tu7t.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com> <tcmay-8BCFAB.23025408122000@news.got.net> <90uddk$31f@gap.cco.caltech.edu> <tcmay-4AACBE.16294509122000@news.got.net>
Message-ID: <3A388DF4.836440B4@caltech.edu>

Tim May wrote:

> In article <90uddk$31f at gap.cco.caltech.edu>, Nathaniel Gray
> <nospam at caltech.edu> wrote:
>
> >  Just out of curiosity, why did you bother to take the time to do a
> >  google
> > search just to castigate the original poster in that thread?
>
> Notice how you're the one fanning the flames in these threads?

Notice how you're avoiding the question?  It's a simple question.  Why go to
the trouble of doing a search, which you claim is such an imposition on your
time, if you're not planning to help out?

On second thought, don't answer.  The answer's pretty obvious, and this thread
has outlived its expiration date.

> > He didn't ask you to do a search.  He asked if anybody had solved the
> > problem and would be willing to share their code.  He was quite polite,
> > and
> > even explicitly stated that he'd already searched dejanews and
> > python.org.
> > What exactly did you take issue with?  Did he use the wrong search engine
> > for your tastes?
>
> He has expressed happiness that he could find 1200 references, many of
> them applicable, to Python and simulated annelaing by using a better
> search engine.

After apologizing profusely under the misguided impression that his question
was somehow inappropriate for this newsgroup.  For the record, Jochen, your
question was just the sort of thing newsgroups were invented for.

> I didn't see _you_ providing any help, either in that thread or in this
> "Python origins" thread. Your role is apparently just to join in later
> and argue for niceness.

You're right.  I definitely should have given Jochen my python code for
simulated annealing.  Oh wait, I don't have any --  that's why I didn't reply
to his post.  See how that works?  If he'd asked for Matlab code simulating
networks of integrate-and-fire neurons, on the other hand, I would have
replied.  I would have told him to take it to the Matlab newsgroup, but I also
would have shared my code. <wink>

> You remind me of Alan Dershowitz.

So does yer mom!  Shall we continue with "I know you are, but what am I," or
shall we go straight to obscenities?  How about we skip it?

Over and out.

-n8gray
-at
-caltech
-dot
-edu



From dsh8290 at rit.edu  Thu Dec 28 08:00:04 2000
From: dsh8290 at rit.edu (D-Man)
Date: Thu, 28 Dec 2000 08:00:04 -0500
Subject: Is this a BUG?
In-Reply-To: <3A4B0C5E.1E7976D2@wielun.dhs.org>; from qdlaty@wielun.dhs.org on Thu, Dec 28, 2000 at 10:48:14AM +0100
References: <3A4B0C5E.1E7976D2@wielun.dhs.org>
Message-ID: <20001228080004.A13937@westsidecnc.rh.rit.edu>

On Thu, Dec 28, 2000 at 10:48:14AM +0100, QdlatY wrote:
> print "!",form["index"].value,"!"
> ! 1 !
> ----------------
>  ^ ^
> Why the hell there are spaces? index property has value '1'!
> not ' 1 '?!

This is a result of print trying to be convenient for you.  If I write

print "abc" , "def"

I will get

abc def

as output.

If I write

print "abc" + "def"

I will get

abcdef


If you use string concatenation when you print you will only get what
you have (no spaces added) plus an EOL at the end.

> 
> Yours
> QdlatY


-D



From michael at stroeder.com  Fri Dec 29 10:09:56 2000
From: michael at stroeder.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Fri, 29 Dec 2000 16:09:56 +0100
Subject: Python vs PHP (& Java?)
References: <92f3kp01cqe@news1.newsguy.com> <slrn94l86m.a3h.tim@degree.ath.cx> <mailman.978021012.21874.python-list@python.org> <3A4C8871.DB734376@stroeder.com> <C56A958929D650C9.FC4D26C518326FAC.CD8FA9082F345503@lp.airnews.net>
Message-ID: <3A4CA944.B02FB62B@stroeder.com>

Cameron Laird wrote:
> 
> In article <3A4C8871.DB734376 at stroeder.com>,
> Michael =?iso-8859-1?Q?Str=F6der?=  <michael at stroeder.com> wrote:
> >Moshe Zadka wrote:
> >>
> >> Apache modules. Yes, now that we have mod_snake, you
> >> can do these things in Python, in theory: but mod_snake is still
> >> young, while mod_perl is quite mature.
> >
> >mod_python?
> 
> Did you intend to supply readers with directions
> to <URL: http://www.modpython.org/ >?

Mainly I wanted to point out that mod_snake is not the only solution
=> IMHO the mod_perl argument is not 100% proven to be right.

Thanks for adding the URL anyway. ;-)

Ciao, Michael.


From andreas at andreas-jung.com  Wed Dec 20 14:02:34 2000
From: andreas at andreas-jung.com (Andreas Jung)
Date: Wed, 20 Dec 2000 20:02:34 +0100
Subject: Problems compiling on Sunos5
In-Reply-To: <m366kfhu94.fsf@boss.dreamsoft.com>; from rvprasad@cis.ksu.edu on Wed, Dec 20, 2000 at 11:34:15AM -0600
References: <m366kfhu94.fsf@boss.dreamsoft.com>
Message-ID: <20001220200234.A5081@yetix.sz-sb.de>

On Wed, Dec 20, 2000 at 11:34:15AM -0600, Venkatesh Prasad Ranganath wrote:
> Hi,
> 
> I am trying to compile Python 2.0 on Sunos 5.  I am able to compile but many
> functions are not being supported.  When I execute configure, the list of
> functions checked by configure are reported as not supported but all those
> functions are available.  For example, strftime() is reported as not available.
> 
> Is this a problem with the configure script or something I have to specify when
> configuring.  My configure command was as follows
> 
> configure --with-threads=/usr/lib --with-cycle-gc --without-libdb

Usually you should not care about the functions available or not. When
"make" works correctly  and you have a running Python interpreter
after the termination of "make" ...be happy...and "make install".
I have never seen any real problems with the installation of Python
under Solaris.

Andreas
 



From amk at mira.erols.com  Wed Dec 13 20:03:14 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 14 Dec 2000 01:03:14 GMT
Subject: Python-like compiled language
References: <913jr3$dmv$1@sylvester.vcn.bc.ca> <3A35835B.AAA92005@bioeng.ucsd.edu> <kuzoi178kg.fsf@lasipalatsi.fi> <9163mr02v2t@news1.newsguy.com> <kulmtl6xqx.fsf@lasipalatsi.fi> <917ipt023tk@news2.newsguy.com>
Message-ID: <slrn93g7rc.o0.amk@207-172-112-249.s249.tnt4.ann.va.dialup.rcn.com>

On Wed, 13 Dec 2000 11:24:43 +0100, Alex Martelli <aleaxit at yahoo.com> wrote:
>What about Modula-3, by the way?  I gather that one has
>similarities, too (having provided some of our BDFL's
>inspiration), but don't know much about its compilers'
>effectiveness.

The Modula-3 compiler is quite elderly, being based on GCC 2.4.5 or
something like that.  There's an updated version, m3cgc1, that patches
against more recent versions, but it doesn't seem to be very
well-known.  

M3 appears to be withering away.  Richard Watts, who maintained a
now-dead branch of the Modula-3 compiler, wants to design Modula-4:
www.modula-4.org, and current discussions in comp.lang.modula3.

--amk



From dsavitsk at my-deja.com  Tue Dec 19 18:09:08 2000
From: dsavitsk at my-deja.com (dsavitsk at my-deja.com)
Date: Tue, 19 Dec 2000 23:09:08 GMT
Subject: python with VB/Delphi
Message-ID: <91opqf$nr7$1@nnrp1.deja.com>

the link form python.org to the tutorial on using vb/delphi interfaces
to python programs seems to be no longer.  does anyone know of a good
tutorial on doing this.

do people have a preference for one of these (or neither)?

thanks
doug


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


From mjackson at wc.eso.mc.xerox.com  Tue Dec  5 20:36:15 2000
From: mjackson at wc.eso.mc.xerox.com (Mark Jackson)
Date: 6 Dec 2000 01:36:15 GMT
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com>
Message-ID: <90k56f$dhc$1@news.wrc.xerox.com>

Nigel <nigeljennison at cs.com> writes:
>  I am doing some research and I would be grateful if anybody could give
> me some history on the origins of Python.

But then you wouldn't really be doing any research, would you?

Hint:  there's *lots* of material to be found on, or linked from,
www.python.org.

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	I know nothing about this subject but I do have
	prejudices, which I am more than happy to share with you.
					- Leon Botstein




From DanielK at aracnet.com  Sun Dec 24 13:29:54 2000
From: DanielK at aracnet.com (Daniel Klein)
Date: Sun, 24 Dec 2000 10:29:54 -0800
Subject: soundex (revisited)
Message-ID: <Var16.263$LU6.109277@typhoon.aracnet.com>

After seeing the post from several days ago on soundex, I gave it whirl to
see if I could come up with something different (and possibly better),
following the rules laid down by Knuth:

def get_soundex(name, digits = 3):
    soundexcodes = "01230120022455012623010202"
    #               ABCDEFGHIJKLMNOPQRSTUVWXYZ
    instring = name.upper()
    soundex = instring[0]
    last = soundex
    instring = instring[1:]
    for char in instring:
        if 65 <= ord(char) <= 90:
            sx = soundexcodes[ord(char) - 65]
            if int(sx) and char != last:
                soundex += sx
                last = char
    if len(soundex) < (digits + 1): soundex = (soundex + ("0" * digits))
    return soundex[:digits + 1]


As this is one of the first (complete) functions I have created in Python,
all comments, suggestions, derogatory remarks :^) are welcome.

One question: how is it that this understands the 'upper()' function without
explicitly importing the 'string' module?

Dan:




From aleaxit at yahoo.com  Fri Dec 22 03:20:31 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 22 Dec 2000 09:20:31 +0100
Subject: COM / Variants / VT_BOOL
References: <mailman.977467384.30398.python-list@python.org>
Message-ID: <91v2sv01rpp@news2.newsguy.com>

"Stefan Migowsky" <smigowsky at dspace.de> wrote in message
news:mailman.977467384.30398.python-list at python.org...
> You do not need to create a VARIANT ! The win32com library automatically
> converts almost all Python objects to the corresponding VARIANT type.
> Since VT_BOOL is just an int which is transmitted through the COM channel
> you can use the simple Python 1 or 0, e.g.
> server = win23com.client.Dynamic("XXX")
>
> server.Enabled = 1 # Set to true
> server.Enabled = 0 # Set to false

*Almost* right... but not *quite*, and we hit this specific problem
quite recently (we weren't driving with Python, but I think that
would have made no difference).

VARIANT_BOOL (the content of VARIANT's discriminated union when
its type-tag is VT_BOOL) is defined as a 16-bit value that is
either 0 ('false') or 0xFFFF ('true'); other values are, it seems,
"not permitted".  Any 'sensible' server would 'of course' take
any non-zero here as 'true'... but we just found out you cannot
*COUNT* on that.

Setting the property .ValidateOnParse of the Xerces XML parser
(originally from IBM, now, I believe, from Apache), in its COM
setting, to 0 *or* 1, still keeps the parser non-validating; you
HAVE to set it to -1 or 65535 to make the parser a validating
one.  I have not checked the behavior of Microsoft's MSXML,
which the packaging of Xerces as a COM server is apparently
intended to generally mimic; but, anyway, I suspect the Xerces
team will not acknowledge this as a 'bug', whatever MSXML does
in the (non-documented, non-supported) case that one sets this
'boolean' property to a non-permitted value -- and, formally,
they would, I guess, be right.  I have not checked Xerces'
sources to confirm this, but I guess that somewhere they're
testing the property's value for "== VARIANT_TRUE".

*Most* COM servers will happily accept 'server.BoolProp = 1'
as setting the property to 'true', but, I suggest, don't count
on that; use 'server.BoolProp = -1' and you might be safer.


Alex





From skip at mojam.com  Fri Dec 22 11:57:50 2000
From: skip at mojam.com (Skip Montanaro)
Date: Fri, 22 Dec 2000 10:57:50 -0600 (CST)
Subject: ANNOUNCE: soundex.py
In-Reply-To: <3A414897.D82C953C@gte.net>
References: <mailman.977338109.32467.python-list@python.org>
	<3A414897.D82C953C@gte.net>
Message-ID: <14915.34830.638362.635454@beluga.mojam.com>

>>>>> "Steve" == Steve Williams <sandj.williams at gte.net> writes:

    Steve> I've found you can good results with a variable length soundex
    Steve> string--the more information you give the routine (first names,
    Steve> middle names, prefixes and suffixes) the better/smaller the
    Steve> result set.

Your wish is my command.  The module at

    http://www.musi-cal.com/~skip/python/soundex.py

now computes a variable-length soundex string if you set soundex.NDIGITS to
0:

    >>> soundex.NDIGITS=0
    >>> soundex.get_soundex("montanaro")
    'M5356'
    >>> soundex.get_soundex("merriweather clark")
    'M6362462'
    >>> soundex.get_soundex('van der tamp, albert')
    'V5363514163'
    >>> soundex.get_soundex('van der tamp, albert c, lieutenant colonel, Phd')
    'V5363514163243553245413'

Skip



From fredrik at effbot.org  Fri Dec  8 10:50:49 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 08 Dec 2000 15:50:49 GMT
Subject: reading directory structure
References: <90qtar$eve$1@narses.hrz.tu-chemnitz.de>
Message-ID: <tr7Y5.2595$O%.399909@newsc.telia.net>

Alexander Groten wrote:
> I'm new in Python.
> For a script I need a algorithm, that read a directory to a defined depth.
> Similar FIND.
>
> I have no plan (in the moment). Any suggestions?

os.path.walk

for some info on how to prevent os.path.walk from
traversing the whole disk, see:

http://www.deja.com/=dnc/getdoc.xp?AN=695333869

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From colaveaj at sirinet.net  Fri Dec  8 05:00:27 2000
From: colaveaj at sirinet.net (Tony)
Date: Fri, 8 Dec 2000 10:00:27 -0000
Subject: window placement w/ Tkinter
Message-ID: <t30nffdgb43t80@corp.supernews.com>

I am creating GUI apps and cant figure out how to set
the placement on the screen of my frames.  I dont see anything
in the constructor for it.  How is this done?






From kragen at dnaco.net  Sun Dec 31 08:03:25 2000
From: kragen at dnaco.net (Kragen Sitaker)
Date: Sun, 31 Dec 2000 13:03:25 GMT
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <92m9pm$d19$1@nnrp1.deja.com> <B3z36.122$9s1.18804@e420r-atl2.usenetserver.com> <3A4EC5B1.8912E571@engcorp.com>
Message-ID: <x8G36.156$9s1.31209@e420r-atl2.usenetserver.com>

In article <3A4EC5B1.8912E571 at engcorp.com>,
Peter Hansen  <peter at engcorp.com> wrote:
>Kragen Sitaker wrote:
>> 
>> The only time I wish I wasn't subject to Python's indentation rules is
>> when I can't outdent my "if finished(): break" statements like I do in
>> C.
>
>But why would you want to outdent it?  It's *part* of the loop, not a
>following statement.  Identation should match the block structure of the
>code, and that statement lies _within_ the block.

I would want to outdent it because, in my view, it isn't part of the
loop body; it's part of the loop structure itself, just like the while
1: at the beginning.  It is not within the block of code controlled by
the loop.  In fact, it comes *between* two blocks of code controlled by
the loop, much as else: comes between two blocks of code controlled by
an if (or, in Python, a for or while).

Indenting it with the rest of the code makes it appear as if there is
only one block of code, when in fact there are two basic blocks that
will run different numbers of times.
-- 
<kragen at pobox.com>       Kragen Sitaker     <http://www.pobox.com/~kragen/>
Perilous to all of us are the devices of an art deeper than we possess
ourselves.
       -- Gandalf the White [J.R.R. Tolkien, "The Two Towers", Bk 3, Ch. XI]




From eff at my-deja.com  Wed Dec 20 06:00:45 2000
From: eff at my-deja.com (eff at my-deja.com)
Date: Wed, 20 Dec 2000 11:00:45 GMT
Subject: while true: !!!
References: <mailman.977268243.15476.python-list@python.org> <NnS%5.482$Qb7.42792@newsb.telia.net> <4bY%5.172261$U46.5444054@news1.sttls1.wa.home.com>
Message-ID: <91q3gr$nbr$1@nnrp1.deja.com>

Greg Jorgensen wrote:
> This thread is a great example of what newsreader filters
> are good for.

so?  this newsgroup is cross-linked with a mailing list.
how about showing the people on that list some respect?

> Obviously quite a few of us are getting something out of it.

lots of new names in the killfile, you mean?

</F>


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


From boud at rempt.xs4all.nl  Fri Dec 29 14:46:45 2000
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 29 Dec 2000 19:46:45 GMT
Subject: Kanji
References: <92dl3u$to0$1@news.idiom.com> <p6qlmt1e3yh.fsf@informatik.hu-berlin.de> <92f3n7$hsf$1@news1.xs4all.nl> <slrn94op28.srq.qrczak@qrnik.zagroda>
Message-ID: <92ipn5$cf4$1@news1.xs4all.nl>

Marcin 'Qrczak' Kowalczyk <qrczak at knm.org.pl> wrote:
> 28 Dec 2000 10:12:55 GMT, Boudewijn Rempt <boud at rempt.xs4all.nl> pisze:

>> And Kura is a language database written entirely in Python and Qt
>> which reached feature-complete stage yesterday...

> "Kura" means "hen" in Polish BTW.

It's Nepali for language - and it's a word in Japanese and Maori, too.
I sometimes get hits from people called Kura who're looking for their
web presence...

By the way, version 0.99 is now on my website, and includes a Pinyin
input method for Chinese. I've used to enter a bit of Mencius:

http://www.valdyas.org:8000/KURA?command=record?table=lng_text?textnr=17

(Make sure you're using the utf-8 encoding; some browsers don't
pick it up from the html.)

-- 

Boudewijn Rempt  | http://www.valdyas.org


From sill at localhost.kitenet.net  Fri Dec 15 16:59:56 2000
From: sill at localhost.kitenet.net (Oldayz)
Date: Fri, 15 Dec 2000 21:59:56 GMT
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <915jcf0c6o@news2.newsguy.com> <9162p3$7t2$1@animus.fel.iae.nl> <MBWZ5.334$h67.37748@newsread1.prod.itd.earthlink.net> <3A397C74.6C8906E1@cosc.canterbury.ac.nz> <91d5520m6j@news1.newsguy.com>
Message-ID: <wvw_5.10652$h67.670217@newsread1.prod.itd.earthlink.net>

On Fri, 15 Dec 2000 14:08:15 +0100, Alex Martelli <aleaxit at yahoo.com> wrote:
>"Greg Ewing" <greg at cosc.canterbury.ac.nz> wrote in message
>news:3A397C74.6C8906E1 at cosc.canterbury.ac.nz...
>> Oldayz wrote:
>> >
>> > I thought something like forever: would be even more lucid
>>
>> If there is a break in the loop somewhere, it's
>> not really forever, though, is it?
>>
>> So perhaps it should *really* be
>>
>>    for a while:
>>       ...
>>
>> :-)

Sarcasm notwithstanding, I think forever: makes
sense because the idiom suggests the code will
execute forever, unless something (say a break
statement) stops it. The question is whether
while 1: is common enough to deserve a separate
name for it - to be honest, i'm almost certain
it does not, even though it's a matter of taste.

Hey, how about if you think while 1: looks ugly
you do true = 1 somewhere on top and use
while true: ? I think it's much more obvious
and clean and you don't need to change syntax.

Ideally, language should stay fairly close to
the way we normally think; in this case it's
the difference between thinking 'walk north
until the heat death of the universe BUT if
you see grocery, get in, buy apple juice and
return home' and 'go to grocery, buy apple
juice and return home'. I usually think in
a latter manner but YMMV ;-).

OTOH, i'm a newbie and I can be totally off
base here, am I?

-- 

	Andrei


From dsh8290 at rit.edu  Mon Dec 18 15:12:25 2000
From: dsh8290 at rit.edu (D-Man)
Date: Mon, 18 Dec 2000 15:12:25 -0500
Subject: while true: !!!
In-Reply-To: <LNBBLJKPBEHFEDALKOLCAEGKIEAA.tim.one@home.com>; from tim.one@home.com on Mon, Dec 18, 2000 at 01:15:39PM -0500
References: <20001218085229.A10917@westsidecnc.rh.rit.edu> <LNBBLJKPBEHFEDALKOLCAEGKIEAA.tim.one@home.com>
Message-ID: <20001218151225.A12029@westsidecnc.rh.rit.edu>

On Mon, Dec 18, 2000 at 01:15:39PM -0500, Tim Peters wrote:
> [D-Man]
> > I think that some people were confused with this statement.  He meant
> > that 'true' and 'false' would be constants like in C++ and Java.  It
> > would be illegal to try
> >
> > true = false
> > ...
> 
> A non-empty string literal is true, and unalterably so, so everyone can be
> happy writing:
> 
> while "forever, because stupid Guido won't define 'true'":
>     wink

Yes, true should be defined as a global constant, but this is almost
as good :

while 'true' :

:-)


Good point Tim.



From qrczak at knm.org.pl  Sun Dec 31 05:31:33 2000
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 31 Dec 2000 10:31:33 GMT
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <t4t9qv73vvvfff@corp.supernews.com>
Message-ID: <slrn94u2o4.fkv.qrczak@qrnik.zagroda>

Sat, 30 Dec 2000 22:21:12 -0500, Andrew Henshaw <andrew_dot_henshaw_at_earthling_dot_net> pisze:

> a)
> do:
>     R=Themistor.ReadResistance()
>     T=R_to_Temperature(R)
> until (T>=25)

> If a), then this looks very unPythonic, because the indentation
> blocking is backwards.  In other words "until (T>=25)" is a control
> statement for an indented block that is above it (and it doesn't
> have a ":").

do:
    R=Themistor.ReadResistance()
    T=R_to_Temperature(R)
:until (T>=25)

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


From kalle at gnupung.net  Fri Dec  8 11:19:44 2000
From: kalle at gnupung.net (Kalle Svensson)
Date: Fri, 8 Dec 2000 17:19:44 +0100
Subject: Refactoring browser
In-Reply-To: <3A31047A.A7E880EF@mail.vega.bg>
References: <3A31047A.A7E880EF@mail.vega.bg>
Message-ID: <Pine.GSO.4.21.0012081711410.29357-100000@luna.dsv.su.se>

On Fri, 8 Dec 2000, Stefan Boiadjiev wrote:

> Do in python world exist something similar to Smalltalk Refactoring
> browser ?

There is a project called Bicycle Repair Man.
http://bicyclerepair.sourceforge.net/

Regards,
  Kalle
-- 
Email: kalle at gnupung.net     | You can tune a filesystem, but you
Web: http://www.gnupung.net/ | can't tune a fish. -- man tunefs(8)
Not signed due to logistical difficulties (I'm not at home).
If you doubt the authenticity of this message, please tell me.



From tim at 2wave.net  Wed Dec 20 13:22:14 2000
From: tim at 2wave.net (tim at 2wave.net)
Date: Wed, 20 Dec 2000 18:22:14 GMT
Subject: DCOracle-1.3.2 makefile problem .er. opportunity?
Message-ID: <91qtcd$dic$1@nnrp1.deja.com>

Hi

I'm no makefile expert, I'm afraid, and am trying to build DCOracle-
1.3.2 in Linux 7 (for Oracle 8.1.6), and not succeeding :-(. I've
followed the makefile instructions provided, and get to the problem
below. I haven't found any references to others having this problem on
comp.lang.python, so I wonder if any of you who have been down this
path before recognise the error - and point me in directions which
might overcome it!

------------------------------
root> makefile -f Makefile.pre.in boot
rm -f *.o *~
rm -f *.a tags TAGS config.c Makefile.pre python sedscript
rm -f *.so *.sl so_locations
VERSION=`python -c "import sys; print sys.version[:3]"`;
installdir=`python -c "import sys; print sys.prefix"`;
exec_installdir=`python -c "import sys; print sys.exec_prefix"`;
make -f ./Makefile.pre.in VPATH=. srcdir=.
     VERSION=$VERSION
     installdir=$installdir
     exec_installdir=$exec_installdir
     Makefile
make[1]: Entering directory `/home/tim/rpms/DCOracle-1.3.2/src'
make[1]: *** No rule to make target'/usr/lib/python1.5/config/Makefile'
needed by 'sedscript'. Stop.
make[1]: Leaving directory `/home/tim/rpms/DCOracle-1.3.2/src'

------------------------------------------------

Thanks

Tim


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


From root at rainerdeyke.com  Fri Dec 15 11:38:47 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Fri, 15 Dec 2000 16:38:47 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <r0pzoi11lhz.fsf@black.cs.washington.edu> <5jvZ5.3588$Kd1.418770@newsb.telia.net> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com> <kuu28971kp.fsf@lasipalatsi.fi> <3A380E4E.8537A0FB@san.rr.com> <91a1vr02b02@news2.newsguy.com> <Co9_5.25465$x6.13756579@news2.rdc2.tx.home.com> <91bbq102mtk@news1.newsguy.com> <rWd_5.26820$x6.14119694@news2.rdc2.tx.home.com> <91d5el0me7@news1.newsguy.com>
Message-ID: <rOr_5.31120$x6.15241856@news2.rdc2.tx.home.com>

"Alex Martelli" <aleaxit at yahoo.com> wrote in message
news:91d5el0me7 at news1.newsguy.com...
> "Rainer Deyke" <root at rainerdeyke.com> wrote in message
> news:rWd_5.26820$x6.14119694 at news2.rdc2.tx.home.com...
>     [snip]
> > > > class MutableInt(UserInt.UserInt):
> > > >   def set(self, value):
> > > >     self.data = value
> > >
> > > UserInt is not in the Python 2 docs, though it's nice to hear it
> > exists:-).
> >
> > Actually it doesn't, but it's trivial (if tedious) to implement.
>
> But, if UserInt is not a funky metaclass, how will MutableInt live
> up to its name?  Surely it should at least redefine all of the
> in-place methods...:
>
>     x = y = MutableInt(23)
>     x += 1

Yes, in-place modification should be supported.  I had something like this
in mind:

class UserInt:
  def __init__(self, value):
    self.data = value
  def __add__(self, other):
    if isinstance(other, UserInt):
      return self.__class__(self.data + other.data)
    else:
      return self.__class__(self.data + other)
  __radd__ = __add__
  def __iadd__(self, other):
    if isinstance(other, UserInt):
      self.data += other.data
    else:
      self.data += other
  # Repeat for all other operations

UserInt as written is more properly UserNumber, since it supports longs ints
and floats.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From Alok_k at infy.com  Thu Dec 28 04:12:42 2000
From: Alok_k at infy.com (Alok_k)
Date: Thu, 28 Dec 2000 14:42:42 +0530
Subject: python faqs by email
Message-ID: <19CC5A519D17D2118E9C0008C7F4327502217AFD@TWRMSG02>

Hi list
I don't have access to the web but I do have email access.
Is it possible to get hold of python faqs thru email? 
some that will be helpful to a c++ programmer in learning python.

Regards,
Alok



From tim.one at home.com  Wed Dec 27 23:32:49 2000
From: tim.one at home.com (Tim Peters)
Date: Wed, 27 Dec 2000 23:32:49 -0500
Subject: One Python 2.1 idea
In-Reply-To: <20001227104140.F25820@xs4all.nl>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEKKIFAA.tim.one@home.com>

[posted & mailed]

[Tim]
> >>> class Oops:
> ...

[Thomas Wouters]
> Funny, I didn't see that original message. I guess the list is
> still a little flabergasted over being moved and Barry apparently
> abandoning it (because of his lack of connectivity) ;)

Best I can tell, mail->news works great now, in that emails to
python-list at python.org show up on local news servers very quickly (minutes).
But for reasons not yet understood, the mailing-list version is experiencing
random long delays, and this is true of all the mailing lists hosted at
mail.python.org.  I haven't seen evidence that any msgs are getting lost,
but multi-hour delays are common, and I've seen a few delayed by a full day.

Barry is indeed whining about his lack of connectivity, as if you need
megabit speeds to look at a stinking mail header <wink>.

About x.foo = x.foo, I never saw anyone suggest that to save time, before
this thread.  Where it did come up several times over the years was people
trying to customize how a method works for a particular instance, rather
than make a proper subclass.  Since Python made instance.__class__ writable,
I don't see any excuse for *that* form of trickery anymore.

but-i-guess-whatever-isn't-forbidden-is-mandatory-ly y'rs  - tim




From aleaxit at yahoo.com  Sat Dec 23 03:17:49 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 23 Dec 2000 09:17:49 +0100
Subject: Good introduction to functional programming with Python?
References: <mailman.977553967.26987.python-list@python.org>
Message-ID: <921nd205si@news2.newsguy.com>

"Tim Peters" <tim.one at home.com> wrote in message
news:mailman.977553967.26987.python-list at python.org...
    [snip]
We agree on far too much, and dwelling on agreements ain't no
fun, so, let's pick on a difference:

> If it is, then if FP is "about" anything, it's about the art of using
> higher-order functions, and I believe mere mortals *need* a strong static
> typing system to keep those straight.  Maps returning maps returning maps
> ... can get hard to follow real fast.

Erlang designers and practitioners seem to disagree: and Erlang is arguably
the FP around which most $$$ turn in today's real-world, what with its use
in Ericsson and Bluetail's acquisition by a Nortel subsidiary.  I have no
real
experience with it, and thus no solid basis on which to judge it, but,
sordid
materialist that I am, I _am_ impressed when hundreds of megabucks are
in play:-).  And Erlang relies on strong dynamic (aka latent) typing, and an
emphasis on concurrency in an FP setting, to (allegedly) reach good levels
of reliability, performance, ease of learning, programmer productivity.

Static typing has its place, but so, it appears, does latent typing, even in
FP.


> > or Unixoid.  (Not that Haskell's and Erlang's implementations
> > are any less free, cross-platform, and good-quality, but, IMHO,
> > Rice's Scheme has characteristics that make it even better).
>
> A caution that this depends on how powerful his computer is:  DrScheme is
> too much a resource hog for a satisfying experience on older Windows boxes
> (voice of experience).  This is a bit ironic, since Scheme is ported at
> least as often as Python to tiny platforms.  You pay through the nose for
> "all the trimmings" that come with DrScheme.  Worth it if you can afford
it!

Thanks for the warning!  The oldest machine on which I had tried Rice
Scheme's had a Pentium-133 and 32M RAM, and they seemed to run fine
there -- with older boxes yet, a smaller implementation is no doubt better!


Alex





From rpm at wag.caltech.edu  Sun Dec 17 13:36:45 2000
From: rpm at wag.caltech.edu (Richard P. Muller)
Date: Sun, 17 Dec 2000 10:36:45 -0800
Subject: Error evaluating numbers starting in zero
Message-ID: <3A3D07BD.85EFEB3D@wag.caltech.edu>

I'm sure people already know about this, but I just got hit with a
really screwy error.

I have a python checkbook program that evaluates date strings, such as
"01/31/99". Most dates are fine, but if I have a date such as
"01/08/99", python runs into an error when it tries to eval("08").

I can see how this might be a feature and not a bug, in that it
interprets numbers starting in "O" as octyl.

My workaround is the function:

def chop_leading_zeros(str):
    # removes a single leading zero from a string to be evaluated
    # I think it interprets these numbers as Octyl
    if str[0] == '0':
        str = str[1:]
    return str

Does anyone have a simpler or more elegant workaround?


From aleaxit at yahoo.com  Tue Dec 19 12:11:16 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 19 Dec 2000 18:11:16 +0100
Subject: Events in win32com
References: <Y1g=Op6IfWSoJn1ISFtOyuvqwBb3@4ax.com>
Message-ID: <91o4sk01jdb@news2.newsguy.com>

"Paul Moore" <paul.moore at uk.origin-it.com> wrote in message
news:Y1g=Op6IfWSoJn1ISFtOyuvqwBb3 at 4ax.com...
> If I create an object using
>
> obj = win32com.client.Dispatch(progid)
>
> can I see events from the object?

You can, although using DispatchWithEvents may be easier.

This is a typical example...:

import win32com.client
from win32com.client import gencache
evmod = gencache.EnsureModule('{D240EC61-B6ED-11D4-9E45-0060B0EB1D67}', 0,
1, 0)

class Evhand:
    def OnOne(self):
        print "eh: One!"
    def OnTwo(self):
        print "eh: Two!"

class Other(evmod._IEvGenEvents2):
    def OnOne(self):
        print "oh: One!"
    def OnTwo(self):
        print "oh: Two!"

ob = win32com.client.DispatchWithEvents("evts.EvGen", Evhand)
oh = Other(ob)
ob.Fire()


where progid 'evts.EvGen' refers to a tiny object I built just
for the purpose of event-handling tests (it generates events
along different sourceinterfaces when you call its Fire method).
This shows we can get events separately from DispatchWithEvents
(and on a different sourceinterface than the default one, if we
know the name of that alternate sourceinterface!).

> In WSH the CreateObject method has
> an extra parameter which is the prefix for subroutine names which are
> called on an event, something like
>
> ie = WScript.CreateObject("InternetExplorer.Application", "ie_")
>
> Sub ie_OnQuit()
> End Sub
>
> While I don't like the "specially named subroutine" approach, I do
> need to be able to sink events in Python. Is this possible, or should
> I resort to using Python via WSH?

It's quite possible to avoid WSH, unless you like the services
it brings to your script.


> (BTW, assuming I can sink events, will Python handle ByRef event
> parameters - for example, a Cancel parameter which is set to True in
> the handler to stop the event occurring).

If your EnsureModule (or equivalent) has been called, so that
Python *knows* exactly what's happening, then, yes, [out] and
[in,out] parameters will be handled (COM has no 'byref', but
that's how VB maps [in,out] -- [out] is a big problem in VB),
but of course they'll be handled for events as in other COM
Python cases -- that is, they become extra _return values_
of your method.


Alex





From rapto at arrakis.es  Mon Dec 18 14:11:15 2000
From: rapto at arrakis.es (Marcos =?iso-8859-1?Q?S=E1nchez?= Provencio)
Date: Mon, 18 Dec 2000 20:11:15 +0100
Subject: More docs in =?iso-8859-1?Q?Spanish=2FM=E1s?= madera
References: <5.0.2.1.0.20001218105723.00a232e0@mail.codeit.com>
Message-ID: <3A3E6153.CA888DF5@arrakis.es>

Julio Flores Schwarzbeck wrote:
> 
> I believe you meant http://pySpanishDoc.sourceforge.net/
> 
> Probablemente te refieres a: http://pySpanishDoc.sourceforge.net/
> 
> Saludos,
> 
> At 06:01 PM 12/18/00 +0000, rapto at arrakis.es wrote:
> >New files in
> >Ficheros nuevos en
> >http://pySpanishDoc/sourceforge.net/
> >
> >
> >Sent via Deja.com
> >http://www.deja.com/
> >--
> >http://www.python.org/mailman/listinfo/python-list

Indeed. Thank you very much.
Corcho, vaya gamba. M?s vale vigilarme.



From pearu at cens.ioc.ee  Sat Dec 30 03:52:10 2000
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Sat, 30 Dec 2000 10:52:10 +0200
Subject: 2001 Enchancement Wishlist
In-Reply-To: <3A4CEACB.220998A0@javanet.com>
References: <3A4CEACB.220998A0@javanet.com>
Message-ID: <Pine.LNX.4.10.10012301008410.5079-100000@kev.ioc.ee>


On Fri, 29 Dec 2000, Raymond Hettinger wrote:

> 3.  __init__ is the only special method that doesn't allow a return value.
>    I think it should continue to return the new object by default but also
>    allow another return item to be substituted.  For example:
> 
>             def __init__( self ):
>                    if MyClass.aSingleton == None:
>                          MyClass.aSingleton = self
>                    return MyClass.aSingleton

I see that many Python experts are against this proposal but anyway, I
would like to support this idea with the following examples.

Say, I would like to define

class Real:
    def __init__(self,num):
        self.num = num
    <other number class methods>

class Complex:
    def __init__(self,real,imag):
        if imag == 0:
            return Real(real)
        self.real,self.imag = real,imag
    <other number class methods> 

That is, if imaginary part of a Complex number is zero, it's constructor
would return Real instance:

>>> a = Complex(1,2)
>>> b = Complex(1,0)
>>> print a.__class__.__name__,b.__class__.__name__
Complex Real

Yes, I could cover this example in Python 2.0 by defining:

class Complex:
    def __init__(self,real,imag):
        if imag == 0:
            self.__class__ = Real
            self.real = real
        else:
            self.real,self.imag = real,imag
    <other number class methods>

but it works only if __init__ should "return" a class instance.

For example, the following behavior cannot be handled in Python 2.0:

class NonZeroInteger:
    def __init__(self,num):
        if num == 0:
            return 0
        self.num = num
    <other number class methods>

so that
>>> type(NonZeroInteger(2)),type(NonZeroInteger(0))
(<type 'instance'>, <type 'int'>)

Sure, one can write wrapper functions (as some of you already suggested):

class _Complex:
    ...
def Complex(real,imag):
    if imag == 0: return Real(real)
    return _Complex(real,imag)

but I think that this way is actually a very unpythonic one,
especially if there is much simpler/cleaner solution available (that is,
possible), that is, allowing __init__ return arbitrary object.

I have used these number classes only for demonstration, but there are 
real situations where 'rich' __init__ would be very useful. For example,
in PySymbolic there are classes Plus, Symbol, etc and gmpy numbers that
are objects. E.g. class Plus represents a sum of its elements (arguments):

c = Plus(a,b,c)       -> c = a + b + c  # c is Plus instance

where a,b,c are Symbol instances.
Now, if one constructs a Plus instance only with one argument, 

c = Plus(a)           -> c = a          # c is Symbol instance

***lots of code*** could be avoided with rich Plus.__init__ that would
return in this case Symbol instance a.
If the argument to Plus constructor is gmpy number:

c = Plus(mpz(3))      -> c = mpz(3)     # c is mpz type object

then rich Plus.__init__ would return mpz object.

So, in my humble opinion, rich __init__ would be very useful feature.

	Pearu



From erno-news at erno.iki.fi  Sun Dec 24 14:17:50 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 24 Dec 2000 21:17:50 +0200
Subject: PEP 222 draft
References: <slrn947bj1.1rs.amk@207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com>
Message-ID: <kug0jdtyqp.fsf@lasipalatsi.fi>

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

|     The cgi.py module would be deprecated.  (XXX A new module or
| package name hasn't been chosen yet: 'web'?  'cgilib'?)

i would like it more to add the new stuff to the existing cgi module
rather than having old/new modules with slightly different
variations of the same name.

|     XXX open issues: naming convention (studlycaps or
| underline-separated?); need to look at the cgi.parse*() functions
| and see if they can be simplified, too.

StudlyCaps for classes and underline_separated for the rest!

|     # The Response class borrows most of its methods from Zope's #
| HTTPResponse class.


    
|         def redirect(self, url): "Redirect the browser to another
| URL" pass
    
i think this is spurious, there is the status method already,
and you don't know which of the various 30x responses is most
appropriate here.

|         # XXX methods for specific classes of error:serverError, #
| badRequest, etc.?

i like cacheability, maybe expires/last-modified/vary
but maybe that's spurious too.
    
    
|     class Request:
    
|         """ Attributes:

|         XXX should these be dictionaries, or dictionary-like
| objects?  .headers : dictionary containing HTTP headers .cookies :
| dictionary of cookies .fields : data from the form .env :
| environment dictionary """

probably dictionary-like, since there can be several occurrences
of a single header and the order might matter.
        
[...]
    
|         # Should people just use the dictionaries directly?  def
| getHeader(self, name, default=None): pass

"getList" too?

    
|         # Possibilities?  I don't know if these are worth doing in
| the # basic objects.  def getBrowser(self): "Returns
| Mozilla/IE/Lynx/Opera/whatever"

do you mean attempting something more than just returning the
user-agent header? that seems highly unwise to me. many browsers
(opera, konqueror, msie atleast) lie in the user-agent header
and it's not possible to decipher it robustly and future-proofly.
and most of the situations where misguided cgi programmers want to do
user-agent sniffing is exactly differentiating between these.

   -- erno




From junaftnoon at nospamplzyahoo.com  Sat Dec 30 14:29:59 2000
From: junaftnoon at nospamplzyahoo.com (June Kim)
Date: Sun, 31 Dec 2000 04:29:59 +0900
Subject: Running Boa Contructor with the latest versions of Python & wxPython
Message-ID: <92ld65$4un$1@news.nuri.net>

I tried to run BC on my windows machine installed with the lastest versions
of Python(2.0) and wxPython(2.2.2) and got the following error.

Any roundabout ways to run it besides downgrading the Python and wxPython?

Traceback (most recent call last):
  File "boa.py", line 119, in ?
    main()
  File "boa.py", line 112, in main
    app = BoaApp(0)
  File "boa.py", line 49, in __init__
    wxApp.__init__(self, redirect)
  File "c:\python20\wxPython\wx.py", line 1646, in __init__
    _wxStart(self.OnInit)
  File "boa.py", line 58, in OnInit
    self.main = Palette.BoaFrame(None, -1, self)
  File "Palette.py", line 157, in __init__
    wxMenu(), self.componentSB, app)#palettePage.menu
  File "Editor.py", line 135, in __init__
    self.addShellPage()
  File "Editor.py", line 310, in addShellPage
    self.tabs.AddPage(ShellEditor.ShellEditor(self.tabs, -1), 'Shell')
  File "ShellEditor.py", line 26, in __init__
    PythonStyledTextCtrlMix.__init__(self, wId, -1)
  File "Views\StyledTextCtrls.py", line 137, in __init__
    self.SetKeywords(0,
AttributeError: 'ShellEditor' instance has no attribute 'SetKeywords'





From fredrik at effbot.org  Fri Dec 29 15:36:23 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 29 Dec 2000 20:36:23 GMT
Subject: Question re: eval()
References: <978109654.339714327@news.silcom.com>
Message-ID: <bB636.2482$Qb7.271736@newsb.telia.net>

Clarence Gardner wrote:
> I want to eval() an expression, but have some of the variables in it be
> lazily evaluated.  E.g.,
>
> class D:
>     def __getitem__(self, name):
>         if name == 'n':
>             return 3
>
> d = D()
> x = eval('n + 1', d)
>
> but the eval function accepts only a dictionary as the second parameter.
> Is there any amazing way to do this?

not really "lazy" evaluation, but the following trick
lets you figure out what names you need to define
before evaluating it:

def getvalue(name):
    # get value of variable "name"
    print "fetch", name
    return len(name)

myexpr = "a + bb + ccc"

# compile expression
code = compile(myexpr, "<string>", "eval")

# populate dictionary
dict = {}
for name in code.co_names:
    dict[name] = getvalue(name)

print "result", eval(code, dict)

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From aahz at panix.com  Fri Dec 22 01:31:00 2000
From: aahz at panix.com (Aahz Maruch)
Date: 21 Dec 2000 22:31:00 -0800
Subject: Thread time-out in python
References: <91m3vv$hfg$1@nnrp1.deja.com>
Message-ID: <91usf4$1db$1@panix3.panix.com>

In article <91m3vv$hfg$1 at nnrp1.deja.com>,  <kataru1 at my-deja.com> wrote:
>
>In my application, I spawn a thread and give it some work. Here, I need
>to be notified if that thread does not finish the assigned task in
>certain time limit. Could anyone suggest me what would be the best
>way to do it, if any?

You'd probably need a second thread to act as a timer.  The problem is
that you need a way to notify the first thread to stop working, and that
can only happen if the first thread is processing work in small enough
chunks that it is easily interruptable.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

I don't really mind a person having the last whine, but I do mind
someone else having the last self-righteous whine.  --Aahz


From dsh8290 at rit.edu  Sun Dec 17 23:57:56 2000
From: dsh8290 at rit.edu (D-Man)
Date: Sun, 17 Dec 2000 23:57:56 -0500
Subject: while true: !!!
In-Reply-To: <B43D149A9AB2D411971300B0D03D7E8B0FBCBB@natasha.auslabs.avaya.com>; from tdelaney@avaya.com on Mon, Dec 18, 2000 at 11:03:31AM +1100
References: <B43D149A9AB2D411971300B0D03D7E8B0FBCBB@natasha.auslabs.avaya.com>
Message-ID: <20001217235756.A9804@westsidecnc.rh.rit.edu>

On Mon, Dec 18, 2000 at 11:03:31AM +1100, Delaney, Timothy wrote:
> 
> while 1:
> 
> 	try to read a line from a file
> 
> 	if we didn't get a line:
> 		break
> 
> 	do something with the line
> 

This is precisely where having assigment as an expression (like C/C++)
is useful :

while ( (line = file.readline()) != None ) :
	do something with the line

Even a do-while loop doesn't solve this idiom ideally, you need to
read the line before the loop test and the test must occur before
using the line.


Maybe adding some sort of peek function to file objects could solve this :

while ( file.exists_next_line() ) :
	line = file.readline()
	do something with the line

-D




From aahz at panix.com  Thu Dec  7 20:39:30 2000
From: aahz at panix.com (Aahz Maruch)
Date: 7 Dec 2000 17:39:30 -0800
Subject: problems with urllib in Python20?
References: <90pdir$k7$1@bob.news.rcn.net>
Message-ID: <90pe4i$7fk$1@panix6.panix.com>

In article <90pdir$k7$1 at bob.news.rcn.net>,
Z 3 Penguin <z3penguin at penguinpowered.com> wrote:
>
>whenever i try to use urllib.urlopen() in Python20/Win32, i get this error:
>-START-
>Fatal Python error: Interpreter not initialized (version mismatch?)

By "Win32" do you mean Win 3.1 with Win32 extensions or what?  I've used
Python20 and urllib on Win98SE with no problems.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

What if there were no rhetorical questions?  --Aahz


From sholden at holdenweb.com  Tue Dec 19 08:22:31 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 19 Dec 2000 08:22:31 -0500
Subject: while true: !!!
References: <mailman.977127721.4398.python-list@python.org> <DnD%5.166633$U46.5342944@news1.sttls1.wa.home.com>
Message-ID: <ckJ%5.3097$Yv6.81688@e420r-atl1.usenetserver.com>

Greg Jorgensen <gregj at pobox.com> wrote in message
news:DnD%5.166633$U46.5342944 at news1.sttls1.wa.home.com...
> "Blair Lowe" <Blair.Lowe at compeng.net> wrote:
>
[snip]
>
> You can define true and false yourself, as many C/C++ header files do, but
> then you run the risk of stuff like this:
>
>     true = 1
>     false = 0
>     ...
>     a = len(s)
>     if a == true:  # fails unless a is 1
>
I've always tried to discourage students from comparing a Boolean with a
constant, but it can be difficult to get the point over that (a) can
evaluate as true even though (a == some_constant) may well evaluate to
false.

Clearly, if what you have is a Boolean value (i.e. one which can be used as
the expression in an "if" statement) then it's adding inefficiency to
compare it with something, when what should be used is either

    if a:

or

    if not a:

This works quite nicely in Python:

>>> not "forever"
0
>>> not ()
1
>>> not(1,2,3)
0

although I have worked in languages where, due to bad semantics, any
non-zero value was treated as true, and the "not" operator simply did a
bitwise inversion, so (not {something true}) also evaluated true.

Be that as it may, I would still argue it's bad pedagogy to suggest that
people write "if Boolean == true_constant", since this *should* (but
sometimes isn't, in poorly designed languages) be equivalent to "if
Boolean".

> It's much harder to track this kind of thing down than learning once in
your
> career what "while 1" means.
>
> > While "while 1" (pardon the pun) is a standard fixture that anyone
> > with any experience knows what is going on, the whole point is to
> > make all our code readable so that we, and others can maintain it and
> > reuse it easily.
>
> Exactly--that's why learning idioms is as important as learning syntax.
>
This is why c.l.p. is so helpful: the idioms are regularly presented.

> > Although the definition is cryptic, the resulting code is
> > very readable. That is what we want.
>
> Until it doesn't work. Encountering "while true:" in a program, a newbie
may
> assume that true is a Python keyword. I remember a C++ programmer I worked
> with who though that true and false were keywords until he discovered this
> (typical) definition in a header file:
>
>     enum { false = 0, true = -1 };
>
Although one could argue that he was using a local idiom.  But in general,
and especially in a learning environment, cryptic definitions aren't always
waht you want.  If someone then wants to look "under the hood" they have to
reach a new level of understanding.  "while 1" isn't really that difficult!

regards
 Steve






From bartig at my-deja.com  Mon Dec 11 14:15:14 2000
From: bartig at my-deja.com (bartig at my-deja.com)
Date: Mon, 11 Dec 2000 19:15:14 GMT
Subject: dictionaries thread safe
Message-ID: <91393q$opb$1@nnrp1.deja.com>

I'm writing a program that passes dictionaries
into a function that's being threaded.  Are
dictionaries thread safe?


Sent via Deja.com http://www.deja.com/
Before you buy.


From MarkH at ActiveState.com  Sat Dec  9 19:50:56 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sun, 10 Dec 2000 00:50:56 GMT
Subject: Implementing IPersistStorage using PythonCOM
References: <t308m58pkecra3@corp.supernews.com> <3A307D8C.10909@ActiveState.com> <t325urdmtq1t07@corp.supernews.com> <3A317651.5000109@ActiveState.com> <t339eb6pvurf67@corp.supernews.com>
Message-ID: <3A32D22B.3050901@ActiveState.com>

IPersistStorage is pretty badly broken when trying to implement it from 
Python. A really obvious error in PyGPersistStorage.cpp.

I have fixed it here, but not too sure how that helps you - if you are 
building from sources, then you can simply get a new CVS update and 
rebuild.  If you aren't I could try mailing you a new pythoncom20.dll

Mark.



From pbjorn at uswest.net  Sat Dec 23 12:32:25 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Sat, 23 Dec 2000 10:32:25 -0700
Subject: Some Python 2.1 ideas
References: <mailman.977590932.19335.python-list@python.org>
Message-ID: <3A44E1A9.D3A9FC35@uswest.net>

Bob Alexander wrote:

> Here are some thoughts on possible enhancements to Python. If discussion
> shows any positive enthusiasm maybe they could turn into PEPs.
>
> [assignment expression]

This has been discussed to death (at least once a year for the three years I've
been here ;-) I happen to think it's a good idea, but I have no illusions of it
ever being implemented <wink>

> [range literals]

PEP 204 (http://python.sourceforge.net/peps/pep-0204.html) Allready rejected.

> [string.chomp]

A platform independent way of doing this (that handles the case where the file
doesn't originate on the native platform) would indeed be nice, although I don't
think it's a big enough feature to require a PEP (i.e. if you create a patch it
might be accepted ;-)

-- bjorn



From ssthapa at harper.uchicago.edu  Fri Dec 15 03:25:30 2000
From: ssthapa at harper.uchicago.edu (Suchandra Thapa)
Date: Fri, 15 Dec 2000 08:25:30 GMT
Subject: Solution: Direct access to Printer I/O lines
References: <917k9b$9i9$1@nnrp1.deja.com>
Message-ID: <slrn93hrjn.2vp.ssthapa@localhost.localdomain>

jkndeja at my-deja.com <jkndeja at my-deja.com> wrote:
>CAVEATS:
>    - It's intended to work under Windows 95 only. It won't work under
>NT (you need a kernel-mode driver to access I/O ports under NT, and I
>haven't written that yet ;-). I don't know about W98. It won't work
>under linux either ;-). I'm happy to write a similar extension for
>linux if people are interested - it's good practice for me...

    It should work in w98 since I've used the same approach you seem to 
have used in w98.  It should also work in windows me.

>    - to be safe, I have restricted the range over which I/O port
>access is allowed. This is currently limited to the address spaces of
>the LPT ports:
>
>    0x278 to 0x27a (LPT1:)
>    0x378 to 0x37a (LPT2:)
>    0x3bc to 0x3bf (LPT3:)
>

    Actually there is a second range of addresses at base+300(?) that
controls ECP/EPP/SPP operations.  The specific offset is in the parallel
port faq.  I don't think people will need much access to this in python
except to check to make sure SPP mode is on or to switch to SPP mode.

>I assume you know what you're doing...
>
>    - Oh, I built it with Python v1.5.2 (ie. this is the import Library
>I used). Does this mean it will need re-building for V2.0? not sure...

    I believe you need to rebuild due to changes in the interface API.
    
>I haven't even checked yet which of the LPT: port addresses are Read-
>only, which are Write only, and which are Read/Write - that's how
>tested this is! Track down Ralph Browns Interrupt list for the gen. on
>this.

    It's even worse than this.  base+2 is the control register and certain
bits are readonly while others are read-write. The status register base+1
has a similar feature.  BTW, I'm not sure if I've gotten the addresses 
for the registers reversed.


-- 
------------------------------------------------------------------
			    |
Suchandra Thapa             | "There are only two kinds of math books. 
s-thapaNO at SPAMuchicago.edu  | Those you cannot read beyond the first 
			    | sentence, and those you cannot read 
			    | beyond the first page."
			    |                       -C.N. Yang
------------------------------------------------------------------


From grante at visi.com  Wed Dec  6 17:49:37 2000
From: grante at visi.com (Grant Edwards)
Date: Wed, 06 Dec 2000 22:49:37 GMT
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <3A2DD2CA.14A5D7E2@engcorp.com> <90l3j9$4bn$1@kapsel.intranet> <976139270.161152@clam-ext>
Message-ID: <5ozX5.1613$TU6.39593@ptah.visi.com>

In article <976139270.161152 at clam-ext>, Mark Hadfield wrote:
>"Piotr Kuchta" <piotr at itam.zabrze.pl> wrote in message
>
>> ...Dutchmen are very talented: common Dutchman speaks well two foreign
>> languages, English and German.
>
>Whereas common Australian speaks poorly one foreign language: English.

;)

Good on ya, mate!

-- 
Grant Edwards                   grante             Yow!  HELLO, everybody,
                                  at               I'm a HUMAN!!
                               visi.com            


From jh at web.de  Tue Dec  5 15:18:48 2000
From: jh at web.de (Jürgen Hermann)
Date: Tue, 5 Dec 2000 21:18:48 +0100
Subject: CGI POST binary file upload?
References: <90j8j0$cmt$1@nnrp1.deja.com>
Message-ID: <90jij9$j2$07$1@news.t-online.com>

"Bill Seitz" <fluxent at yahoo.com> schrieb im Newsbeitrag
news:90j8j0$cmt$1 at nnrp1.deja.com...
> msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)

You DO realize that a file upload, seen from the server side, is _INPUT_? ;)




From aleaxit at yahoo.com  Thu Dec 21 09:43:13 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 21 Dec 2000 15:43:13 +0100
Subject: please help
References: <91spce$tc1$1@nnrp1.deja.com>
Message-ID: <91t5ir02hc9@news2.newsguy.com>

<monamant at freemail.sx.cn> wrote in message
news:91spce$tc1$1 at nnrp1.deja.com...
> Hi,
>   I am just thinking about taking Python as my first programming
> language for serious learning (after reading "how to become a hacker"),

Good idea.

> but some distrusts with python's applicability prevent me from fully
> applying myself to it. Would you kindly help clear up some of my doubts
> about Python?

Sure!

> 1. I want to use IIS 4.0 for my webserver, can I use scripts compiled
> by Python in it?

With Mark Hammond's 'win32all' extensions (already bundled in the
ActiveState distribution of Python for win32, see www.activestate.com)
Python is installed as an 'Active Scripting Language', so you can
use it for most any task for which you could use any other Active
Scripting language (such as vbscript or jscript).  IIS 4.0 is one
example of the variety of active script host programs, and chapter
21 "Active Scripting" of Hammond and Robinson's "Python Programming
on Win32" book (O'Reilly) specifically deals on how to configure
it (virtual directory -> physical directory with the Python ASP's).

> 2. What about the ASP? If I want to use ASP, can I use Python to do
> such jobs like user authentication? How well can Python be used in Web
> programming on Windows platform?

See above.  'Yes', and 'Very very well', are the answers.  Once
you've installed Python (in the ActiveState version, or with the
separate win32all extensions that you'll also find on their site
applied on top of the standard Python 2 Win32 distribution), and
set up the virtual->physical directory, you're all set.  You'll
start your ASP page with <%@ Language=Python @> and live happily
ever after (directing your commands and requests to Application,
ObjectContext, Request, Response, Server, just as you would
for any other ASP).

> 3. I used VB in the past and you can make a '.exe' file after
> compiling. But how can I actually execute the Python program in
> Windows? It seems that I can not make a .exe file for it to execute. It
> is dependent on the interpreter?

You can try Gordon McMillan's "installer" at
    http://www.mcmillan-inc.com/install1.html
to bundle the various files a given Python application needs
into a single, 'self-unpacking-and-installing', .EXE file.

This only matters if you want to distribute applications to
others who don't have Python installed, actually.

Else, to run a foo.py program, once you have Python installed,
you enter at the command prompt (or in a .BAT, or...):
    python foo.py
normally, or
    pythonw foo.py
to execute without a console (stdin/out/err), useful for GUI
apps; and/or associate python.exe and/or pythonw.exe to the
appropriate file-extensions (.py and .pyw are normally used)
so the python file or a shortcut to it can be 'double-clicked'
(a single click may actually suffice, with Active Desktop:-).

You also have many other options, for example all those
connected with Windows Scripting Host (WSH), HTML Applications
(HTA), and so on.  That YOUR program is not itself an .EXE
(except if you package with Installer) matters little.  Python
itself is mostly a DLL, by the way; python.exe and pythonw.exe
are just tiny wrappers to run the code in the DLL -- this
matters because it's easy to access the DLL itself from your
programs (in C or C++).

But on Windows, among the other myriad options, COM shines:
with Python (and win32all; or, ActiveState's python) you can
build COM objects and instantiate-and-run-them in any way
you like from any other program (including Python ones, of
course).  COM is also how ActiveScripting works, how WSH
works, how HTML pages can embed/talk with your objects and
be scripted by you, etc, etc -- even how the graphical Shell
of Windows is interacted with.


> 4. If I really sit my self to Python, does it reward equally for the
> same effort in learning C? In the aspect of the quality of programs.

Python will let you be orders of magnitude more productive
than C at most tasks: the flip side of this is *performance*.

Python is a very high-level language, while C is pretty low
level, close to the machine and far from the human being.

Python thus lets YOU, the human being, be much faster and
more productive; while C optimizes the speed with which the
*computer* can run your programs.  (C++ tries to do both,
and _also_ keeps C compatibility; but, as a result, is HUGE
and terribly complex, alas).

For most tasks these days, by far, human productivity is
what matters -- computer power has become incredibly cheap
compared to what it was back when C was invented (as the
price/performance ratio halves every 18 months or so, in
30 years there have been about 20 halvings -- i.e., a
factor of *a million to one* or so...), while human beings
are much the same as 30 years ago.  Crucial performance
aspects, such as choice of good algorithms, are _easier_
in Python, as you can better experiment and try out various
things, so you could even end up with a better/faster
program...!-)  And when the algorithm is settled and you
*DO* need to squeeze every last cycle of computation out
of your machine, that's where extensions to Python, built
in C or C++, enter the picture, *selectively* replacing
Python code with much-speedier, lower-level code.  You will
find a lot of GREAT existing extensions for such things as
numerical work ('Numeric') and image processing ('PIL'), the
kind of tasks that are computationally very intensive...
not to speak of all the existing components which you may
access (on Windows) from Python via COM (that lets you
reuse/automate such things as Microsoft Excel's engine
to handle your computations, for example).

You may eventually want to learn C, or C++, yourself, in
order to be able to write some little but crucial very
high speed component for your own programs -- but that
will only be, say, 1/10 or 2/10 of your overall program;
most of it can stay in Python, if you wish...


Alex





From tim.one at home.com  Sun Dec 31 17:02:12 2000
From: tim.one at home.com (Tim Peters)
Date: Sun, 31 Dec 2000 17:02:12 -0500
Subject: Is 100,000 entries big for a dictionary?
In-Reply-To: <92o5cg$ooc$1@panix2.panix.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCMECDIGAA.tim.one@home.com>

[/F]
> The implementation limit is currently one billion objects.

[Aahz]
> It occurs to me, in light of the earlier thread on "is", that the
> limit is likely to hit even faster than that, depending on how id() is
> implemented.

Yes, it will be hit much faster.  id() is a 32-bit int on most machines, and
is simply the address of the object.  Since all Python objects contain at
least a refcount and a type pointer, on most machines a PyObject occupies a
bare minimum of 8 bytes, so divide your virtual address space by at least 8.
If you use nothing but ints, at least 12 <wink>.  Python also ensures that a
hash table is never more than 2/3 occupied.  So 60 million is a more
realistic-- but still wildly optimistic --upper bound (e.g., on Win32 the
user VM space spans 31 bits, not 32, so lose another factor of 2 right
there).

> ...
> Incidentally, I think you were insufficiently pedantic in that thread
> when you suggested "id(a) == id(b)".  It really should be
>
> import __builtin__
> if __builtin__.id(a) == __builtin__.id(b):
>     ....

Actually, since __builtin__ is writable, you should really arrange to
capture the then-current __builtin__.id at startup time.  Watch out for
import hooks too:  "import" may not do what you expect <wink>.

the-only-python-program-i'm-sure-about-is-"x=1"-ly y'rs  - tim




From wolfgang.grafen at marconi.com  Tue Dec  5 05:15:55 2000
From: wolfgang.grafen at marconi.com (Wolfgang Grafen)
Date: Tue, 05 Dec 2000 11:15:55 +0100
Subject: Using dictionaries
References: <3a2c9264.503856587@news-server.bigpond.net.au>
Message-ID: <3A2CC05B.DEDE735B@marconi.com>

sorular wrote:
> 
> G'day
> 
> I am trying to write a python program where a "configuration.py" will
> have all the user entered values in a dictionary format and a
> "main.py" which will import these and use it as local parameters.
> 
> But, there are quite a few values in the dictionary
> ie
> dict = { 'key1':'value1' , ' key2':'value2' , ......'keyN':'value2'}
> 
> and in the "main.py" these will be all renamed with the local names
> ie
> Val1 = dictionary.dict['key1']
> Val2 = dictionary.dict['key2']
> ....
> ValN = dictionary.dict['keyN']
>         etc....
> 
> But overall, I still have a large numbe of values (ValN) to list... !
> 
> Q1- Is there a more efficient way to do this ? I thought about using a
> for loop but that didnt work....
> 
> Q2 - would it be possible only to list the parameters in first file
> (ie dictionary.py
>         key1 = value1
>         key2 = value2
>         .....
>         keyN = valueN
> )
> and inside main.py, do a simple import...
>         import dictionary *
> so that key1 can be used inside main.py as a local parameter?
> 
> thank
> 
> HK
What about this?

>>> locals().update(dictionary)

or

>>> globals().update(dictionary)

>>> dictionary={"b":3}
>>> globals().update(dictionary)
>>> b
3


From pyth at relax.dk  Thu Dec  7 08:01:06 2000
From: pyth at relax.dk (pyth at relax.dk)
Date: Thu, 7 Dec 2000 14:01:06 +0100 (CET)
Subject: list -> tuple
In-Reply-To: <31575A892FF6D1118F5800600846864D5B1572@intrepid>
Message-ID: <Pine.LNX.4.21.0012071400000.22360-100000@relax.dk>

> > From:	pyth at relax.dk [SMTP:pyth at relax.dk]
> > My problem is, that I'm getting a list, so I have to convert it to a
> > tuple, before I can use it... any easy way to do that
> > list2tuble(list,tuple) ? :o)
>  
> tuple(list)
> 
DOH!... thank you very much... I hate when problems gets to
complicated... hehe... :o)

/Dan Larsen




From aleaxit at yahoo.com  Mon Dec 25 19:50:29 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 26 Dec 2000 01:50:29 +0100
Subject: One Python 2.1 idea
References: <mailman.977590932.19335.python-list@python.org> <9237md$orm$1@nnrp1.deja.com> <924bjt0e3n@news1.newsguy.com> <925baq$5sj$1@nnrp1.deja.com> <925tak01tpu@news1.newsguy.com> <926ici$vi9$1@nnrp1.deja.com> <92722s0l1f@news1.newsguy.com> <927se2$q7n$1@nnrp1.deja.com>
Message-ID: <928q6p02mdu@news1.newsguy.com>

<rturpin at my-deja.com> wrote in message news:927se2$q7n$1 at nnrp1.deja.com...
    [snip]
> > C++ offers lots of very-low-level constructs AS WELL
> > AS pretty high-level ones, and, as I said, the low-level
> > parts keep coming up then and again even when one is
> > trying to work more abstractly ..
>
> To me, this is a BIG reason for preferring Python to
> C++. Yes, you can implement any abstraction in C++.

Nolo contendere: I *do* prefer Python for most tasks.
Squeezing the last little bit of performance out of some
key hotspot, and interfacing to existing libraries and
frameworks that I haven't nicely dressed up yet, being
the occasional exceptions.

> When you write Python, you keep present in your mind the
> memory management strategy for sequences, the lookup,

I do keep in mind a few well-known traps that could
lead to huge performance problems (building a string
by sequentially catenating lots and lots of pieces being
the foremost example).  Apart from that, I feel no real
such need when I _write_ code -- simplicity and clarity
being by far more important issues than a few % worth
of performance.  (Occasionally, when I _tune_ code, I
may do a few experiments based on Python's internals
being a NON-black box -- I don't just thrash around at
random seeing what might make it faster, I do let my
knowledge lead me to see what may be worth trying).


> point release? If so, I think C++ might be a better
> language for you!

It may be better *when the ultimate quest for top-notch
performance is on*.  That is, AT MOST, in 10% of the
code I write.


> > .. IF I have a semantic constraint about there being
> > no deletes on this object, it's a CRUCIAL part of my
> > abstraction for it -- I want to state it clearly and
> > readably, and let compiler and human readers make of
> > my statement what they wish.
>
> But this is NOT what we are talking about. IF it is
> important functionally to some list object that there are
> no deletes from it, if deleting from it is an application
> error, then you SHOULD create a new type, and build in
> this constraint, EVEN IF it imposes a performance penalty
> to ensure this constraint. But that is NOT the issue. The

Oh yes it is.  I want *DECLARATIVE* ways to assert my
known semantic constraints: easier for me to write,
easier for the human readers to follow, easier for the
compiler to insert checking code for OR base inferences
for optimization on.  A win-win-win-win situation.

Hiding declarative knowledge in executable code is a
serious design mistake.  Are you SERIOUSLY saying
that if what I know is "for every item x of X there exists
one item y of Y such that foo(x,y)" then I should code
at each step:
    for x in X:
        for y in Y:
            if foo(x,y): break
        else:
            raise AssertionError, "no foo(x,y) for "+str(x)
...?  I want (in my dreams) the COMPILER to generate
this code for me based on a DECLARATIVE ASSERTION
using quantifiers -- when I throw some switch on it
telling to pedantically check out everything it possibly
can; and I want my declarative assertion to stand as
an interesting tidbit of information to readers AND,
who knows, possibly help some future super-duper
compiler get some extra optimization too.


> issue is the much more common case where it just so
> happens that there are no deletes from a list, in the
> code as it appears now, and this can be used for
> performance optimization. It is NOT part of the
> application's semantics, but just a happenstance of
> current code.

With Python's fluidity and non-encapsulation, 'current
code' will have to comprise every last single user
typed string whose exec might possibly change this
'happenstance'.  Talk about whole-program optimizers...
_at the very least_, a declarative assertion about
lack of access to 'this list here' "from the outside" is
going to be needed for any practical purposes --
"nobody but us chickens is going to mess with X --
feel free to infer everything about X based on this
*finite* set of code and/or to hide it from outside".


Alex





From tim at degree.ath.cx  Wed Dec 27 21:07:13 2000
From: tim at degree.ath.cx (Tim Hammerquist)
Date: Thu, 28 Dec 2000 02:07:13 GMT
Subject: Python vs PHP (& Java?)
References: <mailman.977897649.4877.python-list@python.org>
Message-ID: <slrn94l86m.a3h.tim@degree.ath.cx>

Ben Ocean <zope at TheWebsons.com> wrote:
> I'm an intermediate-level programmer that's new to Python (via Zope) and I
> like it a lot. However, it appears to me that, while Python is a good
> replacement for Perl, it cannot replace much of the functionality of PHP,
> which can be scripted into HTML (whereas Python, it appears, must be
> called).

Ok, everyone else made very good points, but this little comment slipped
out (not that I'm surprised, honestly)  =)   There is no "replacement
for Perl!"  And Python would do a poor job trying to be Perl (and vice
versa), nor does anyone here seem to _want_ Python to be/replace Perl.

Let me paraphrase:
Perl is Perl and should remain Perl.
Python is Python and should remain Python.
Java is long-winded and...um, "pragmatic" and should be given a fair
shot.
PHP is crap and should be left to those to don't want control over their
scripts.  (Ok, this last one was significantly biased; did ya notice?)

Just my $.02.  =)

And Ben: don't leave language choices up to others...unless "others" are
your boss, of course.  Research the languages, not just hearsay.
-- 
-Tim Hammerquist <timmy at cpan.org>
I am opposed to millionaires, but it would
be dangerous to offer me the position.
	-- Mark Twain


From merlyn at stonehenge.com  Fri Dec 15 10:46:57 2000
From: merlyn at stonehenge.com (Randal L. Schwartz)
Date: 15 Dec 2000 07:46:57 -0800
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <tcmay-FCA22F.14073713122000@news.got.net> <918v7a02scn@news1.newsguy.com> <tcmay-04E80C.16181313122000@news.got.net> <91a1ed02a4j@news2.newsguy.com> <3A389B1A.4F33FFD4@cfmu.eurocontrol.be> <91ae8s02llf@news2.newsguy.com> <3A38CF3E.8BBD2216@cfmu.eurocontrol.be> <91avrn08ql@news2.newsguy.com> <3A39EAB0.1A0528C3@cfmu.eurocontrol.be>
Message-ID: <m1snnplma6.fsf@halfdome.holdit.com>

>>>>> "Ian" == Ian Wild <ian at cfmu.eurocontrol.be> writes:

>> > classes exist only by convention [in Perl]
>> 
>> 'only by convention'!?  Have they abrogated the 'bless'
>> keyword, now?

Ian> My understanding is that 'bless' does some internal
Ian> tweakery of a magic array (@INC).  If you promise not
Ian> to look too hard, you've got classes.  Anything that
Ian> needs a promise from me I'll regard as convention, not
Ian> a language construct.  YM, as they say, MV.

'bless' changes a portion of the datastructure holding the object
so that it remembers which package it's associated with, which
then permits runtime location of a method called against that object.

And isn't *everything* a "promise" at some point?  I don't see
anything less direct about Perl's object structure than say, C++'s
version of the same thing.  Isn't a C++ object just a struct behind
the scenes?

You dismiss Perl too easily.  As a hybrid (some objects, some not)
language, it's a very usable object model.  It's not smalltalk, which
is a cleaner interface, but it's certainly more usable *and*
introspective than C++ (which Larry Wall promised us Perl's OO would
*not* emulate {grin}), and more flexible (and therefore useful) than
Java's hybrid object model.

-- 
Randal L. Schwartz - Stonehenge Consulting Services, Inc. - +1 503 777 0095
<merlyn at stonehenge.com> <URL:http://www.stonehenge.com/merlyn/>
Perl/Unix/security consulting, Technical writing, Comedy, etc. etc.
See PerlTraining.Stonehenge.com for onsite and open-enrollment Perl training!


From spahievi at vega.bg  Tue Dec  5 16:01:24 2000
From: spahievi at vega.bg (Niki Spahiev)
Date: Tue, 5 Dec 2000 23:01:24 +0200
Subject: Smalltalk/ROSE-like prog. environment?
In-Reply-To: <90j2gg$kpf$1@news.rz.uni-karlsruhe.de>
References: <90j2gg$kpf$1@news.rz.uni-karlsruhe.de>
Message-ID: <7843403270.20001205230124@vega.bg>

05.12.2000, 17:44:16, Bjoern Giesler wrote:

BG> Hi,

BG> I have a dream... :-)

BG> I'd like to have a programming environment that allows you to do graphical
BG> OO design (UML, ROSE-like) with access to all classes a project consists of
BG> (Smalltalk-like). This thing would allow the programmer to graphically
BG> design class interfaces and interactions, and edit method source code on a
BG> per-method basis. I.e. no editing-around in large .py files but working just
BG> on the job at hand.

BG> This would have a couple of other nifty features: A test mode in which an
BG> exception is presented to the user in a dialog box that allows modification
BG> of the code _while the program is running_ (this could be done with
BG> continuations, I believe) and stuff like that.

BG> This is a very hazy notion that has been floating around in my head for a
BG> while. I believe that Python would be the ideal language for this kind of
BG> application. Does anybody know of such an environment?

BG> If not, I'd like to give it a shot. Anybody have any ideas or want to
BG> collaborate with me?

BG> Regards,
BG>                 --Bj?rn

Closest thing to this dream (but not too close) is called SNiFF+. It
has personal version for linux and groks Python.

-- 
Best regards,
 Niki Spahiev





From mdtompkins at home.com  Fri Dec 15 18:08:29 2000
From: mdtompkins at home.com (Mark Tompkins)
Date: Fri, 15 Dec 2000 23:08:29 +0000
Subject: win32wnet.WNetAddConnection2, win32wnet.WNetCancelConnection2
Message-ID: <3A3AA46C.5E1F2F8C@home.com>

Hi,

Well, the last time I called, I couldn't map a local drive to a remote
network share.  I was able to solve this problem by assigning the
following privileges first to the Administrator group using the user
administrator, and then, from within the service itself, changing the
privileges of the service process (SYSTEM and LocalSystem are members of
the local Administrators group by default, I believe):


        # SE_TCB_NAME - Act as Part of the Operating System
        # SE_CHANGE_NOTIFY_NAME - Bypass Traverse Checking (This
privilege is granted to all users by default).
        # SE_ASSIGNPRIMARYTOKEN_NAME - Replace a Process Level Token

Next I try to create a connection to a remote share

win32wnet.WNetAddConnection2(win32netcon.RESOURCETYPE_DISK,'Y:','\\\\clun1-172\\source',None,'orabatch','orabatch')

This works just great.

Then, I try to cancel the connection

win32wnet.WNetCancelConnection2('Y:',0,1)

This fails, with error (the error message gives the pywintypes.api_error
information, wrapped by my error message):

Unable to delete net file share: Y: + Code: 2250 + Func:
WNetCancelConnection2 + descr: This network connection does not exist.

I've tried running the service as default (SYSTEM account - a local
account only), (LocalSystem account) and orabatch, an account that is a
domain account that is enabled to make the connection to the remote
share.

Why is the current process not owner of the connection that it has just
created???? Why can't the current process see the connection that it has
just made?

What must I do so that when orabatch user creates a connection to a
remote share, that the local drive (in this case Y:) can be deleted from
within the service?

thx

Mark




From jwbnews at scandaroon.com  Wed Dec  6 00:13:46 2000
From: jwbnews at scandaroon.com (John W. Baxter)
Date: Tue, 05 Dec 2000 21:13:46 -0800
Subject: Rounding error in Python 2.0
References: <90jscs$uno$1@nnrp1.deja.com> <AnjX5.2006$O%.346064@newsc.telia.net>
Message-ID: <jwbnews-9B1BDE.21134605122000@news.olympus.net>

In article <AnjX5.2006$O%.346064 at newsc.telia.net>, "Fredrik Lundh" 
<fredrik at effbot.org> wrote:

>     A. Binary floating point cannot represent decimal fractions
>     exactly, so some rounding always occurs (even in Python
>     1.5.2).

Well, 20% of the numbers with a single digit to the right of the 
decimal, and 4% of the numbers with two digits to the right of the 
decimal (and so on) are precise.  It's only the other 80% or 96% or... 
which aren't.  ;-)  --John

-- 
John W. Baxter   Port Ludlow, WA USA  jwbnews at scandaroon.com


From andrew_dot_henshaw_at_earthling_dot_net  Tue Dec 19 21:36:49 2000
From: andrew_dot_henshaw_at_earthling_dot_net (Andrew Henshaw)
Date: Tue, 19 Dec 2000 21:36:49 -0500
Subject: while true: !!!
References: <mailman.977127721.4398.python-list@python.org> <DnD%5.166633$U46.5342944@news1.sttls1.wa.home.com> <ckJ%5.3097$Yv6.81688@e420r-atl1.usenetserver.com>
Message-ID: <t4073164g9shaf@corp.supernews.com>

"Steve Holden" <sholden at holdenweb.com> wrote in message
news:ckJ%5.3097$Yv6.81688 at e420r-atl1.usenetserver.com...
...snip
> I've always tried to discourage students from comparing a Boolean with a
> constant, but it can be difficult to get the point over that (a) can
> evaluate as true even though (a == some_constant) may well evaluate to
> false.
>
> Clearly, if what you have is a Boolean value (i.e. one which can be used
as
> the expression in an "if" statement) then it's adding inefficiency to
> compare it with something, when what should be used is either
>
>     if a:
>
> or
>
>     if not a:
>
> This works quite nicely in Python:
>
> >>> not "forever"
> 0

I had never thought of this before, but it is kind of interesting that
    not 'this is a true string'
doesn't produce
    ''
but instead produces 0.

Similarly, not [0, 1, 2] could produce [] .

In fact, I could see where that might be useful.  On the other hand (to be
consistent), what would
    not []
produce? :)

Also, shouldn't false == false ?
    '' == []
returns 0.
  :)

Well, at least
    (not '') == (not [])
returns 1.


Andrew Henshaw




From sill at localhost.kitenet.net  Tue Dec  5 17:54:33 2000
From: sill at localhost.kitenet.net (Oldayz)
Date: Tue, 05 Dec 2000 22:54:33 GMT
Subject: Newbie question on 'cmpfiles':
References: <90i074$d73$1@nnrp1.deja.com>
Message-ID: <JmeX5.39553$II2.3513820@newsread2.prod.itd.earthlink.net>

On Tue, 05 Dec 2000 05:59:02 GMT, harisri at my-deja.com <harisri at my-deja.com> wrote:
>Hello,
>
>I am a newbie user to python 2.0
>
>Can someone help me by explaining 'cmpfiles' function with a simple
>example?
>
>I am trying to compare 2 directories, but not sure about the syntaxof
>this function. The "Python Reference Library" did not help me on this.
>
>[What does "common" mean in "cmpfiles(dir1, dir2, common)"?]

This function is used to compare *files*, not directories. Multiple files
in these directories.

To compare directories, use dircmp class:

dircmp (a, b[, ignore[, hide]])
Construct a new directory comparison object, to compare the
directories a and b. ignore is a list of names to ignore, and
defaults to ['RCS', 'CVS', 'tags']. hide is a list of names to
hid, and defaults to [os.curdir, os.pardir].

report ()
Print (to sys.stdout) a comparison between a and b.

etc.. It's in 2.0 lib ref under filecmp

>
>Thanks in advance,
>Hari.
>
>
>Sent via Deja.com http://www.deja.com/
>Before you buy.


-- 

	Andrei


From com at gahnstrom.se  Sat Dec 30 16:36:14 2000
From: com at gahnstrom.se (Tim Gahnström)
Date: Sat, 30 Dec 2000 22:36:14 +0100
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <3A4D6A05.EEC97FA6@uswest.net> <slrn94qqvg.mgg.grey@teleute.rpglink.com> <ije36.96813$x6.43604873@news2.rdc2.tx.home.com> <slrn94qt4f.mgg.grey@teleute.rpglink.com> <CMo36.97857$x6.44474719@news2.rdc2.tx.home.com> <slrn94s9qg.134.grey@teleute.rpglink.com> <gdq36.2696$Qb7.302774@newsb.telia.net> <slrn94shbe.150.grey@teleute.rpglink.com> <92li6g$hju$1@yggdrasil.utfors.se>
Message-ID: <92lkd8$ifh$1@yggdrasil.utfors.se>

>The only way to acieve this exact behavior with is while is like this

uops I got a nose punc for this :-)

>a=1
>while a=1 or (T >=25)
>    a=0
>    R=Themistor.ReadResistance()
>    T=R_to_Temperature(R)

First of all, this doesn't produce the exact same behavior unless python
uses lazy evaluation (does it?)
Two this is ofcourse not the ONLY way to acive the behavior.
Bj?rn Peterssen said another one in the first message.
I have just benn tought to not use goto and I counted break as a nobler art
of goto :-)
Tim



From aleaxit at yahoo.com  Tue Dec 19 08:29:41 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 19 Dec 2000 14:29:41 +0100
Subject: while true: !!!
References: <mailman.977182622.11891.python-list@python.org>
Message-ID: <91nnsu016ji@news2.newsguy.com>

"Delaney, Timothy" <tdelaney at avaya.com> wrote in message
news:mailman.977182622.11891.python-list at python.org...
> > >The idiom is actually
> > >
> > >while 1:
> > >
> > > do something
> > >
> > > if condition:
> > > break
> > >
> > > do something else
    [snip]
> I go well out of my way to avoid this idiom, as I feel you should always
> have the termination condition at one end of the loop, preferably the top

I disagree intensely with this stance: I strongly feel that the
loop-exit condition should be *in the most natural place for it*,
which is often the top, but not all that rarely elsewhere.

Anyway, one doesn't have to go 'well' out of one way to hide
this clean, general loop structure as if it was otherwise --
one 'status-variable' will suffice, e.g:

looping = 1
while looping:
    do something
    if condition:
        looping = 0
    else:
        do something else

I think this is *substantially* less readable than the
clean, natural form:

while 1:
    do something
    if condition: break
    do something else

but it does satisfy fetishistic adherents of the Jacopini
and Boehm school (and, I believe, _is_ isomorphic to the
local transformation J&B's second lemma would implicitly
perform on the flow diagram of the second form to produce
that of the first).


Alex





From ahopkins at dynacare.com  Tue Dec 26 15:07:30 2000
From: ahopkins at dynacare.com (Albert Hopkins)
Date: 26 Dec 2000 14:07:30 -0600
Subject: Formatting for hh:mm:ss
References: <Pine.GSO.4.21.0012261309370.13920-100000@z.glue.umd.edu>
Message-ID: <slrn94hu4r.17s.ahopkins@ahopkins.dynacare.com>

On Tue, 26 Dec 2000 13:12:33 -0500, Roy Katz <katz at Glue.umd.edu> wrote:
>Hello,
>
>
>I've searched the 'time' and 'locale' modules  for any function which
>would take a difference of two time.time()'s and format the result as
>hour:minutes:seconds, but I can't find any.  All I am doing is this:
>
>  t1 = time.time()
>  # some processing here
>  t2 = time.time()
>
>  total = t2 - t1
>
>  print_time( total )  
>
>where print_time() would format the time as above. Does anyone know if
>this function already exists?
>

It sounds to me that you can accomplish what you need with the 
time.strftime() function.  Of course you'll need to convert 'total' to a 
tuple to pass to the function.  Something like

def print_time(t)
	return time.strftime("%H:%M:%S", time.gmtime(t))


-- 
                                                     Albert Hopkins
                                             Sr. Systems Specialist
                                              Dynacare Laboratories 
                                              ahopkins at dynacare.com



-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----


From aleaxit at yahoo.com  Mon Dec 11 12:13:07 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 11 Dec 2000 18:13:07 +0100
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A350405.B8F09EC7@gte.net>
Message-ID: <9131vd0pnq@news2.newsguy.com>

"Steve Williams" <sandj.williams at gte.net> wrote in message
news:3A350405.B8F09EC7 at gte.net...
    [snip]
> I wouldn't call "While true" a python-idiom, I'd call it a
> python-solecism.

With some effort, I guess one MIGHT find assertions with
which I'd disagree more drastically than this one -- but,
it WOULD take effort.

How *WOULD* you prefer to express Knuth's classical
"N-times-and-a-half" loop (in Python, C, C++, Java, &c)
if not by (pseudocode):

    'loop until otherwise terminated':
        'first-part' suite of statements
        if 'termination-condition': break
        'second-part' suite of statements

...?

The 'loop until otherwise terminated' clause is _best_
spelled
    while 1:
in Python, while other languages offer you a choice
between
    for(;;)
and
    while(1)
as well as
    do
(with a while(1) at the end), etc.


What do you claim is "barbarous", "grammatically faulty",
etc, about using "while 1:" to express "loop until
otherwise terminated" in Python?


Alex





From tcmay at got.net  Sat Dec  9 01:54:34 2000
From: tcmay at got.net (Tim May)
Date: Fri, 08 Dec 2000 22:54:34 -0800
Subject: simulated annealing
References: <udn1eluhzo.fsf@yogi.pc1.uni-duesseldorf.de> <roy-9B2D39.19561027112000@news.panix.com> <tcmay-98F57B.17064527112000@news.got.net> <udvgt8xr1t.fsf@yogi.pc1.uni-duesseldorf.de> <tcmay-849C4E.19535003122000@news.got.net> <udwvddompp.fsf@box.home.de>
Message-ID: <tcmay-BFA4BD.22543408122000@news.got.net>

In article <udwvddompp.fsf at box.home.de>, "Jochen K?pper" 
<jochen at pc1.uni-duesseldorf.de> wrote:

> Tim> Glad you have "found" Google. I had been hearing about how great it 
> was 
> Tim> for several months before trying it. 
> 
> Me too.
> 
> Tim> When I finally did, I realized how much better it was than Alta
> Tim> Vista and Metacrawler.  (I used to like Metacrawler because it,
> Tim> ostensibly, searches with N different search engines. In
> Tim> practice, it doesn't find as much as just Google does.)
> 
> It actually searches google! And that is why I was confident it was
> the right thing to use metacrawler...

Don't be misled by this claim. While it is claimed that Metacrawler 
searches N different other search engines, it often (usually) returns 
fewer results than Google does.

How can this be? Without researching the ins and outs carefully, I 
surmise it's the limitations the search engine companies place on 
Metacrawler using their system. After all, Yahoo, Infoseek, Lycos, and 
the other search engine companies make some/most of their money from 
advertising. So if Metacrawler searches N of them and puts its _own_ 
advertisements on the results page....

Metacrawler often reports "search engines that time out..."

In any case, the moral is to try many different engines...and then 
"anneal" the results, so to speak. Curiously appropriate for the topic 
at hand. (With such a small sample set, of perhaps 10 major search 
engines, SA is obviously not needed.)

--Tim May

-- 
---------:---------:---------:---------:---------:---------:---------:----
Timothy C. May              | Crypto Anarchy: encryption, digital money,  
ComSec 3DES:   831-728-0152 | anonymous networks, digital pseudonyms, zero
W.A.S.T.E.: Corralitos, CA  | knowledge, reputations, information markets, 
"Cyphernomicon"             | black markets, collapse of governments.


From pbjorn at uswest.net  Sun Dec 24 15:45:04 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Sun, 24 Dec 2000 13:45:04 -0700
Subject: Some Python 2.1 ideas
References: <mailman.977590932.19335.python-list@python.org> <Cw616.23054$qN6.3308205@typhoon.we.rr.com> <3A452881.7686E717@uswest.net> <%vb16.187233$U46.5801214@news1.sttls1.wa.home.com> <3A464D5D.B541C04E@uswest.net> <20001224210742.A18246@xs4all.nl>
Message-ID: <3A46604F.CF0A1DA1@uswest.net>

Thomas Wouters wrote:

> On Sun, Dec 24, 2000 at 12:24:13PM -0700, Bjorn Pettersen wrote:
> > Greg Jorgensen wrote:
>
> > > "Bjorn Pettersen" <pbjorn at uswest.net> wrote in message
> > > [snip]
> > > How should these be handled?
>
> > >   "foo\n\n\n"
> > >   "foo\n\r\n\r"
> >
> > "foo\n\n\n" -> "foo\n\n"
> > "foo\n\r\n\r" -> "foo\n\r"
>
> Note that the latter was a trick question! The proper order for DOS-EOL is
> '\r\n', *not* '\n\r'. So the string "foo\n\r\n\r" is a Macintosh string, not
> a DOS one :-)

Sorry, I thought I remembered that Macintosh used \n\r and not just \r. Shows you how
much I work with Macs ;-)

I think the rule I was trying to follow was "the longest sequence that can be
recognized as a line terminator".

-- bjorn

ps: anyone know what the line terminator for OS X is? (My old NeXT used the regular
unix conventions..)




From fredrik at effbot.org  Sat Dec 16 12:48:29 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sat, 16 Dec 2000 17:48:29 GMT
Subject: Unicode and default encodings
References: <91g9e1$4t2$1@news1.xs4all.nl>
Message-ID: <NVN_5.199$u17.33869@newsb.telia.net>

Boudewijn Rempt wrote:
> I'm currently trying to work with Unicode strings, and
> I was kind of wondering where sys.setdefaultencoding has
> gone to - it's still mentioned in Misc/unicode.txt, but
> it isn't there anymore ;-)

You can use it in sitecustomize.py, but not in user code.
See Lib/site.py for details.

</F>




From eq3pvl at eq.uc.pt  Tue Dec  5 05:27:16 2000
From: eq3pvl at eq.uc.pt (Pedro Vale Lima)
Date: Tue, 05 Dec 2000 10:27:16 +0000
Subject: Programming with Python
References: <90fq39$jgq$1@nnrp1.deja.com> <9iOW5.55847$L26.16633720@typhoon.southeast.rr.com> <90gj77$7a0$1@nnrp1.deja.com> <3A2BFD12.7AB3A628@anubis.network.com>
Message-ID: <3A2CC304.8C451CD6@eq.uc.pt>

"Joseph A. Franzmeier" wrote:

> The other to recommend is  "The Python Reference"  by  Guido Van
> Rousson
> the creater of Python.

I never heard of this one. Are you talking of the "Library Reference"
coming with the
python docs, or are you talking about "Python Essential Reference" by
Beazley the author
of SWIG (by the way, this is the book I always carry around).

The best book for learning is Programming Python, it's a bit dated but
still you learn all from
the basic to the expert. ORA is probably going to do a 2nd edition.

regards
pedro



From prestonlanders at my-deja.com  Fri Dec 22 11:15:27 2000
From: prestonlanders at my-deja.com (Preston Landers)
Date: Fri, 22 Dec 2000 16:15:27 GMT
Subject: mail processing module?
References: <91rcig$rgu$1@nnrp1.deja.com> <m266kd45m0.fsf@freak.kaiserty.com>
Message-ID: <91vumo$di3$1@nnrp1.deja.com>

In article <m266kd45m0.fsf at freak.kaiserty.com>,
  Michael Haggerty <mhagger at alum.mit.edu> wrote:

> Is there a reason to put sending and retrieving into a single module?
> They seem easily factorable.

You're really going to like this: it's all methods of one object. You
see, my main design goal here was to make working with mail as simple as
possible.

> In fact maybe a better breakdown would be one module that represents
> an email message and allows data to be MIMEd into and out of it, one
> module for sending such messages via SMTP, one module for retrieving
> such messages via POP, and one for retrieving via IMAP.

All of these modules exist already, are widely used and well tested.  My
module uses them.  Why would I abstract each module up a level but keep
them separate?

> interfaces for the two retrieving modules should be as close as
> possible, and ideally the interface to the mail-sending routine should
> be generic enough that one could write a module with the same
> interface but that sends emails by writing them to, e.g., sendmail or
> qmail.

My maillib module unifies the interface to POP and IMAP using a
technique that hard core OOPers will find abhorent:

if self.is_pop:
   do_one_thing()
elif self.is_imap:
   do_another()

Currently, only sending with smtp is supported but I could easily put in
support for using a command line mail sender.  Also, you could subclass
the maillib class and reimplement the sending methods.

> (Some of this functionality might overlap with existing Python
> modules.)

Ya don't say? ;-)  I plan to use them.  If all goes well I should be
able to post some code (or a pointer to it) within a couple of weeks,
and everyone will have a chance to show me the error of my ways then.
;-)

Like I said, I am trying to:

a) reimplement as little as possible

b) make the interface for using the module as simple as possible

Here's an example of use (sorry if there are formatting problems):

 import maillib

 # make mail exception availible in this namespace
 MailError = maillib.MailError

 # set up a configuration dictionary with the recommended options

 config_dict = {

     'sendmail_from': 'Your Name <your at address.here>',
     'sendmail_smtphost': 'smtp.yourdomain.com',
     'sendmail_user_agent':  'Catchy Slogan Here',

     'getmail_method': 'imap',  # or 'pop'
     'getmail_user': 'uname',
     'getmail_passwd': 'pwd',
     'getmail_server': 'imap-server.yourdomain.com',
     'getmail_server_port': 143, # standard imap port, 110 is POP port

     }

 Debug = None  # set to 1 for more verbose output to stderr
 my_mail = maillib.maillib(config_dict, Debug)  # you now have a maillib
object

 ### SENDING MAIL

 mailto_list = ['destination at address.com']
 subject = 'Hello, maillib!'
 mail_body = 'Hello, world.'

 attachment_list = [] # no attachments, empty list or None

 ### ATTACHMENT 1

 description_1 = 'Picture of Tux the Penguin'
 filename_1 = 'tux.jpg'
 mimetype_1 = None # let the module guess from the filename, only works
with common types
 item_1 = open_file_handle.read()
 attachment_1 = (description_1, mimetype_1, filename_1, item_1)

 attachment_list.append(attachment_1)

 try:
    my_mail.send_mail(mailto_list, subject, mail_body, attachment_list)
 except MailError, msg:
    print 'There was a mail sending problem'
    print msg
    raise

 print 'Mail sent ok.'


 # RECIEVE EMAIL

 try:
    my_mail.getmail_login()
    Delete = 1 # delete the message after you fetch it
    raw_message = my_mail.fetch_first_msg(Delete)
    my_mail.do_cleanup()  # logs off mail server!
 except MailError, msg:
    print 'There was a problem retrieving the email.'
    print msg

 ### The raw unprocessed message is now in raw_message

 ### DECODING MESSAGES

 try:
   (header_dict, message_body, attachments_list) =
my_mail.decode_email(raw_message)
 except MailError, msg:
    print 'There was a problem decoding the email.'
    print msg

  print 'From: ', header_dict.get('From', 'unknown')
  print 'To: ', header_dict.get('To', 'unknown')
  print
  print message_body


--- Preston <pibble at yahoo dot com>


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


From fredrik at effbot.org  Sun Dec 31 07:37:48 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sun, 31 Dec 2000 12:37:48 GMT
Subject: Line Benchmarks - Tkinter vs. wxPython
References: <91uljk$12as$1@news2.ottawa.cyberus.ca>
Message-ID: <wMF36.2800$Qb7.322181@newsb.telia.net>

Gordon Williams wrote:
> In November of 1999 I did some benchmark tests to gauge the performance
> difference between Tkinter and wxPython when large numbers of lines are
> plotted.

As most Tk/Tkinter users know, the Canvas widget isn't the best
thing to use for widgets like this.  There are several Tk extensions
that provide better support for rapidly changing data, including
BLT.

And then there's Tkinter 3000.

As some of you know, I've been involved in a commercial UI library
called uiToolkit (see www.pythonware.com for more info). Tkinter
3000 is a reimplementation of some uiToolkit concepts, on top of
Tkinter (and/or Tk).

One such concept is something called the "Widget Construction
Kit", which is a small interface that allows you to implement new
Tkinter widgets in pure Python.

Using my current "proof of concept" implementation of the WCK,
I implemented a simple PlotWidget, and used it to run Gordon's
benchmark.

Here are the results on my Win2K box:

    Python/Tkinter 1.5.2: 0.4356 seconds
    Python/Tkinter 2.0: 0.0832
    Tkinter 3000, Tk driver, floating point coordinates: 0.0018
    Tkinter 3000, Tk driver, integer coordinates: 0.0016

I don't have a working wxPython on this computer, but it should
be clear that Tkinter can match wxPython also on a benchmark
like this, if you use the right widget...

(in either case, 625 fps should be enough for anyone ;-)

For more info on Tkinter 3000, see

    http://tkinter.effbot.org (the project)
    http://online.effbot.org#1811114 (benchmarks)
    http://www.egroups.com/subscribe/tkinter3000 (mailing list)

</F>




From pbjorn at uswest.net  Wed Dec 27 22:15:40 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Wed, 27 Dec 2000 20:15:40 -0700
Subject: Some Python 2.1 ideas
References: <3A452881.7686E717@uswest.net> <%vb16.187233$U46.5801214@news1.sttls1.wa.home.com> <3A464D5D.B541C04E@uswest.net> <mailman.977688488.25438.python-list@python.org> <20001225150250.6FA06A84F@darjeeling.zadka.site.co.il> <mailman.977760789.10224.python-list@python.org> <roy-EA8C3B.11361325122000@news.panix.com> <3A479C29.2631F27B@uswest.net> <mailman.977876470.23568.python-list@python.org> <3A4A50E6.11BBE0FD@uswest.net> <mailman.977962871.16241.python-list@python.org>
Message-ID: <3A4AB05C.D41F4022@uswest.net>

D-Man wrote:

> Yeah, it is a problem if you don't know where the file came from.  How
> about creating a standard text file format?  All files could be opened
> in binary mode on all platforms, then EOL is handled by each
> application (or better yet make a standard module that does the
> implementation correctly once).

Feeling like fighting windmills today, eh <wink>? I think the current wish is to
do it from the other end, i.e. instead of relying on the text file format to be
"correct", implement a method on string object that is smart enough to handle all
text file format line endings intelligently.

-- bjorn




From sill at localhost.kitenet.net  Thu Dec  7 19:19:32 2000
From: sill at localhost.kitenet.net (Oldayz)
Date: Fri, 08 Dec 2000 00:19:32 GMT
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu> <slrn92tu7t.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com> <3A2FE630.5C113AB@bioeng.ucsd.edu>
Message-ID: <oOVX5.95$U4.1605@newsread1.prod.itd.earthlink.net>

On Thu, 07 Dec 2000 11:34:08 -0800, Curtis Jensen <cjensen at bioeng.ucsd.edu> wrote:
>"A.M. Kuchling" wrote:
>> 
>> On Wed, 06 Dec 2000 17:06:55 -0800, Nathan Gray <nospam at caltech.edu> wrote:
>> >It's really upsetting how often I've seen this sort of response to a polite
>> >question on this newsgroup lately.  Tell me, what's so terrible with asking a
>> >group of experts for pointers when researching a subject?  It's a perfectly
>> >legitimate and highly effective way to learn about something.  There's plenty
>> 
>> I've noticed this trend, too, though T.C. May, the person you're
>> replying to, has always had a low tolerance for cluelessness, as I
>> noted when lurking on the cypherpunks mailing list in the early 90s;
>> May's posts were often informative and entertaining, but not often
>> placid.  (I was pleased to turn up here, though I'm not sure why he's
>> taken up Python.)
>
>A while back, someone wrote in and asked what a text editor was.  The
>answers to this were polite and informative, and non-demeaning.  I'm
>curious why this one and why others get ripped on, and others don't. 
>Luck of the draw?
>

Question like 'what's a text editor' shows that the inquieror (is that a
right word?) is a newbie, while a question like 'guys I'm writing a paper
on python, can you tell me where it comes from?' shows laziness. By the way,
I tried google search for 'what's a text editor' and out of top 10 hits, only
one explained the difference and was titled 'word processor vs. text editor',
in other web searches you'd probably be even less lucky, while search on
'history of python' came up with self-descriptive    [40]History of:
GreatSoftware - Python.

-- 

	Andrei


From micsaund at home.com  Wed Dec 13 20:51:49 2000
From: micsaund at home.com (Mike Saunders)
Date: Thu, 14 Dec 2000 01:51:49 GMT
Subject: Linux version would be really cool
References: <917k9b$9i9$1@nnrp1.deja.com>
Message-ID: <VIVZ5.22058$x6.12697491@news2.rdc2.tx.home.com>

> haven't written that yet ;-). I don't know about W98. It won't work
> under linux either ;-). I'm happy to write a similar extension for
> linux if people are interested - it's good practice for me...

Since you mentioned it :^)

I think that with all the embedded Linux projects, and
the fact that Linux makes it so easy to pipe stuff
together, I think a Linux version of your program would
be very useful.  I, for one, would like to add it to
my repetoire (sp?) for future projects.

Thanks for taking the time to work on your .dll and
possibly the Linux version!!!!

Mike





From rodsenra at correionet.com.br  Wed Dec 13 21:48:08 2000
From: rodsenra at correionet.com.br (Rodrigo Senra)
Date: Thu, 14 Dec 2000 00:48:08 -0200
Subject: Parsing Java source files in Python ?
Message-ID: <3A3834E8.19C9D495@correionet.com.br>

Hi everybody,

 I would appreciate any pointers to tools or
 strategies in order to parse Java source
 files from Python (from CPython not Jython,
 I know this may sound a little bit awkward,
 but other requirements made me rule out Jython
 for the time being).

 Do you think using regualr expressions is  feasible,
 rephrasing..., is advisable ?

TIA
Rod
-- 
Rodrigo Senra         
Computer Engineer   (GPr Sistemas Ltda)  rodsenra at correionet.com.br 
MSc Student of Reflection (IC- UNICAMP) Rodrigo.Senra at ic.unicamp.br
http://www.ic.unicamp.br/~921234 (see also   http://www.gpr.com.br)



From syver at NOSPAMcyberwatcher.com  Sun Dec 10 15:19:50 2000
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Sun, 10 Dec 2000 21:19:50 +0100
Subject: reloading modules question
References: <90scdf$hpe$1@troll.powertech.no> <ouk43t494isd9m8fk92udjjbg2pb6crq7q@4ax.com>
Message-ID: <910om3$2vk$1@troll.powertech.no>

> I would suggest you are better off just stopping and starting the affected
web sites. That's what we
> do and it only takes a second.

You mean something like pushing the unload button in the applications
property sheet?
Starting and stopping the default web site in inetmgr is not enough.

> You can automate the procedure using WMI or ADSI. I'm sure I used to have
an example script but I'm
> buggered if I can find it right now.
I think I saw some documentation on it in the documentation for IIS. I'll
look it up.





From fredrik at effbot.org  Sun Dec 10 05:34:06 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sun, 10 Dec 2000 10:34:06 GMT
Subject: sgmllib nit
References: <eyErVLAEHlM6EwA4@jessikat.fsnet.co.uk> <EAsY5.2808$O%.435892@newsc.telia.net> <JagDRIAEbuM6EwxY@jessikat.fsnet.co.uk>
Message-ID: <y_IY5.2910$O%.457391@newsc.telia.net>

Robin Becker wrote:
> I realise that it does that! Unfortunately Zope's dtml isn't HTML.

after a bit more research, I found this little piece of text from
the SGML specification:

    If "SHORTTAG YES" is specified on the SGML declaration,
    the name and [the separator] can be omitted if the
    attribute value specification is an undelimited name token
    that is a member of a group specified in the declared value
    for that attribute.

or in other words, you can leave out the *attribute name*, not
the value...

(this is why you cannot use the same enumeration value for
more than one attribute in the same tag)

...so if DTML expects

    <dtml-var standard_html_header>

to mean

    <dtml-var standard_html_header=''>

instead of

    <dtml-var standard_html_header='standard_html_header'>

it's neither SGML, HTML, nor XML.

> Which parser should one use to analyse dtml which can contain tags like
> <dtml-var standard_html_header>?

dtmllib?

I suggest asking the Zope folks if/why they're using an embraced
and extended syntax, and what parser they recommend...

if you want to fix this in sgmllib, the id hack is way too ugly
-- better move

            if not rest:
                attrvalue = attrname
            elif attrvalue[:1] == '\'' == attrvalue[-1:] or \
                 attrvalue[:1] == '"' == attrvalue[-1:]:
                attrvalue = attrvalue[1:-1]
            attrs.append((string.lower(attrname), attrvalue))

into a separate method, and override it from a dtmllib module.

another way to solve it (in dtmllib) would be to have a map of
attributes that are supposed to be minimized (derived from the
DTML DTD).  just use a simple dictionary:

    minimize = {
        ("dtml-var", "standard_html_header"): 1,
        ...
    }

    ...

    if attrname == attrvalue and minimize.has_key((tag, attrname)):
        out.write(" %s" % attrname)
    else:
        out.write(" %s=%s" % (attrname, repr(escape(attrvalue))))

    ...

> I guess I need to write my own parser that doesn't make unnecessary
> assumptions.

well, if you think that assuming SGML is an "unnecessary assumption"
for something called sgmllib, there's not much I can do to help.

</F>




From Bjoern.Giesler at gmx.de  Thu Dec  7 05:12:22 2000
From: Bjoern.Giesler at gmx.de (Bjoern Giesler)
Date: 7 Dec 2000 10:12:22 GMT
Subject: MVC module anyone?
Message-ID: <90nnq6$8n0$1@news.rz.uni-karlsruhe.de>

Hi,

I'm thinking of writing a Model-View-Controller based GUI wrapper around a
couple of the more popular Python GUI toolkits because
	a) I'm tired of apps that work with only a certain GUI, so I want
	   an abstraction layer
	b) I like Model-View-Controller.

Is there already such a project? I don't want to duplicate efforts.

If not, can anyone point me to the Smalltalk MVC specification? I've worked
with Smalltalk a while ago, but don't remember exactly how the classes
looked like. I'd really like to do this the way Smalltalk does it. (If I
can't, I'll use the OPENSTEP AppKit as a model.)

Regards,
		--Bj?rn


From DanielK at aracnet.com  Mon Dec  4 22:29:58 2000
From: DanielK at aracnet.com (Daniel Klein)
Date: Mon, 4 Dec 2000 19:29:58 -0800
Subject: Calculating factorial
References: <3a2a9496.647995@uutiset.nic.fi>
Message-ID: <pdZW5.117$%c4.34041@typhoon.aracnet.com>

Other than the the previous recommendation to use 1L, you could reduce the
number of iterations by 1 by starting at 2 in the 'for' loop, ie

    for number in range(2, f + 1):

No need to multiply 'result' times 1.

Daniel Klein


"Olli Rajala" <ollir at linuxfreak.com> wrote in message
news:3a2a9496.647995 at uutiset.nic.fi...
> Hi!
> I've learned a few time Python and programming.  I wrote a function
> that calculates a factorial of N and I was wondering if it's the best
> (I think it's not. =) way to do that. I'm glad to hear your comments.
>
> <script>
>
> def factorial(f):
>     result = 1
>     for number in range(1, f + 1):
>         result = result * number
>     return result
>
> print factorial(10)
>
> </script>
>
> Ps. Sorry my poor English.
> --
> Olli Rajala
> A Finnish student who loves Linux.




From Blair.Lowe at compeng.net  Tue Dec 12 18:11:22 2000
From: Blair.Lowe at compeng.net (Blair Lowe)
Date: Tue, 12 Dec 2000 16:11:22 -0700
Subject: compiling BeOpen-Python-1-2.0.src.rpm
In-Reply-To: <v04220807b651d381086e@[192.168.11.33]>
References: <9097pc+pspc@eGroups.com>
 <v04220804b64f0e47b847@[192.168.11.33]> <3A29DBF6.6A074D47@tridog.com>
 <v04220807b651d381086e@[192.168.11.33]>
Message-ID: <v04220808b65c5f049943@[192.168.11.33]>

Hi,

We are using Stock RedHat 6.2. We want to re-configure the 
BeOpen-Python source code for SSL, and general use.

We were compiling the BeOpen-Python-2.0-1.src.rpm with the command:

rpm -bi BeOpen-Python-2.0.spec

, and got the following error:


gcc -fpic  -I/usr/local/include/xmlparse -g -O2 -Wall 
-Wstrict-prototypes -I./..
/Include -I.. -DHAVE_CONFIG_H -c ./pyexpat.c
./pyexpat.c: In function `newxmlparseobject':
./pyexpat.c:639: warning: implicit declaration of function `XML_ParserCreateNS'
./pyexpat.c:639: warning: assignment makes pointer from integer without a cast
./pyexpat.c: In function `pyxml_SetStartNamespaceDeclHandler':
./pyexpat.c:1024: `XML_SetNamespaceDeclHandler' undeclared (first use 
in this fu
nction)
./pyexpat.c:1024: (Each undeclared identifier is reported only once
./pyexpat.c:1024: for each function it appears in.)
./pyexpat.c: In function `pyxml_SetEndNamespaceDeclHandler':
./pyexpat.c:1032: `XML_SetNamespaceDeclHandler' undeclared (first use 
in this fu
nction)
./pyexpat.c: In function `pyxml_SetStartCdataSection':
./pyexpat.c:1040: `XML_SetCdataSectionHandler' undeclared (first use 
in this fun
ction)
./pyexpat.c: In function `pyxml_SetEndCdataSection':
./pyexpat.c:1048: `XML_SetCdataSectionHandler' undeclared (first use 
in this fun
ction)
./pyexpat.c: At top level:
./pyexpat.c:1077: `XML_SetCommentHandler' undeclared here (not in a function)
./pyexpat.c:1077: initializer element for `handler_info[8].setter' is 
not consta
nt
./pyexpat.c:1089: `XML_SetDefaultHandlerExpand' undeclared here (not 
in a functi
on)
./pyexpat.c:1089: initializer element for `handler_info[12].setter' 
is not const
ant
./pyexpat.c:1092: `XML_SetNotStandaloneHandler' undeclared here (not 
in a functi
on)
./pyexpat.c:1092: initializer element for `handler_info[13].setter' 
is not const
ant
make[1]: *** [pyexpat.o] Error 1
make[1]: Leaving directory `/home/blade/src/BUILD/Python-2.0/Modules'
make: *** [sharedmods] Error 2
Bad exit status from /var/tmp/rpm-tmp.8833 (%build)



Why doesn't this compile cleanly? Does it need a patch for RH6.2? 
Seems like an unlikely thing to do.

Thanks,
Blair Lowe
Computer Engineering Inc. http://www.compeng.net
Phone: 780 499 5687 (9 - 5 MST) Fax:   780 435 0693 (24 Hours)


>From "Online LoanApplications-MoneyLendersarewaiting...."@cwi.nl  Tue Dec 12 18:17:10 2000
Return-Path: <"Online LoanApplications-MoneyLendersarewaiting...."@cwi.nl>
Delivered-To: python-list at python.org
Received: from hera.cwi.nl (hera.cwi.nl [192.16.191.1])
	by dinsdale.python.org (Postfix) with ESMTP id 5DD091CD72
	for <python-list at python.org>; Tue, 12 Dec 2000 18:17:10 -0500 (EST)
Received: from zeus.cwi.nl (zeus.cwi.nl [192.16.201.120]) by hera.cwi.nl with ESMTP
	id AAA12057 for <python-list at cwi.nl>; Wed, 13 Dec 2000 00:17:11 +0100 (MET)
Received: from Backstretch.leads.com (d83b350d.dsl.flashcom.net [216.59.53.13]) by zeus.cwi.nl with SMTP
	id XAA21233 for <python-list at cwi.nl>; Tue, 12 Dec 2000 23:53:59 +0100 (MET)
Subject: Interest RateOnline Loan Rates lows, tis` the season to spend money la la la .. lalalala -gtqof
MIME-Version: 1.0
X-Encoding: MIME
Content-Type: multipart/alternative;
	boundary="----=_NextPart_53036561383234"
Message-Id: <31n28ei5k2bo52.0u7x5x5bjx1l00f at Backstretch.leads.com>
Date: Tue, 12 Dec 2000 17:54:39 -0600
Reply-To: poogus at hotmail.com
From: Online at python.org, Loan at python.org, Applications at python.org,
	- at python.org, Money at python.org, Lenders at python.org, are at python.org,
	waiting.... at cwi.nl
X-Mailer: shelby090 at rocketmail.com
To: python-list at cwi.nl
Sender: python-list-admin at python.org
Errors-To: python-list-admin at python.org
X-BeenThere: python-list at python.org
X-Mailman-Version: 2.0
Precedence: bulk
List-Help: <mailto:python-list-request at python.org?subject=help>
List-Post: <mailto:python-list at python.org>
List-Subscribe: <http://www.python.org/mailman/listinfo/python-list>,
	<mailto:python-list-request at python.org?subject=subscribe>
List-Id: General discussion list for the Python programming language <python-list.python.org>
List-Unsubscribe: <http://www.python.org/mailman/listinfo/python-list>,
	<mailto:python-list-request at python.org?subject=unsubscribe>
List-Archive: <http://www.python.org/pipermail/python-list/>

This is a multi-part message in MIME format.

------=_NextPart_53036561383234
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

*******************************************************************************
***************
python-list Removal Instructions: Click on the link at the bottom of the page to be 
exclude from further communication.  
******************************************************************************
****************
                  a2z National Mortgage Lender referral specialists 

                                For U.S.A. Homeowners 
                                       Save Now! 

                FREE We Shop The Best Loan For You! "FREE"
 

We are a Mortgage lender referral service who specializes in helping people 
get the loans to: 

Debt Consolidation - http://www.a2z123.tv/debt/
2nd Mortgage - http://www.a2z123.tv/2nd/
Refinance - http://www.a2z123.tv/refin/
Credit Repair - http://www.a2z123.tv/credit/
Home Improvement - http://www.a2z123.tv/improv/
New Car - http://www.a2z123.tv/car/
Dream Vacation - http://www.a2z123.tv/pre/
College Tuition - http://www.a2z123.tv/college/
Start a new business - http://www.a2z123.tv/newbiz/ ..and much, much more. 
Pay off bills - htpp://www.a2z123.tv/gen/ and stop creditors from calling 

100% free service which lets you shop for a mortgage loan conveniently and 
securely from the comfort of your home computer. Using our vast network of 
lenders across the U.S., we will search our database of 50 lender rates for 
the best loan that fit your needs.  Even if you're currently working with 
another lender or have been turned down before, we can still help.  

Funding borrowers with less than perfect credit is we can help that's our 
specialty! 

Incredibly low monthly payments 

We can get you the loan you need.  
Regardless of whether you have good or bad credit, we can help you. 

Ready to get started?  
Simply Click http://a2z123.tv/  fill out the form, and we'll begin shopping 
for your loan.  It's that easy!  

Free Mortgage Quote 
 
-------------------------------------------------------------  
Promotional email service provided by Backstretch Leads?
? 2000 a2z Mortgage Lender Refferal Services 
Email Removal Instructions: Click on the below link to be exclude from 
further communication.
 mailto:d_i_n_o_e at hotmail.com?subject=Remove From Listing.
-------------------------------------------------------------
Are you in need of quality leads for any of the following industries?

MLM Leads
Opportunity Seeker Leads
Business Buyers
Investors
Merchant account leads
Travel Leads
Mortgage Leads
Long Distant Phone Services

If you do... We have them and can always generate them for you.

We offer lead generation services for many companies. If your
industry is not listed above we may still be able to help you with 
your lead service needs.

We can fine tune our mass advertising efforts to fit your exact needs.

For quality leads at an affordable price please contact us today.

Phone: 1.877.346.6123
Email: d_i_n_o_e at hotmail.com
http://a2z123.tv/promo/
-----------------------------------------------------------
Promotional email service provided by Backstretch Leads?
? 2000 a2z Mortgage Lender Refferal Services 
Email Removal Instructions: Click on the below link to be exclude from 
further communication.
 mailto:d_i_n_o_e at hotmail.com?subject=Remove From Listing.
-------------------------------------------------------------


------=_NextPart_53036561383234
Content-Type: text/html;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable


------=_NextPart_53036561383234--




From c.evans at clear.net.nz  Sun Dec 10 01:26:05 2000
From: c.evans at clear.net.nz (Carey Evans)
Date: 10 Dec 2000 19:26:05 +1300
Subject: Handling signals and performance issues.
References: <meRU5.211$Mx.12164@sjc-read.news.verio.net>
	<m9dgmAAe+tM6EwnY@gol.com> <roy-FD4A85.20535009122000@news.panix.com>
Message-ID: <878zpo94hu.fsf@psyche.dnsalias.org>

Roy Smith <roy at panix.com> writes:

> You might want to look at the source to the unix "tail" command (you 
> should be able to find it on any of the many GNU/Linux distribution 
> sites) and see how it implements the "-f" option, which is essentially 
> what you are trying to emulate.

One trick I've seen is do a popen('tail -f filename', 'r').  Because
this gives you one end of a pipe, you can use select() to tell when
more data is available, unlike with a file.

-- 
	 Carey Evans  http://home.clear.net.nz/pages/c.evans/


From claird at starbase.neosoft.com  Wed Dec 27 11:39:10 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 27 Dec 2000 10:39:10 -0600
Subject: Python vs PHP (& Java?)
References: <mailman.977897649.4877.python-list@python.org> <92d22f$cnv$1@nnrp1.deja.com>
Message-ID: <9F7B793FB9C11A70.614813E4DCF61C53.380C715902E65086@lp.airnews.net>

In article <92d22f$cnv$1 at nnrp1.deja.com>,
 <jerry_spicklemire at my-deja.com> wrote:
			.
			.
			.
> Python packages for embedding scripts in HTML -
<URL:http://starbase.neosoft.com/~claird/comp.lang.python/web_python.html>
aims to be comprehensive on this subject.
			.
		[various apt
		observations]
			.
			.
>Yes, you can do it, but if you're working server side anyway, why
>choose a language as limiting as Javascript?
To lock yourself into an unhealthily conventional 
vendor relationship with Netscape.  Phrased some-
what more charitably, that seemed like a bang-up
idea to many well-paid IT professionals not so
long ago.
			.
			.
			.
>Still, there are many very easy to use Python modules for such
>functions, and if you are truly scripting InterNet Communications,
>you're dealing with standard protocols. It doesn't matter what language
>either end is implemented in. Java based clients can talk to Python
>based servers, and vice versa.
Like many things Mr. Spicklemire has written, this
paragraph bears reiteration.
			.
			.
			.
-- 

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


From bryan at eevolved.com  Wed Dec 13 01:46:31 2000
From: bryan at eevolved.com (Bryan Mongeau)
Date: Wed, 13 Dec 2000 06:46:31 GMT
Subject: __file__ in 2.0
Message-ID: <bXEZ5.2409$0U2.143972@news20.bellglobal.com>

Hi all,

I am interested in the mcmillan installer to generate self-extracting exe's.
Unfortunately, running source/Make.py results in an AttributeError on 
exceptions.__file__ calls.

I understand that static modules (such is the case) do not define the 
__file__ attribute. Documention pointed me to use .repr() or .str() instead 
to get the filename of a loaded module. No such luck when I edit the script 
(Attribute errors again)...

Which leads me to my next question: is the mcmillan installer project 
stagnant? The release contains __file__ calls which crap out in 2.0 and 
they haven't even mentioned this on the site at all... Kinda leaves newbies 
like me scratching our heads.

And my final question: How do I call __file__ on static modules in 2.0??

Tips, advice welcome. 

-- 
<==================================>
Bryan Mongeau
Lead Developer, Director
eEvolved Real-Time Technologies Inc.
www.eevolved.com
<==================================>

"... the true utility function of life, that which is being maximized in 
the natural world, is DNA survival. But DNA is not floating free; it is 
locked up in living bodies and it has make the most of the levers of power 
at its disposal." -- Richard Dawkins



From jal at faxnet.com  Fri Dec 15 09:55:32 2000
From: jal at faxnet.com (Jonathan LaCour)
Date: 15 Dec 2000 14:55:32 GMT
Subject: HELP - Threading with SCO Unix
Message-ID: <91dbd4$9ot@dispatch.concentric.net>

I am working with a corporation that is beginning development on a Python
based product using Zope.  In order for our product to function, we need
to have Python built with thread support.  One platform that we must
support is SCO OpenServer.  We have been attempting to compile
python-1.5.2 with thread support for several days without success.  We
*must* have thread support, or we will be forced to move to another
solution.

Does Python even support threads in SCO Unix?

Any suggestions?  We have attempted our build using a binary distribution
of GNU PThreads for SCO OpenServer.  Installing another UNIX is not an
option at this stage.

If you are running SCO OpenServer and have python with threading support,
please contact me!  We need your help!

Thanks for you help!  This is absolutely critical for success, and any
assistence would be appreciated!

Jonathan LaCour
Developer
VertiSoft, Corp.


From pbjorn at uswest.net  Fri Dec 29 23:52:21 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Fri, 29 Dec 2000 21:52:21 -0700
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com>
Message-ID: <3A4D6A05.EEC97FA6@uswest.net>

henry_crun3583 at my-deja.com wrote:

> Can someone point me at an explanation of the logic for not having
> do:
> until
>
> or explain the elegant way to achieve this in Python.
>
> eg
> do :
>     R=Themistor.ReadResistance()
>     T=R_to_Temperature(R)
>   until (T >=25)

while 1:
   R=Themistor.ReadResistance()
   T=R_to_Temperature(R)
   if T >=25: break




From tcmay at got.net  Tue Dec  5 21:36:10 2000
From: tcmay at got.net (Tim May)
Date: Tue, 05 Dec 2000 18:36:10 -0800
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com>
Message-ID: <tcmay-A67EAA.18361005122000@news.got.net>

In article <90k3pq$50p$1 at nnrp1.deja.com>, Nigel <nigeljennison at cs.com> 
wrote:

>  I am doing some research and I would be grateful if anybody could give
> me some history on the origins of Python.
> 

If you are doing research, then you should surely understand that the 
large amount of material already on the Web, and findable with standard 
search engines, is a much better source of information than asking 
people to custom-write material for you. Or to do your searches for you. 

A handful of minutes spent with Google will yield you far more than 
waiting and waiting for people here to write histories for you.

Why are you "doing some research" if you are this clueless?


--Tim May

-- 
---------:---------:---------:---------:---------:---------:---------:----
Timothy C. May              | Crypto Anarchy: encryption, digital money,  
ComSec 3DES:   831-728-0152 | anonymous networks, digital pseudonyms, zero
W.A.S.T.E.: Corralitos, CA  | knowledge, reputations, information markets, 
"Cyphernomicon"             | black markets, collapse of governments.




From aleaxit at yahoo.com  Wed Dec 13 17:19:14 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 13 Dec 2000 23:19:14 +0100
Subject: Moving from Perl to Python
References: <3a3a9552.76673185@news.tcn.net> <slrn93fcdt.a8m.scarblac@pino.selwerd.nl> <3a38b798.85448901@news.tcn.net>
Message-ID: <918sud02por@news1.newsguy.com>

"Richard Zilavec" <rzilavec at tcn.net> wrote in message
news:3a38b798.85448901 at news.tcn.net...
    [snip]
> >I started with Learning Python, and I like that book. It explains the
>
> Wish I could see this book first, I don't want anything too basic
> maybe just a quick start, then more advance features.

In this case you may like the "Quick Python Book"; it's faster-paced
and goes farther, while "Learning Python" covers less ground, a bit
more deeply, and more slowly.


Alex





From sh at ttsoftware.co.uk  Fri Dec 22 06:21:30 2000
From: sh at ttsoftware.co.uk (Steve Horne)
Date: Fri, 22 Dec 2000 11:21:30 +0000
Subject: Speed of Python 1.5.x vs 2.0
References: <4vj54t4v0hvjsnstfmtmf36q9irotp59sj@4ax.com> <mailman.977475665.3482.python-list@python.org>
Message-ID: <dud64ts8mb7mlc41kht410q5g87ekb2nc5@4ax.com>

On Fri, 22 Dec 2000 09:59:58 +0100, Andreas Jung
<andreas at andreas-jung.com> wrote:

>On Thu, Dec 21, 2000 at 10:53:32PM -0500, Tyler Eaves wrote:
>> 	I'm soon going to attemp to talk my webhost into upgrading to
>> 2.0, but I figure the only way I'm going to be able to make that
>> happen is if I can show results that show 2.0 being faster / more
>> effcient than 1.5.x. Their servers run on Red Hat 7.
>
>2.0 runs a bit faster than 1.5.X but don't expect a performance
>boost. You should upgrade when you need some of the new features
>of Python 2.0 like unicode support... 
>
>Andreas 

I'd expect list comprehensions to work significantly faster than
equivalent expressions using filter, map and lambdas - the reason
being that the overhead of repeatedly calling the lambda should be
avoided, not to mention avoiding the temp storage between the filter
and map.

I haven't tried this, but it might be worth a go.

-- 
Steve Horne
Home : steve at lurking.demon.co.uk
Work : sh at ttsoftware.co.uk


From removethis_kranki2 at earthlink.net  Fri Dec 22 10:38:32 2000
From: removethis_kranki2 at earthlink.net (Bob W.)
Date: Fri, 22 Dec 2000 15:38:32 GMT
Subject: Newbie Question
References: <05m64tkqev0aavea80ttjse1sj2gd447hg@4ax.com>
Message-ID: <hqs64tc5jjndav3p6nvmu67t7hqaf5go00@4ax.com>

I tried using "self" in the os.path.walk and I get an error saying
that there 3 expected arguments, but got 4 in ntpath.py, line 264.
That was my preferred way of simulating a C++ private method.
Thanks for any help provided!

On Fri, 22 Dec 2000 13:44:23 GMT, Bob W.
<removethis_kranki2 at earthlink.net> wrote:

>I am trying to write the following class, but it fails in the Run()
>method reference to __CheckFile and I am not sure how to get it to
>work.  Thanks in advance for your help!
>
>class RmwExpireCheck:
>    def __init__( self, szFileName, szDir ):
>        self.szFileName = szFileName
>        FileStatList = os.stat( szFileName )
>        if FileStatList:
>            self.iCheckDate = FileStatList[8]
>        self.szDir = szDir
>        self.szFileList = []
>
>    def __getitem__(self, i):
>        line = self.szFileList[i]
>        if line:
>            return line
>        else:
>            raise IndexError
>
>    def __CheckFile( self, szDirName, szFileNames ):
>        for szFileName in szFileNames:
>            szPathName = os.path.join( szDirName, szFileName )
>            if os.path.isfile( szPathName ):
>                iFileTime = os.stat( szPathName )[8]
>                if iFileTime > self.iCheckDate:
>                    self.szFileList.append( szPathName )
>        
>    def Run( self ):
>        os.path.walk( self.szDir, RmwExpireCheck.__CheckFile, self )
>        
>



From SBrunning at trisystems.co.uk  Wed Dec  6 05:13:13 2000
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Wed, 6 Dec 2000 10:13:13 -0000 
Subject: [DAO Database in win32] What is more efficient
Message-ID: <31575A892FF6D1118F5800600846864D5B1552@intrepid>

> From:	Arturo P?rez Mulas [SMTP:arturo_perez at wanadoo.es]
> 
> By the way, is mxODBC any better?
 
I dunno about better, 'cos I've never tried DAO/ADO. But I've used mxODBC
quite a lot recently, and it seems fast, powerful and stable to me. I
recommend it.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From aleaxit at yahoo.com  Tue Dec  5 05:11:13 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 5 Dec 2000 11:11:13 +0100
Subject: "inherited" keyword in Python?
References: <3A2B9EFA.3A64C13B@schlund.de> <90gevs02uu1@news2.newsguy.com> <m2elznipbl.fsf@mycroft.actrix.gen.nz>
Message-ID: <90ievt025bl@news2.newsguy.com>

"Paul Foley" <see at below> wrote in message
news:m2elznipbl.fsf at mycroft.actrix.gen.nz...
    [snip]
> > Metaprogramming magick was deemed inappropriate here (as I
> > read the group consensus).  One *might* have a global function
> > called super(), which tried to gauge the appropriate class
> > object to use (in single-inheritance cases, only, of course),
>
> Why "of course"?  If you can find out the name of the method which is
> calling super() (which can be done, but I don't remember how), and the

Name of calling-function is easy (raise & catch an exception,
it's in the stackframe you get this way).

> class from which it's being called (I can think of ways to avoid
> needing this, if it can't be done),

This one is hellish.  I can't figure out ANY halfway clean way
that doesn't require either mucking with the bytecode, or going
through all the dictionaries in an object's baseclasses until
an entry with the same codeobject as the one previously identified
is found -- slow as well as dirty.

> it shouldn't be very hard to make
> a super() function that works with multiple inheritance.  Use
> cpl(class) [see below] to get a properly ordered list of base classes,
> and search for the first one that implements the named method.  Call
> that.  [You need to know which class the current method belongs to in
> order to get the right cpl, otherwise it'll infloop...but that isn't
> obtainable, you could maybe have super() maintain a stack to tell it
> where it's at, or something...]

I think your cpl function is much more complicated than what's needed
to reproduce Python's very simple rule for walking the DAG of base
classes (it's just depth-first, left-first), even assuming it does
end up reproducing it perfectly.  But that's not the main point.

A more interesting theoretical question is how you propose to solve
the case in which you can't tell which class the current method
'belongs' to (also remember a given function-object may be set as
a method of N classes, for any N, with any set of names -0- but,
even forgetting that for the moment).  E.g., a very simple case:

class A:
    def foo(self):
        print super()

class B(A):
    def foo(self):
        print super()
        A.foo(self)

class C(A):
    def foo(self):
        print super()
        A.foo(self)

class D(B,C):
    def foo(self):
        print super()
        B.foo(self)
        C.foo(self)

Here, the various 'foo's are mercifully separate codeobjects so
IN THEORY this is solvable (e.g., by rooting in the dictionaries
of each ancestor-class of self looking for the codeobject as a
value) -- of course if, say, C just did

class C(A):
    foo = B.foo

or

def bar(something):
    print super()
    A.foo(something)
class C(A):
    foo = bar

we might have a harder time.  But unless the class 'owning'
(hoping there is some 'ownership'...) the calling-method IS
identified (be it by hook or by crook), how does 'super()'
does its job, even theoretically...?


On a *pragmatic* level, I think this kind of "deep and dark
metaprogramming tricks" are most often best eschewed for
practical use.  Great for writing debuggers, class-browsers,
and all sorts of infrastructure, sure, but simpler and cleaner
ways will generally prove more understandable and thus more
maintainable.  "Do the simplest thing that could possibly
work" rules... and the "simplest thing" _rarely_ (not 'never')
involves such reflection/introspection.  It's too easy to get
lost on a misguided quest to provide 'convenience', and
actually end up obfuscating and obscuring things for no
real good purpose.

'super()' in a multiple-inheritance context is a good example.
There is no single superclass that will do for all attributes
one might wish to access, so
    super().foo()
and
    super().bar()
just won't work.  'supercall' might work (without needing
any real complication -- just a loop of getattr over the
__bases__, given Python's quite-simple semantics here), IF
you tell it which class it must find the 'supers' of -- the
simplest and most effective way to do that is to have a
convention that each class starts out by setting some
specific variable to itself:

class Foo:
    __thisclass=Foo
    # etc

class Bar(Foo):
    __thisclass=Bar
    # etc

and 'supercall' is always (but always!-) called in the form

    supercall(self, __thisclass, methodname, *args, **kwds)

which reduces it to something decently-simple, like:

def supercall(_obj, _cls, _metnam, *args, **kwds):
    for _bas in _cls.__bases__:
        _metobj = getattr(_bas, _metnam, None)
        if not _metobj is None:
            return _metobj(_obj, *args, **kwds)
    else:
        raise NameError, "No superclass of %s has method %s" % (
            _cls, _metnam)


But, how often is it going to be satisfactory to call 'just
the one' (of several) inherited-method of a certain name, in
a multiple-inheritance context?  Just as often we may want
to call _all_ methods of a certain name (__init__ and __del__
are frequent candidates, and indeed the last time this was
discussed, a couple weeks ago I think, __init__ was the
focus of the discussion) -- and shall we want to pass each
of them the same arguments...?  Complications flood us
again pretty soon, and the harked-for "convenience" remains
elusive.

To quote an old local saying, I think that, in most cases,
going for sophisticated metaprogramming tricks in the
attempt to provide "convenience" is "quenching your thirst
with ham".  (P.S.: Bolognese ham tends to be pretty salty:-).


Alex





From dale at out-think.NOSPAMco.uk  Wed Dec  6 08:18:36 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Wed, 06 Dec 2000 13:18:36 +0000
Subject: Rummaging around inside an Access DB
Message-ID: <c5es2ts2l66bhg8qb7ne28cj0qa3gi9595@4ax.com>

I've been trying to dig some information out of an Access DB using Python. I need to access as much
information as possible about the composition of all the reports. 

After a few false starts and lots of head scratching (I can't find a help file on the object model
anywhere!!), I found something in the Python Win32 distribution called daodump.py, presumably
written by Mark H.

This helped a lot and produced a listing of all the reports but I can't seem to break into the
Report object.

This is what I've come up with so far:


from win32com.client import Dispatch
import sys
eng = Dispatch("DAO.DBEngine.36")
db = eng.OpenDatabase(sys.argv[1])
rep = db.Containers.Item('Reports')
for doc in rep.Documents:
	print doc.Name


Does anyone know how I can get at the cuts of the Report object? The only other properties I can
find are: 

AllPermissions, CLSID, Container, CreateProperty, DateCreated, LastUpdated, Name, Owner,
Permissions, Properties, UserName


Thanks for any help.
--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants




From francois.granger at free.fr  Thu Dec  7 15:45:05 2000
From: francois.granger at free.fr (=?ISO-8859-1?Q?Fran=E7ois_Granger?=)
Date: Thu, 07 Dec 2000 20:45:05 GMT
Subject: Help/example of bit testing
References: <90llq7$8fc@dispatch.concentric.net>
Message-ID: <1elagig.djja7k1ujl97mN@paris11-nas7-52-81.dial.proxad.net>

Bryan Webb <bww00 at amdahl.com> wrote:

> Hi,
>     I am reading a file (binary)of 1024 char recs (bitmaps) . I can read the
> file , bit im having trouble finding examples of testing bits. I need to
> look at each bit of the above record.
> Looking for examples and ideas.

I got inspired by an algorithm by Tim Peters, so I did a quick hack in
order to display in binary form. It need some polishing.

l = [1, 2, 3, 4, 8, 16, 32, 64, 128, 255, 512, 1024, 2048, 4096, 8192,
16384, 32768]

def toBin(x):
    a = ''
    y = x
    while x:
        x, y = x / 2, y % 2
        if y:
            a = '1' + a
        else:
            a = '0' + a
        y = x
    return
for x in l:
    print toBin(x)


-- 
Une faq de frj: http://faq.jardin.free.fr/, pr?s de l'encyclop?die
http://nature.jardin.free.fr/


From sdm7g at virginia.edu  Wed Dec 20 13:09:45 2000
From: sdm7g at virginia.edu (Steven D. Majewski)
Date: Wed, 20 Dec 2000 13:09:45 -0500 (EST)
Subject: Python interpreter in python?
In-Reply-To: <91qas9$i10$1@talia.mad.ttd.net>
Message-ID: <Pine.NXT.4.21.0012201243290.224-100000@localhost.virginia.edu>


On Wed, 20 Dec 2000, Raymond Biurrun wrote:

> "Steven D. Majewski" <sdm7g at virginia.edu> escribi en el mensaje
> news:mailman.977254923.336.python-list at python.org...
> >
> > That is absolutely untrue!
> >
> > Lisp and Scheme are frequently implemented in themselves -- a
> > 'metacircular interpreter' is a fairly typical student exercise.
> >
> 
> Please excuse my ignorance but what's a "metacircular interpreter" and where
> can I find info about them.
> 

Thats just the fancy name they sometimes  use for a Lisp interpreter
written in Lisp. The top level loops on (PRINT (EVAL (READ))) -- 
and usually the exercise is to implement EVAL. 

It's a not uncommon exercise in Lisp or Scheme books.
( See <http://www.lisp.org/table/books.htm> for a list. ) 
The Structure and Interpretation of Computer Programs by Abelson,
Susman & Sussman (commmonly abreviated SICP) is one. 
Lisp In Small Pieces by Christian Queinnec, or John Allen's 
  (out of print) Anatomy of Lisp are a couple other I recall. 

Any of those or any of the other books on implementing Lisp are 
useful for understanding the implementation of any other symbolic
interpreted language (like Python, for example.) 
 Parsing is completely different -- Python has a more complicated
syntax -- but you can find info on parsing in any compiler book. 
 Python's solutions for symbol management and scoping are different,
but the problems are similar.

Lisp and Forth just happen to have a significant implementation
literature. Ralph Griswold has a good book on the implementation
of Icon, and I think had one on Snobol. For everything else --
Python, Perl, Tcl, Ruby, et.al., you mostly have just the source
code. 

-- Steve Majewski <sdm7g at Virginia.EDU> 





From glenfant at equod.com.nospam  Tue Dec 19 08:29:25 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Tue, 19 Dec 2000 14:29:25 +0100
Subject: Timers in Python?
References: <3a3e42ef.16144323@news.skynet.be> <91li8l$33h$1@reader1.imaginet.fr> <3a3e61cf.24049354@news.skynet.be> <91lq8r$gu2$1@panix6.panix.com>
Message-ID: <91nnj3$djm$1@reader1.imaginet.fr>

Yes...

Sorry for this !

"Aahz Maruch" <aahz at panix.com> a ?crit dans le message news:
91lq8r$gu2$1 at panix6.panix.com...
> In article <3a3e61cf.24049354 at news.skynet.be>,
> Daniel <Daniel.Kinnaer at AdValvas.be> wrote:
> >
> >class TimeScheduled(threading.Thread):
> >    def __init__(self, function, delay):
> >        self.function = function
> >        self.delay = delay
>
> Gilles made a boo-boo.  Here's the correct code:
>
> class TimeScheduled(threading.Thread):
>     def __init__(self, function, delay):
>         threading.Thread.__init__(self)
>         self.function = function
>         self.delay = delay
>
> For more info on threading, take a look at
> http://starship.python.net/crew/aahz/
> --
>                       --- Aahz (Copyright 2000 by aahz at pobox.com)
>
> Androgynous poly kinky vanilla queer het    <*>
http://www.rahul.net/aahz/
> Hugs and backrubs -- I break Rule 6
>
> This is a signature anti-virus.
> Please stop the spread of signature viruses!



From jepler at inetnebr.com  Sat Dec 30 13:43:04 2000
From: jepler at inetnebr.com (Jeff Epler)
Date: Sat, 30 Dec 2000 18:43:04 GMT
Subject: 2001 Enchancement Wishlist
References: <3A4CEACB.220998A0@javanet.com> <Pine.LNX.4.10.10012301008410.5079-100000@kev.ioc.ee>
Message-ID: <bcb.3a4e2cb8.85fd7@potty.housenet>

On Sat, 30 Dec 2000 10:52:10 +0200, Pearu Peterson
>I see that many Python experts are against this proposal but anyway, I
>would like to support this idea with the following examples.
>
>Say, I would like to define
>
>class Real:
>    def __init__(self,num):
>        self.num = num
>    <other number class methods>
>
>class Complex:
>    def __init__(self,real,imag):
>        if imag == 0:
>            return Real(real)
>        self.real,self.imag = real,imag
>    <other number class methods> 
>
>That is, if imaginary part of a Complex number is zero, it's constructor
>would return Real instance:

I think that this can be accomplished by using a 'transmute' function, defined
something like the following:
	def transmute(instance, newclass, *args, **kw):
		instance.__class__ = newclass
		apply(instance.__init__, args, kw)
		return instance

it can be useful other places than __init__, for one thing.  See
Complex.__iadd__.

class Real:
	def __init__(self, num):
		self.num = num
	# numeric methods

class Complex:
	def __init__(self, real, imag=0):
		if not imag:
			transmute(self, Real, real)
			return
		self.real, self.imag = real, imag

	def __iadd__(self, other):
		self.real += other.real
		self.imag += other.imag
		if not self.imag:
			transmute(self, Real, self.real)
		return self
	# Other numeric methods

>>> complex(1,0)
<Real 1>
>>> c = complex(1,1)
>>> print c
<Complex 1+1i>
>>> c += complex(1,-1)
<Real 2>

Is there anything that you would like to do which transmute cannot?  One issue
I see is whether the dictionary should be cleared in the object before
it is transmuted (for instance, "c" still has attributes 'real' and
'imaginary' after it is transmuted in Complex.__iadd__) but I don't see
anything that is just plain impossible.

Toss transmute in your standard library somewhere, and you have a nice way to
write this code---just a function call.  Or make it a method on a common base
class.  I don't know whether "self.transmute(Real, self.real)" makes any more
or less sense than I wrote it above.

Jeff


From tony at lsl.co.uk  Thu Dec 14 09:34:10 2000
From: tony at lsl.co.uk (Tony J Ibbs (Tibs))
Date: Thu, 14 Dec 2000 14:34:10 -0000
Subject: How to use mxTextTools
In-Reply-To: <2bc4OuYTMVp87e49nlcz01wv0ybq@4ax.com>
Message-ID: <003501c065da$ec392da0$f05aa8c0@lslp7o.int.lsl.co.uk>

Paul Moore wrote:
> More information, or better still, some complete examples, would be
> very helpful. (All the examples in the distribution just use
> print_tags() to display the tags, and don't do anything with them...)

Well...

the "obvious" (and incestuous?) examples, referenced off the mxTextTools
web page itself, are my "meta language" for mxText Tools [1] and Mike
Fletcher's SimpleParse [2]. They're rather different in style, and
(perhaps) not a *simple* example of what one can do, but they *are*
examples.

A "quick" search with Google also suggests that PyFontify.py in
py2html.py might be using it, and maybe dalke is as well.

Of course, Marc might have some examples as well...


[1] http://www.tibsnjoan.demon.co.uk/mxtext/Metalang.html
[2]
http://members.home.com/mcfletch/programming/simpleparse/simpleparse.htm
l

Have to dash...
	Tibs

--
Tony J Ibbs (Tibs)      http://www.tibsnjoan.co.uk/
Well we're safe now....thank God we're in a bowling alley.
- Big Bob (J.T. Walsh) in "Pleasantville"
My views! Mine! Mine! (Unless Laser-Scan ask nicely to borrow them.)




From thomas at xs4all.net  Sat Dec 23 15:53:25 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Sat, 23 Dec 2000 21:53:25 +0100
Subject: Why Won't This Work?
In-Reply-To: <LOBBJMHOIFOLFIPFANHJIEGACFAA.lists@TheWebsons.com>; from lists@TheWebsons.com on Sat, Dec 23, 2000 at 12:18:40PM -0800
References: <20001223184357.G5515@xs4all.nl> <LOBBJMHOIFOLFIPFANHJIEGACFAA.lists@TheWebsons.com>
Message-ID: <20001223215325.L5483@xs4all.nl>

On Sat, Dec 23, 2000 at 12:18:40PM -0800, Ben Ocean wrote:

> Hmmm. *python* at a command prompt, sans *./* is a command.
> This worked when I configured 2.0, which I had to rip out because it's not
> compatible with Zope.

'python' *with* './' is a command as well. './python' attempts to start the
python binary in your current directory. If that fails, you don't have a
python binary in your current directory, or it isn't executable :)

The problem with 'python' seems to be that there is no 'python' binary in
your PATH (the standard search-path for executables) according to your
shell. If you really ripped out Python 2.0, and did 'make install' in a
1.5.2 tree, it could very well be that your shell hasn't registered the fact
that you have a new binary. Try doing 'rehash' or 'hash -r' (depending on
your shell) to 'reset' it.

Another possibility is that you don't have /usr/local/bin/ in your path, or
you passed a different --prefix to python's configure and don't have that in
your path. All the above can only be the problem if you didn't already do
'make install' in the Python 1.5.2 sourcetree. From your mail it isn't clear
whether you did that yet or not.

> Also, I tried *make test*
> [webmaster at thor Zope-2.2.4-src]$ make test
> make: *** No rule to make target `test'.  Stop.

> Any other ideas as to why I'm not getting my installation of Python 1.5.2
> to respond?

Well, is your problem python 1.5.2, or Zope ? If you want to install python
1.5.2, you shouldn't first try to install Zope ;) The 'make test' I was
referring to is the python 'make test', so you should do it in the python
1.5.2 sourcetree, not the zope one.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From root at rainerdeyke.com  Sat Dec 30 23:42:23 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Sun, 31 Dec 2000 04:42:23 GMT
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <slrn94shbe.150.grey@teleute.rpglink.com> <X8s36.2731$Qb7.305339@newsb.telia.net> <slrn94smds.150.grey@teleute.rpglink.com> <FOt36.17$vd1.2224@e420r-atl2.usenetserver.com> <slrn94t1vj.150.grey@teleute.rpglink.com>
Message-ID: <POy36.102241$x6.45606433@news2.rdc2.tx.home.com>

"Steve Lamb" <grey at despair.rpglink.com> wrote in message
news:slrn94t1vj.150.grey at teleute.rpglink.com...
> On Sat, 30 Dec 2000 23:00:53 GMT, Kragen Sitaker <kragen at dnaco.net> wrote:
> >Hahaha!  You just plonked Fredrik Lundh in comp.lang.python.  He was
> >being helpful and polite, and you plonked him because you disagreed
> >with him.  I can guess how much help you'll get next time you have a
> >Python question.  :)
>
>     I can tell you I'll get far better without him than with him when he
is
> giving advice that runs contrary to about 10+ years of programming that
I've
> been taught from various different sources.  I don't care if he is God
> himself, advocating not initing a variable is stupid, plain and simple.
He
> wanted to argue the point for nothing more than argument's sake, so be it.
I
> don't need his advice any further.

Uninitialized variables do not exist in Python, therefore the concept of
initializing variables is meaningless here.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From auto103690 at hushmail.com  Sun Dec 10 23:29:03 2000
From: auto103690 at hushmail.com (Jeff)
Date: Mon, 11 Dec 2000 04:29:03 GMT
Subject: Executable python program?
Message-ID: <jKYY5.54278$df5.1795043@news1.crdva1.bc.home.com>

A non-text attachment was scrubbed...
Name: not available
Type: text/plain  charset=iso-8859-1
Size: 534 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20001211/be66d09c/attachment.bin>

From dale at out-think.NOSPAMco.uk  Wed Dec 13 10:37:33 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Wed, 13 Dec 2000 15:37:33 +0000
Subject: Source Code Management
Message-ID: <fb4f3tsgffnmvme64aol1cn397te7ln349@4ax.com>

I'm looking for a source code management system for looking after Python, Java, Javascript and
possibly other languages too.

Does anyone know of such a system that matches or comes close?

Essential features:
* Runs on Windows 2000
* Works over the Internet (ftp would be good)
* Secure (ftp level security would pobably suffice)
* Module level version control with deltas or compressed archives
* Package level (multiple modules) sign in/sign out
* System level release control. (ie, this is now release x, note all module versions for this
release.)

Desirable features:
* Extendible - preferably scriptable with Python
* Manage arbitrary files including documentation
* GUI or command line interfaces
* Low maintenance
* Library stored as plain files and not in some odd database format

That's not too much to ask, is it!  :-)

Thanks
--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants


From Blair.Lowe at compeng.net  Wed Dec  6 18:05:03 2000
From: Blair.Lowe at compeng.net (Blair Lowe)
Date: Wed, 6 Dec 2000 16:05:03 -0700
Subject: XML-RPC was [pygtk] Python and Orbit Examples
In-Reply-To: <20001206165355.B12869@linfoo.eink.com>
References: <Pine.LNX.4.21.0012050828550.24722-100000@tupelo.daa.com.au>
 <200012060632.eB66WIK27614@ZOOstation.cc>
 <20001206165355.B12869@linfoo.eink.com>
Message-ID: <v04220802b654744d4fc6@[192.168.11.33]>

Sounds cool.

I found pyXML. at http://download.sourceforge.net/pyxml/.

Can one send XML to an SSL site with pyXML as well? If so, where is 
some good docs for this. The version I got (PyXML-0.6.2-2.0.i386.rpm) 
has not docs, and the overview at 4suite say I need to download 
something else.

Are there some better tools for doing this. I started to look at 
Zope, but it was rather cumbersome and knowing where to start with 
the documentation was not easy to find either.

TTYL,
Blair.

At 16:53 -0500 2000/12/06, Pehr Anderson wrote:
>Dear Folks,
>
>I would highly recommend using XML-RPC as a minimally complicated
>way to communicate data between your apps.
>
>This might sound complicated but it really only means
>	1. specify XML encodings for your data, use these throught your code
>	2. open a socket on the target
>	3. send an XML query and wait for an XML response
>	4. parse the response
>
>The beautiful thing about XML-RPC (remote proceedure call) is
>that it is just dirt simple. You open a socket, dump in
>human readable text, and wait for a human-readable response.
>Everything gets completely defined somewhere and servers and
>clients could be coded in *anything*!
>No complicated tools are required to implement any parts of
>this system though you might want to look at how others
>parse XML rather than doing it all by hand.
>
>And the best part is you'll have an app that anyone can understand.
>No secret church of CORBA, COM or other "special tools".
>XML is directly comprehensible by HUMAN BEINGS and that makes it
>a sustainable encoding with simple tcp sockets for transport.
>
>	-pehr
>
>
>On Tue, Dec 05, 2000 at 06:32:17PM -1200, Rob Brown-Bayliss wrote:
>>  >or getting it to work with gnome-python, that is a bit more difficult.
>>
>>  Sounds a little ominous...
>>
>>  What I am doing (or planning at this stage) is to write a POS and Stock
>>  management system for a small retail chain.
>>
>>  Would I be better of giving Pyhton a miss, or ditching the Corba idea?
>>
>>  I have very little knowledge of linux programming, though I do have
>>  years of basic programming on the Amiga (like 18 I think if 
>>include vic 20 and c64)
>>  and a few months VC++ on windows.  I have used Linux for 18 months 
>>now, and played
>>  a bit with C and a very very little with Python.
>>
>>  I am confident I can tackle the task in either language, but I 
>>know python will be
>>  quicker to develop with.
>>
>>  I had looked to corba as a means of communicating between the 
>>stores, I could always
>>  use some form of file transefer, but it's not as elegant or nearly 
>>as real time.
>>
>>
>>
>>
>>  --
>>
>>    Rob Brown-Bayliss
>>   ---======o======---
>>
>>    www.ZOOstation.cc
>>
>>  _______________________________________________
>>  pygtk mailing list   pygtk at daa.com.au
>>  http://www.daa.com.au/mailman/listinfo/pygtk
>
>--
>http://www.python.org/mailman/listinfo/python-list

Computer Engineering Inc. http://www.compeng.net
Phone: 780 499 5687 (9 - 5 MST) Fax:   780 435 0693 (24 Hours)



From bruce_dodson at bigfoot.com  Fri Dec 22 23:29:08 2000
From: bruce_dodson at bigfoot.com (Bruce Dodson)
Date: Sat, 23 Dec 2000 04:29:08 GMT
Subject: coding convention conversion
References: <20001222122015.A7987@llnl.gov> <mailman.977522226.5104.python-list@python.org>
Message-ID: <oSV06.176$g03.67749@sapphire.mtt.net>

Better yet, import the file as a module, and scan through its public symbols
(via dir()) to find out which names need to be changed, and then use that
information to build a sed script which can be applied globally.  You can
easily take advantage of other meta-information such as the type, e.g. limit
it to just functions if you want.  This approach is good because the parser
will recognize more variations than a regexp.


Something like this is what I had in mind:

from types import ModuleType, ClassType

def print_symbols(symbols, prefix=''):
    for sym in dir(symbols):
        if sym[:2] == '__': continue

        v = getattr(symbols, sym)
        if type(v) is ModuleType: continue

        print prefix + sym

        if type(v) is ClassType:
            print_symbols(v, prefix + sym + '.')


See also the pyclbr module for a way to do something similar without
executing the module.

Bruce



"D-Man" <dsh8290 at rit.edu> wrote in message
news:mailman.977522226.5104.python-list at python.org...
>
> This shouldn't be too hard with a python script and some regexes with
> some more string processing in between (or maybe multiple pass).
>
> Everything you want to change is bounded by "def" on one side and "("
> on the other.  Then replace the capitals with _ followed by the lower
> case version.
>
> A quick (partial) solution:
>
> re = re.compile( "def ([a-z]+)([A-Z])([a-z]*)\(" )
>
> for line in source.readlines() :
> match = re.match( line )
> if ( match ) :
> res = "def " + re.group( 0 ) + "_" + \
> string.lower( re.group( 1 ) + re.group( 2 ) + "("
>
>
> This isn't complete and won't have the arguments in the result (nor
> preserve indentation before the "def").
>
> I think the re.sub() function will do more of what you want it to.
>
>
> HTH,
> -D
>
>
> On Fri, Dec 22, 2000 at 12:20:15PM -0800, Issac Trotts wrote:
> > Does anyone know of a utility to change between the coding conventions
> > of
> >
> > def fooBarBaz(): pass
> >
> > and
> >
> > def foo_bar_baz(): pass
> >
> > It is a pain to do this by hand.
> >
> > Thanks,
> > Issac
>



From matthewrussell at my-deja.com  Wed Dec  6 07:49:59 2000
From: matthewrussell at my-deja.com (matthewrussell at my-deja.com)
Date: Wed, 06 Dec 2000 12:49:59 GMT
Subject: urllib2, proxy, basic authentication
References: <90hk8k$fqh$1@bob.news.rcn.net> <3A2D61C1.9182431D@motorola.com> <90k859$adl$1@bob.news.rcn.net>
Message-ID: <90lclk$38p$1@nnrp1.deja.com>

A way to pass authentication details would be to specify them in the
url as follows

import urllib2

urllib2.urlopen("http://username:password at www.mydomain.com")

In article <90k859$adl$1 at bob.news.rcn.net>,
  Albert Chin-A-Young <china at the.web.com> wrote:
> Jeremy Lamplough <jeremy.lamplough at motorola.com> wrote:
> > try setting the proxy host prior to importing the other modules.
>
> > import os, sys
> > os.environ['http_proxy'] = 'http://[PROXY_HOST]:5865'
> > import urllib2
> > from urllib2 import urlopen
>
> > The way i'm doing basic stuff, with NO authentication is (to
download a
> > stock quote)
>
> We don't have a problem with http_proxy and the remote host not
> requiring authentication. However, if the remote host does require
> authentication, then we have our problem.
>
> --
> albert chin (china @at@ thewrittenword .dot. com)
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From rjroy at NO-SPAM.magma.ca  Wed Dec 20 14:34:30 2000
From: rjroy at NO-SPAM.magma.ca (Robert Roy)
Date: Wed, 20 Dec 2000 19:34:30 GMT
Subject: Q: dynamic function call
References: <Ko606.141$G66.2664@vixen.cso.uiuc.edu>
Message-ID: <e1124tgv39ria0drqf3c76c5sr2g2ekg26@4ax.com>

On Wed, 20 Dec 2000 11:56:28 -0600, "Hwanjo Yu" <hwanjoyu at uiuc.edu>
wrote:

>Hi,
>
>Is there any way to bind a function name at runtime ?
>For instance, we don't know the function name to call, but we know the
>function name is saved in a variable, "fname".
>How to call it in this case ?
>Thanks in advance.
>
Functions are first-class objects in python so it is no big deal to
assign them to other names and then call them. There are also several
ways to resolve names at runtime. Here are few quick examples of how
this could be done.

Bob
#########
# this is the function we will play with
def cb(name):
    print 'callback:', name
    
# this gets passed the function
def foo(name, callback):
    callback(name)

#this gets passed the function name
def bar(name, callbackname):
    globals()[callbackname](name)

# call foo 
foo('joe',cb)

#bind cb to c and call
c=cb
c('ralph')

# use eval to run from function name
eval('cb')('called from eval')

# find the function in the locals dictionary
locals()['cb']('called from locals')

#find the function in the globals dictionary
globals()['cb']('called from globals')

# call bar
bar('cb called from globals in function', 'cb')
# note that we bound to c earlier so can call it as well
bar('c called from globals in function', 'c')





From pbjorn at uswest.net  Sun Dec 17 00:10:14 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Sat, 16 Dec 2000 22:10:14 -0700
Subject: Source Code Management
References: <fb4f3tsgffnmvme64aol1cn397te7ln349@4ax.com> <91aq96$tba$1@nnrp1.deja.com> <3A3AA3B7.3AC3F4BF@uswest.net> <91ggld$ble$1@nnrp1.deja.com>
Message-ID: <3A3C4AB6.E264370F@uswest.net>

Thanks!

(The feature seems like it would still require some text parsing in most
cases however...)

-bjorn

gtalvola at my-deja.com wrote:

> In article <3A3AA3B7.3AC3F4BF at uswest.net>,
>   Bjorn Pettersen <pbjorn at uswest.net> wrote:
> > Do you have a link to how to get the result in Python
> marshalled dictionary
> > format? (I looked at the perforce website but couldn't
> find anything...)
> >
>
> Here's a snippet from the Perforce mailing list about the feature
> (I don't have the original author's name):
>
> > You can use the -G option to Perforce to cause Perforce to do input
> > and output as Python marshalled dictionaries.  (This is a new but
> > unsupported feature in 2000.1 -- see "p4 help undoc".)
> >
> > Using this feature it's easy to query Perforce and read the results.
> >
> >      import marshal
> >      import win32pipe
> >      stream = win32pipe.popen('p4 -G changes', 'r')
> >      changes = []
> >      try:
> >          while 1:
> >              changes.append(marshal.load(stream))
> >      except EOFError:
> >          for c in changes:
> >              print c['change'], c['desc']
> >
> >
>
> Sent via Deja.com
> http://www.deja.com/



From rturpin at my-deja.com  Thu Dec 28 13:00:45 2000
From: rturpin at my-deja.com (rturpin at my-deja.com)
Date: Thu, 28 Dec 2000 18:00:45 GMT
Subject: Python vs. PHP (& Java?)
References: <69BC4C6AE83ED311BC9400805FA7B14AF84D2C@il93exp01.css.mot.com> <mailman.978018791.20357.python-list@python.org>
Message-ID: <92fv4a$l2l$1@nnrp1.deja.com>

"Kemp Randy-W18971" <Randy.L.Kemp at motorola.com> writes:
>> Can we nip language comparisons in the bud?

In comp.lang.*? I doubt it.

In article <mailman.978018791.20357.python-list at python.org>,
  "Alex Martelli" <aleaxit at yahoo.com> wrote:
> In each case, there are some _facts_ that can be pointed
> out, ..

Yep.

And some of these facts point to the danger of reasoning
from natural to programming languages. People use natural
language for human communication. And in almost all
practical contexts, small competencies bring signficant
benefit. If you can ask where the bathroom is in Japanese,
the waiter, recognizing your deficiency in that tongue,
but pleased that you made the attempt, will point the
way. This keeps you from having to translate the Japanese
equivalent of "behind the screen with the crouching tiger."
Which is good, because your back teeth are afloat, and the
only part your limited Japanese knows is "behind."

In contrast, texts in computer languages are more studied
as artifacts. "How did this work?" "Why did it stop
working?" "What did the original authors intend?" They,
the rascals, are no longer around to explain, take credit,
or defend their honor. This kind of comprehension requires
a high level of competency in the language. Knowing
one-tenth of Perl is not enough to understand most Perl
scripts.

Computer languages have more precisely defined semantics
than natural languages. The niggling details are important,
such as what "[[0]*10]*3" means. Natural languages are
more fluid, and it is the nuances of the tacit, multiple
meaning, and almost ambiguous that are important in their
competence.

For these reasons, I am a great believer in small computer
languages, ones that can be explained well in a book no
thicker than my pinkie. The use of extentions should be
recognizable, so that when you come across something you
don't know, you know *what* to look up and where to find
it. "One way to do everything" IS better than "many ways
to do anything."

Of course, being small has its hazards, if core features
are omitted. That's why I am fond of saying that a language's
productivity is proportional to its expressive power, and
inversely proportional to its complexity. (Not that these are
the only considerations.) That is why I am happy to see
Guido resist the temptation to continually add new bells and
whistles.

People tend to focus on the benefit of expressive power,
while overlooking the problem of complexity. C, Lisp, and
Python are examples of languages that pack a lot of
expressive power into simple packages with a small set of
orthogonal concepts. C++, trying to build OO on top of C
while retaining maximal compile time efficiency, grew
bigger than its expressive power requires. But that set
of goals is hard to achieve, and to its credit, C++ does
this. Perl is overly complex for what it does. I think
this was due to its syncretic history, rather than the
scope of what it achieves. Like Alex, I can't imagine any
task that, from a pure language viewpoint, is better done
in Perl than in Python. Practical constraints, such as
a company's preferred language, or which language someone
already knows, could determine otherwise.

Russell


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


From jbalding at hotmail.com  Thu Dec 14 20:14:44 2000
From: jbalding at hotmail.com (jbalding at hotmail.com)
Date: Fri, 15 Dec 2000 01:14:44 GMT
Subject: Is Stackless and uthread.py SMP friendly?
Message-ID: <91br9v$qa4$1@nnrp1.deja.com>

I read with interest a recent thread asking about threading, the global
interpreter lock, and symmetric multiprocessing. Nobody mentioned, as
far as I could see, Stackless Python and microthreads. Having just
picked up a dual PIII machine, will Stackless and uthread.py make my SMP
experience with Python a happy one?

Thanks - John


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


From mfletch at tpresence.com  Thu Dec 14 12:01:03 2000
From: mfletch at tpresence.com (Mike Fletcher)
Date: Thu, 14 Dec 2000 12:01:03 -0500
Subject: How to use mxTextTools
Message-ID: <CC59868A0A76D311B1F50090277B321C1F3745@VRTMAIL>

Something you might find useful would be to look at the mcf.vrml.parser
module, which uses simpleparse (which just spits out mxTextTools tuples) to
process a file into an in-memory node graph.

See http://members.home.com/mcfletch/programming/mcf_vrml.htm for the
mcf.vrml distribution.  Here's some code from there...

	def readNext( self):
		'''Read the next root-level construct'''
		success, tags, next = TextTools.tag( self.data,
ROOTITEMPARSER, self.position )
##		print 'readnext', success
		if self.position >= self.datalength:
			print 'reached file end'
			return None
		if success:
			#print '  successful parse'
			self.position = next
			if self.parseOnly:
				return success
			map (self.rootItem_Item, tags )
			return success
		else:
			return None
	def rootItem (self, (type, start, stop, (item,))):
		''' Process a single root item '''
		self.rootItem_Item( item )
	def rootItem_Item( self, item ):
		result = self._dispatch(item)
		if result is not None:
##			print "non-null result"
##			print id( self.sceneGraphStack[-1] ), id(self.result
)
			self.sceneGraphStack[-1].children.append( result )
	def _getString (self, (tag, start, stop, sublist)):
		''' Return the raw string for a given interval in the data
'''
		return self.data [start: stop]
	def _dispatch (self, (tag, left, right, sublist)):
		''' Dispatch to the appropriate processing function based on
tag value '''
##		print "dispatch", tag
		try:
			function = getattr (self, tag)
		except AttributeError:
			raise AttributeError( '''Unknown parse tag "%s"
found! Check the parser definition!'''%(tag))
		return function( (tag, left, right, sublist) )
	def Proto(self, (tag, start, stop, sublist)):
		''' Create a new prototype in the current sceneGraph '''
		# first entry is always GI
		GI = self._getString ( sublist [0])
##		print "PROTO",GI
		newNode =Prototype (GI)
##		print "\t",newNode
		setattr ( self.sceneGraphStack [-1].protoTypes, GI, newNode)
		self.prototypeStack.append( newNode )
		# process the rest of the entries with the given stack
		map ( self._dispatch, sublist [1:] )
		self.prototypeStack.pop( )
	def fieldDecl(self,(tag, left, right, (exposure, datatype, name,
field))):
		''' Create a new field declaration for the current
prototype'''
		# get the definition in recognizable format
		exposure = self._getString (exposure) == "exposedField"
		datatype = self._getString (datatype)
		name = self._getString (name)
		# get the vrml value for the field
		self.fieldTypeStack.append( datatype )
		field = self._dispatch (field)
		self.fieldTypeStack.pop( )
		self.prototypeStack[-1].addField ((name, datatype,
exposure), field)

HTH,
Mike

-----Original Message-----
From: Paul Moore [mailto:paul.moore at uk.origin-it.com]
Sent: Thursday, December 14, 2000 8:36 AM
To: python-list at python.org
Subject: How to use mxTextTools


Hi,
I'm looking at mxTextTools to see if it would be suitable for some
types of text parsing work I am interested in (nothing concrete yet,
so I can't give specifics...)

The example in the documentation of tagging HTML looks fine - I
understand what's going on there, and as I understand it, this will
give me back a taglist, which is (effectively) the text stream with
portions tagged as I ask.

What I dont't see (yet), and I can't find any good examples for, is
what to do with the resulting taglist. There seem to be no functions
for working with taglists, and the lists themselves seem like
relatively complex data structures, so is it right that I should be
manipulating them "by hand"?

More information, or better still, some complete examples, would be
very helpful. (All the examples in the distribution just use
print_tags() to display the tags, and don't do anything with them...)

Paul

-- 
http://www.python.org/mailman/listinfo/python-list



From kalle at gnupung.net  Tue Dec 12 09:07:06 2000
From: kalle at gnupung.net (Kalle Svensson)
Date: Tue, 12 Dec 2000 15:07:06 +0100
Subject: Q: Print whithout \n
In-Reply-To: <aboZ5.3538$Kd1.410322@newsb.telia.net>
References: <aboZ5.3538$Kd1.410322@newsb.telia.net>
Message-ID: <Pine.GSO.4.21.0012121444320.9661-100000@luna.dsv.su.se>

On Tue, 12 Dec 2000, Henrik M?rtensson wrote:

> How do you print a string whithout a linefeed in the end.

As others have said, print with a , works, but it gives you a space 
instead.  If this disturbs you as it does me, you can use the write method
of sys.stdout (or sys.stderr):

import sys
sys.stdout.write("Hej, ingen nyrad")

If you do something like

import sys
if filename:
  outfile = open(filename, "w")
else:
  outfile = sys.stdout
outfile.write("Whatever...")

outfile can be anything with a write method, which can be nice, e.g. for
handling different output formats.

Peace,
  Kalle
-- 
Email: kalle at gnupung.net     | You can tune a filesystem, but you
Web: http://www.gnupung.net/ | can't tune a fish. -- man tunefs(8)
Not signed due to logistical difficulties (I'm not at home).
If you doubt the authenticity of this message, please tell me.



From cg at schlund.de  Mon Dec 11 08:02:00 2000
From: cg at schlund.de (Carsten Gaebler)
Date: Mon, 11 Dec 2000 14:02:00 +0100
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com>
Message-ID: <3A34D048.62889EE2@schlund.de>

Jaroslav Gergic wrote:

> Do you know better construction to avoid both - duplicate lines
> and horrible 'while 1' construction?
> 
> Something like:
> 
> while( "" != (line = fh.readline()) ):
>   ... do something ...

You could write a wrapper class like this:

class MyFile:
	def __init__(self, filename):
		self.f = open(filename, "r")

	def readline(self):
		self.line = self.f.readline()
		return self.line


And then something like:

F = MyFile("somefile")
while F.readline():
	print F.line


cg.


From moshez at zadka.site.co.il  Wed Dec 27 11:49:00 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Wed, 27 Dec 2000 18:49:00 +0200 (IST)
Subject: Python and Napster
In-Reply-To: <92ap0g12b6r@news1.newsguy.com>
References: <92ap0g12b6r@news1.newsguy.com>, <sx226.18970$RC1.511898@newsread1.prod.itd.earthlink.net>
Message-ID: <20001227164900.DF7E6A84F@darjeeling.zadka.site.co.il>

On Tue, 26 Dec 2000 19:40:14 +0100, "Alex Martelli" <aleaxit at yahoo.com> wrote:

> Python would probably let you be more productive, I suspect.  It sure has
> all the network functionality you need; in your shoes, I'd look into pyro --
> building file exchange on top of it should be easier than on sockets or on
> protocols such as HTTP.

I'm not so sure I agree -- SimpleHTTPServer/urllib/httplib is wonderful 
for file transfers. For anyone who wants to see a practical exmple
(supports both upload and download), have a look at PMEP in the PMS
source distribution (http://pythonms.sourceforge.net)
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From garachop at club-internet.fr  Thu Dec  7 19:09:35 2000
From: garachop at club-internet.fr (Patrice Garachon)
Date: Fri, 08 Dec 2000 01:09:35 +0100
Subject: win32evtlogutil: How does it works ?
References: <3A2D556F.6240C661@club-internet.fr> <3A2D8223.4070408@ActiveState.com>
Message-ID: <3A3026BF.CB70C2E2@club-internet.fr>

Mark Hammond a ?crit :
> 
> Garachon Patrice wrote:
> 
> > Any ideas ?
> 
> Not really - but you are not alone.  I have the same problem, and havent
> been able to track it down.  The PythonService support for the event log
> works fine, and I can't see the difference between the 2 :-(
> 
> I am sure it is just a registry setting, but I can't see what it is...
> 
> Mark.

Today, I have tried to use PythonService.exe as Dll message.
If you use the 1001  or 1002 Id number, it doesn't works but if you  use
0x4000100 1 or 2 It works ?

--
Patrice


From phd at phd.pp.ru  Thu Dec 28 11:24:12 2000
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 28 Dec 2000 19:24:12 +0300 (MSK)
Subject: cgi parameters
In-Reply-To: <3A4B4BD8.6A696387@wielun.dhs.org>
Message-ID: <Pine.LNX.4.30.0012281923020.2669-100000@fep132.fep.ru>

On Thu, 28 Dec 2000, QdlatY wrote:
> <form action="newpost.cgi?index=1" method="post">
                           ^                 ^ you are trying to combine
both GET and POST methods. WWW Servers do not allow it. Pass your value in
a hidden field in the form.

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




From gerhard.nospam at bigfoot.de  Thu Dec 28 14:21:34 2000
From: gerhard.nospam at bigfoot.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Thu, 28 Dec 2000 20:21:34 +0100
Subject: Python with Apache on WinNT newbie question
References: <3a4b6391.4061375@localhost>
Message-ID: <slrn94n584.2cg.gerhard.nospam@lilith.hqd-internal.de>

I recommend to add the following line to httpd.conf under WinNT:

ScriptInterpreterSource Registry

and give your Python scripts the extension ".py". A Python CGI script can
then look like:

#!/usr/bin/env python
print 'Content-type: text/plain\n\nHello!'

and work perfectly under WinNT and Linux!

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

On Thu, 28 Dec 2000 16:07:07 GMT, Victor Muslin <victor at prodigy.net> wrote:
>To get Apache server to run a Python CGI script, the script needs to
>have on its first line a directive telling Apache the file name of the
>Python interpreter, like so:
>
>	#!d:/apps/python/python.exe
>	#
>	# The rest of my cgi script follows
>	# ....
>
>The problem with this is deploying this script on different machines
>that may have Python interpreter installed in different directories.
>Surely others had run into the same issue. What is a typical solution?
>Unfortunately, if this directive can take environmental variables I
>could not figure out how to do it.


From henrik.martensson at frontconsulting.se  Tue Dec 12 06:43:02 2000
From: henrik.martensson at frontconsulting.se (Henrik Mårtensson)
Date: Tue, 12 Dec 2000 11:43:02 GMT
Subject: Q: Print whithout \n
Message-ID: <aboZ5.3538$Kd1.410322@newsb.telia.net>

Hi

How do you print a string whithout a linefeed in the end.

/Henrik




From pbjorn at uswest.net  Sat Dec 23 17:34:41 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Sat, 23 Dec 2000 15:34:41 -0700
Subject: Some Python 2.1 ideas
References: <mailman.977590932.19335.python-list@python.org> <Cw616.23054$qN6.3308205@typhoon.we.rr.com>
Message-ID: <3A452881.7686E717@uswest.net>

string.chomp() (a better name would be nice) needs to handle all of:

 - "foo\n"
 - "foo \n"
 - "foo \r\n"
 - "foo \n\r"
 - "foo"

etc. correctly, which means it should delete between zero and two bytes
(characters) from the end of the string. This is very different from
removing leading or trailing whitespace (whitespace can be significant
you know <wink>).

Before someone suggests:

   line = line[:-1]

I would note that this removes one character too many if you're reading
a file that doesn't end in a newline.

-- bjorn

Pete Shinners wrote:

> "Bob Alexander" <bobalex at home.com> wrote
> > String method to remove trailing newline
>
> there is a string method named "strip()" that will
> remove all whitespace (newline&linefeeds included)
> from a string.
>
> actually, since strings are immutable, it returns a
> new string with the appropriate changes made.
>
> there are also "rstrip" and "lstrip" functions which
> will remove whitespace from only the end or beginning
> of the string.
>
> >>> print 'Example string gotten from readline\n'.strip()
> 'Example string gotten from readline'



From richard at richardgordon.net  Sun Dec 10 21:45:02 2000
From: richard at richardgordon.net (Richard Gordon)
Date: Sun, 10 Dec 2000 21:45:02 -0500
Subject: recursion problem
Message-ID: <v04205500b659eaae1ad5@[64.1.38.58]>

hi-

I'm stuck on something that I thought would be easy, but has got me stumped.

The scenario is that I am using python to automate the construction 
of a web index page that uses one of the assorted 
expanding/collapsing file tree javascripts (e.g., it is organized 
into folders and files within folders based on the physical location 
of stuff on the server). This particular one uses <ul> and <li> tags 
along with some trick style stuff to expand or collapse various 
levels of the hierarchy and looks more or less like the windoze 
explorer stuff once the dust settles.


Anyway, walking the tree, figuring out what's a file and what's a 
directory and that kind of thing are easy, but where I'm getting 
stuck is in dealing with deeply nested folders. This will probably 
get screwed up in email, but what I want is something like:

<ul>Level0folder
	<li>L0file
	<li>L0file
	<ul>Level1folder
		<li>L1file
		<ul>Level2folder
			<li>L2file
			<li>L2file
		</ul>
		<li>L1file
	</ul>
</ul>

So I am basically just reading the top level directory, then testing 
each item to see if it is a file or a folder. If it's a file, I just 
call a function that prints '<li><a href="' + fullpath + '">' + 
filename + '</a>' while if it's a directory, I go into another 
function that is supposed to do the initial <ul> stuff, then calls 
the file printer function for each file, then closes things out with 
</ul> before passing control back to the initial treewalker.

Anyway, this is getting all screwed up as far as the closing </ul> 
tags are concerned and I have managed to create the world's widest 
web page (kind of cool if you like fractals), hit a recursion limit, 
and otherwise toyed with it until I am exasperated. The question is 
how to drill down thru an arbitrary number of subdirectories and get 
the closing </ul> tags in  the right places since the nesting 
prevents just writing one out every time you exit a directory?

Thanks for any help you may have to offer.

Richard Gordon
--------------------
Gordon Design
Web Design/Database Development
http://www.richardgordon.net



From peter at engcorp.com  Sat Dec 30 21:12:13 2000
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 30 Dec 2000 21:12:13 -0500
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <slrn94shbe.150.grey@teleute.rpglink.com> <X8s36.2731$Qb7.305339@newsb.telia.net> <slrn94smds.150.grey@teleute.rpglink.com> <FOt36.17$vd1.2224@e420r-atl2.usenetserver.com> <slrn94t1vj.150.grey@teleute.rpglink.com>
Message-ID: <3A4E95FD.D6C737A7@engcorp.com>

Steve Lamb wrote:
> 
> On Sat, 30 Dec 2000 23:00:53 GMT, Kragen Sitaker <kragen at dnaco.net> wrote:
> >Hahaha!  You just plonked Fredrik Lundh in comp.lang.python.  He was
> 
>     I can tell you I'll get far better without him than with him when he is
> giving advice that runs contrary to about 10+ years of programming that I've
> been taught from various different sources.  I don't care if he is God
> himself, advocating not initing a variable is stupid, plain and simple.  He
> wanted to argue the point for nothing more than argument's sake, so be it.  I
> don't need his advice any further.

It appears to me that you've missed some of the points made in this
"discussion".  I would suggest you go back with an open mind and
reconsider the validity of the opposing views.  The advice that runs
contrary to your 10+ years seems very reasonable to my 20+ years of
programming experience, and apparently many others with a variety of
experiences agree.  

I understand the point you are making; I'm just pointing out that the
opposing views seem well-founded.  Areas where I think you missed the
point are in the importance of code maintainability, the increased
complexity with your suggested approach, and the feelings of the "entire
CS community" on this issue.

If I may attempt to clarify what I believe is the opposing argument:  you
suggested an alternative which relies on initialization of a variable to a
special value (anything less than 25) solely in order to enter a 'while'
loop which you've rewritten to use the conditional at the top instead of
at the bottom (because you believe there is some inherent advantage in
doing this, though you've yet to explain what it is other than that the
use of 'break' is somehow bad), and in the process you've demonstrated the
*reason* others feel this is a poor approach, by leaving out the required
initialization.  And the first attempt to correct this was given using a
wrong value for the initializer!  That's the whole point: your approach is
inherently open to such complexity/maintainability problems.  Even
experienced programmers can slip up in this way.

I think those of us objecting have in our 10+ years of experience seen
this kind of thing cause so many problems that we are happy with the
single _idiomatic_ way which Python provides, in spite of a little
so-called "ugliness".  

I suspect some of us would even argue that the conditional in 'do/until'
can be difficult for some people to get right and that forcing the use of
'while/if/break' may actually increase the likelihood the resulting code
will be correct the first time, and be interpreted correctly during
maintenance.  My guess is that for this precise reason the alternative
loop construct was removed from/never added to Python in the first place.


From chris_barker at my-deja.com  Tue Dec 26 14:58:03 2000
From: chris_barker at my-deja.com (chris_barker at my-deja.com)
Date: Tue, 26 Dec 2000 19:58:03 GMT
Subject: Source for Linux Binaries?
Message-ID: <92at88$rig$1@nnrp1.deja.com>

Hi all,


I'm looking for a source for Linux binaries for Python 2.0, and a whole
bunch of modules to go along with it (PIL, Numeric, wxPython, etc.) I
know the main Python web page has rpms for the main distro, but none of
the add on packages.

Has anyone set up a comprehensive source for Linux binaries? My foirst
choice is RedHat 6.1 compatable rpms, but I'm flexible on that one. What
I'm looking for is something along the line of Oliver Andrich's old
site.

If this doesn't exist, I may just have to set it up myself!

-Chris



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


From phd at phd.pp.ru  Tue Dec  5 07:07:27 2000
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 5 Dec 2000 15:07:27 +0300 (MSK)
Subject: Random string
In-Reply-To: <200012042351.QAA24052@mail.cybermesa.com>
Message-ID: <Pine.LNX.4.30.0012051506370.24848-100000@fep132.fep.ru>

On Mon, 4 Dec 2000, Jay O'Connor wrote:
> 	Is there a good algorithm or existing function for creating a random string such
> as "AFYGADFGWETOQW6DSJD", suitable as a 'session id' for a web session

   Run md5 or sha1 over random int, convert the hash to a string and take
any number of characters.

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




From jepler at inetnebr.com  Fri Dec 15 09:03:30 2000
From: jepler at inetnebr.com (Jeff Epler)
Date: Fri, 15 Dec 2000 14:03:30 GMT
Subject: __file__ in 2.0
References: <bXEZ5.2409$0U2.143972@news20.bellglobal.com>
Message-ID: <slrn93k95h.lb.jepler@potty.housenet>

On Wed, 13 Dec 2000 06:46:31 GMT, Bryan Mongeau
 <bryan at eevolved.com> wrote:
>And my final question: How do I call __file__ on static modules in 2.0??

You can't.  But if modulename in sys.builtin_module_names, then you know
that no __file__ attribute will be defined.

Jeff


From phrxy at csv.warwick.ac.uk  Tue Dec 12 08:39:58 2000
From: phrxy at csv.warwick.ac.uk (John J. Lee)
Date: Tue, 12 Dec 2000 13:39:58 +0000
Subject: gcc and Python 2.0 on windows
Message-ID: <Pine.SOL.4.30.0012121312330.16154-100000@mimosa.csv.warwick.ac.uk>

It seems that the Python 2.0 headers have been patched to allow the use of
gcc on windows following Paul Sokolovsky's patch to 1.5.2, but only in
config.h, not in object.h.  I haven't noticed any reference to these
changes in the C extension tutorial or reference.  I managed to compile
and run something with the plain Python 2.0 headers, so I presume it's
fully working - but I don't know what the changes to object.h in Paul's
patch are supposed to do - something to do with __declspec(dllimport).
Are these changes redundant now?

Maybe some mention of gcc / win should be made in the docs somewhere.

Thanks for any help.


John



From landauer at apple.com  Wed Dec 27 19:42:32 2000
From: landauer at apple.com (Doug Landauer)
Date: Wed, 27 Dec 2000 16:42:32 -0800
Subject: A Suggestion for Python Colon Syntax
References: <mailman.977962510.16011.python-list@python.org>
Message-ID: <landauer-EB07F4.16423227122000@news.apple.com>

> ... You should be able to tell from the overwhelming lack of
> other response on the newsgroup that there's no groundswell
> of desire (for making trailing colons optional) being suppressed
> here! 

Well, there is the fact that it's currently between Chrismas
and New Year's, and a lot of people aren't "here".  But aside from
that, I suspect

 - that 90% of the responses that are being suppressed are from
   lack of interest and/or a confidence that The Right Thing
   (i.e., No change) Will Happen; and

 - that most of those responses, were they to be written, would say
   something like this:   "Why ever would you want to make your
   code less consistent and less readable by omitting the colons?"

barely here-ly,
-- 
    Doug Landauer        landauer at apple.com (work)
                         landauer at scruznet.com (not-work)


From darrell at dorb.com  Tue Dec  5 19:06:59 2000
From: darrell at dorb.com (Darrell Gallion)
Date: Tue, 5 Dec 2000 19:06:59 -0500
Subject: an apply question
References: <007b01c05f15$34dcc750$3eaf2cd5@malanwin2000>
Message-ID: <0ded01c05f18$749ffdf0$6401a8c0@home>

>>> def a():
...     print 'a'
...
>>> myMap={'a':a}
>>> for k, f in myMap.items():
...     apply(f,())
...
a
>>>

func1 in MyMap shouldn't be a string.

--Darrell

----- Original Message ----- 
From: "Alan Marchand" <malan at club-internet.fr>
> In Python 2.0:
> 
> What am I doing wrong?
> 
> #!/bin/env python
> #############
> MyMap = {
> 'Key1' : 'func1',
> 'Key2' : 'func2',
> 'Key3' : 'func3',
> 'Key4' : 'func4'
> }
> 
> def func1 ( ):
>     print "Here's func1"
> 
> items = MyMap.items( )
> for x in range( len(items) ):
>     key, func = items[x]
>     if key == "Key3":
>         apply( func, ( ) )
> -------------------------
> c:\devpy\python test.py
> 
> Traceback (most recent call last):
>   File "test.py", line 17, in ?
>     apply( func, ( ) )
> TypeError: Call of non-function (type string)





From loewis at informatik.hu-berlin.de  Fri Dec  8 15:07:38 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 08 Dec 2000 21:07:38 +0100
Subject: newbie to python: can't load tkinter -- please help
References: <90kg3p$n2p$0@pita.alt.net>
Message-ID: <p6qofymk779.fsf@informatik.hu-berlin.de>

"Mayer Goldberg" <gmayer at cs.bgu.ac.il> writes:

> Can someone please help?

Where did you get Python from? If it is ActivePython, then you need to
install Tcl 8.3 separately.

Regards,
Martin



From cg at schlund.de  Fri Dec 22 11:52:29 2000
From: cg at schlund.de (Carsten Gaebler)
Date: Fri, 22 Dec 2000 17:52:29 +0100
Subject: Again: compiling Python 2.0 against Berkeley DB
Message-ID: <3A4386CD.5DE7EE06@schlund.de>

Hi!

I have to correct my posting regarding Berkeley DB 3.0.55. I *thought* I
was trying to read a 3.0.55 file. In fact it's a version 1.85 file, and
im trying to read it with Python compiled against BerkeleyDB 3.0.55
(dbhash.open() fails with "Invalid argument"). I have compiled
BerkeleyDB 3.0.55 with 1.85 compatibility API but it doesn't like 1.85
files. Then I downloaded version 1.85 from sleepycat.com but that won't
compile on my machine (numerous 'parse error before "("' messages).

I'm stuck. How can I make my Python read those DB1.85 files?

Regards
Carsten.


From robin at jessikat.fsnet.co.uk  Mon Dec  4 06:17:29 2000
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Mon, 4 Dec 2000 11:17:29 +0000
Subject: PyThreadState_Get error again :(
References: <42+P6LAmqiK6Ewl1@jessikat.fsnet.co.uk>
 <3A2AD10F.8040803@ActiveState.com>
Message-ID: <uTVaDtAJ13K6EwUw@jessikat.fsnet.co.uk>

In article <3A2AD10F.8040803 at ActiveState.com>, Mark Hammond
<MarkH at ActiveState.com> writes
>Robin Becker wrote:
>
>> I'm running BEOPEN Python 2.0, Win95 OSR2, PythonWin 134 (with the fixed 
>.pyds).
>
>I'm afraid that using "popen4("dir")" as the test, using ActivePython 
>201 (ie, win32all 135), on Windows 98, I can not get this code to fail 
>either from python.exe or pythonwin.exe :-(
>
>Mark.
...
Well it's your co-author who claims it works with 2000 :)

I can certainly 'claim' that there's something odd happening under
win95. The good news is that we don't need the code any more ie we
brought the calculation inboard.

The crash happens when the output file is closed even when I try
popen4("dir \\tmp"). The command seems to work and the output looks ok,
just crashes when either of the files is closed.

Is there any way to get 135 without reinstalling 2.0 or perhaps there
has been a fix in that? Is it worth pursuing further?
-- 
Robin Becker


From MarkH at ActiveState.com  Tue Dec  5 19:12:18 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Wed, 06 Dec 2000 00:12:18 GMT
Subject: activex scripting and debugging with Python win32com
References: <90irub$24b$1@troll.powertech.no>
Message-ID: <3A2D8317.20807@ActiveState.com>

Syver Enstad wrote:

> Anybody knows if there are any plans to rectify this, and in what time
> perspective. I would be interested in doing some work on making this feature
> work as it would help me work more productively with Python at work.


I would like to help fix this, but don't currently have an ASP 
environment running.  A good first step would be to confirm it _does_ 
work with IE and WSH (as it does for me, and I can work on that if it 
doesnt for you).  After that, we need to work out what is different 
about ASP.

Just mail me, and we can work through it - the more people that 
understand this the better!!!

Mark.



From SBrunning at trisystems.co.uk  Wed Dec 13 06:04:22 2000
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Wed, 13 Dec 2000 11:04:22 -0000
Subject: Smalltalk and Python
Message-ID: <31575A892FF6D1118F5800600846864D5B15DB@intrepid>

> From:	Alex Martelli [SMTP:aleaxit at yahoo.com]
> > | It's an absolutely stunning language for free, and although not as
> fast
> > | as Perl
> >
> > (it's as fast in my experience)
 
This site - <http://www.bagley.org/~doug/shootout/> - shows Python being
outperformed by Perl in a number of categories, sometimes by a fair margin.

Not that I care too much about this myself - I'd rather write a pretty
program in Python that works than an ugly, fast one in Perl which probably
doesn't . Still, I wonder if we could suggest some tweaks to his Python
scripts which might improve Python's score.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk
There is no pleasure in having nothing to do; the fun is in having lots to
do and not doing it. - Mary Little




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From db3l at fitlinxx.com  Fri Dec 22 19:20:07 2000
From: db3l at fitlinxx.com (David Bolen)
Date: 22 Dec 2000 19:20:07 -0500
Subject: IP adresses are not unique enough.
References: <3a413a17.3416402@192.168.0.1> <3A41475A.5DC4ABD8@san.rr.com> <3a4148ad.7150467@192.168.0.1> <uvgscy5fr.fsf@roundpoint.com>
Message-ID: <uk88srnt4.fsf@ctwd0143.fitlinxx.com>

Ben Hutchings <ben.hutchings at roundpoint.com> writes:

> gogo at bluedynamics.com (Georg Gogo. BERNHARD) writes:
> <snip>
> > The MAC adresses are unique, as I understood.
> <snip>
> 
> I have some anecdotal evidence of cases where a manufacturer assigned
> duplicate MAC addresses.  Someone found me a documented case, too:
> <http://puck.nether.net/lists/cisco-nsp/2731.html>.

The global MAC address space is designed to be unique, but given that
most devices depend on programmed information during the manufacturing
process to be assigned the MAC address, there's always the risk of
error - either user (programming the wrong base into a programmer) or
machine (a broken programmer).

There is also the possibility (the feasibility of this on PCs probably
depends on the card interface and OS) of programming a MAC to use a
portion of the local address space, in which case it's up to local
management to manage this.  This sort of thing was much more prevalent
on end stations I believe with Token Rings rather than Ethernets but
its possible.  And it's actually used to good effect in systems where
machines back up each other (like a pair of Cisco routers) where the
backup takes over the primary's MAC address in a fallover situation.

In a locally reprogrammed case, you're still pretty safe assuming
local uniqueness to the wire (or things wouldn't function too well),
but not necessarily global uniqueness.  I suppose you could detect a
local address (bit 1 of the first octet) and treat that specially or
something.

But aside from chip serial numbers (and the ruckus that raised), MAC
addresses from ethernet chipsets are probably the closest thing to a
unique computer stamp (while that chipset is present at least) as we
have.  But it's not without some risks of duplication nonetheless.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From qdlaty at wielun.dhs.org  Thu Dec 28 09:19:04 2000
From: qdlaty at wielun.dhs.org (QdlatY)
Date: Thu, 28 Dec 2000 15:19:04 +0100
Subject: cgi parameters
Message-ID: <3A4B4BD8.6A696387@wielun.dhs.org>

Hello!

I need to pass index parameter to script executed by post method (via
form)

I tried to use following code:

<form action="newpost.cgi?index=1" method="post">

But this won't work
(fieldstorage doesn't have "index" property)

I'm counting on Your help :)

Yours
QdlatY


From pbjorn at uswest.net  Tue Dec 26 16:33:33 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Tue, 26 Dec 2000 14:33:33 -0700
Subject: <> vs != was Re: A Suggestion for Python Colon Syntax
References: <mailman.977595017.22018.python-list@python.org> <92av8a$38f$1@hecate.umd.edu>
Message-ID: <3A490EAD.BFE1EDDA@uswest.net>

William Djaja Tjokroaminata wrote:

> I personally think != is better than <>, just because not equal is != in
> C/C++ (or Java too?  Now I don't remember what is not equal in Java).  I
> like Python and I use it as much as I can.  However, there isn't
> a single language which is good for everything.  I still use C/C++ a lot
> (and also Perl for one-liners and some text processing), and therefore I
> have to switch between the languages quite often.  Isn't it great if all
> the languages agree on what not equal should look like?  We will get
> syntax error less often and we have to open the reference books less often
> too.  Well, I still have to find my Java books to see what not equal is in
> Java.... (sigh....)

Well, then it is clear, we should never use !=, since if it was meant to mean
not equal it would certainly have been spelt !==. != should of course be
changed to the not assingment operator which would work perfectly in e.g.
loops, where nobody sane would want to have an assignment. <> of course,
should be read as either smaller or greater which everyone intuitively knows
means not equal.

this-is-all-the-timbot's-fault-and-I-should-probably-include-a-<wink>-somewhere'ly
y'rs
-- bjorn



From aleaxit at yahoo.com  Tue Dec 26 19:00:32 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 27 Dec 2000 01:00:32 +0100
Subject: saving NNTP bodies to disk
References: <3a4925f7.51675327@news.skynet.be>
Message-ID: <92bbhp02vfb@news1.newsguy.com>

"Daniel" <Daniel.Kinnaer at AdValvas.be> wrote in message
news:3a4925f7.51675327 at news.skynet.be...
    [snip]
> try to write a list to disk using f.write() and Python complains about
> this with following report
>
>  f.write(TheBody[i])
> TypeError: read-only character buffer, list
>
> Below is the script. Can someone please explain me what I'm doing
> wrong?  All help is greatly appreciated.  Thanks

Python is saying: I need a read-only character buffer, you give me
a list -- implying, a list has no conversion to such a buffer.

If TheBody[i] is a list of strings, you can write it to file with
    f.writelines(TheBody[i])


Alex





From thomas.wright1 at ntlworld.REMOVETHIS.com  Thu Dec 28 12:52:43 2000
From: thomas.wright1 at ntlworld.REMOVETHIS.com (Tom wright)
Date: Thu, 28 Dec 2000 17:52:43 -0000
Subject: sniffing btw my web browser and a server
References: <92aush$rhc$1@news.nuri.net> <92fs3l$5gg$01$1@news.t-online.com> <92ftpn$e2j$1@news.nuri.net>
Message-ID: <P7L26.12668$Yy.360733@news2-win.server.ntlworld.com>

I can email it to you if u like, its not very big, but is seems abit buggy,
doesnt like webdav at all. This is under Python 2.0

Regards

Tom

"June Kim" <junaftnoon at nospamplzyahoo.com> wrote in message
news:92ftpn$e2j$1 at news.nuri.net...
> "Eduard Hiti" <KatMouse at gmx.de> wrote :
>
> [snip]
> >
> > There's also a Python only solution: 'WebDebug'. It also acts as a
proxy,
> > logging all transfers and generating a nice and comprehensive HTML
report.
> >
> > The URL: http://www.cyberclip.com/webdebug/download.html
>
> This one's what I've been looking for, but the link seems broken. Does
> anyone have it or know alternative links?
>
> Best regards,
>
> June
>




From aleaxit at yahoo.com  Mon Dec 11 04:35:41 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 11 Dec 2000 10:35:41 +0100
Subject: type vs. class
References: <3A34244A.415F4DFB@t-online.de> <uE_Y5.133498$U46.4497870@news1.sttls1.wa.home.com>
Message-ID: <91275n02trc@news2.newsguy.com>

"Greg Jorgensen" <gregj at pobox.com> wrote in message
news:uE_Y5.133498$U46.4497870 at news1.sttls1.wa.home.com...
> "Thomas Thiele" <tnt-audio at t-online.de> wrote:
>
> > >>> class X:
> > pass
> >
> > >>> x = X()
> > >>> type(x)
> > <type 'instance'>
> >
> > How can I realize that I will not get type 'instance' but type 'X'? I
> > have different classes, more than one, and I want to check out which
> > class it is.  Are all classes type instance?
>
> In Python a class does not create a new type. All classes are of type
> ClassType, and all class instances are of type InstanceType.
>
> The built-in function isinstance(object, class) may be what you need:

Perfectly true -- isinstance() is often more useful than type()!

However, to satisfy Thomas's request as he posed it...: when x is an
instance of some class, you can learn WHICH class it's an instance
of, through x's __class__ attribute (you can get the NAME of the
class as a string, through the class-object's own __name__ attribute):

>>> class X:
...   pass
...
>>> x=X()
>>> x.__class__
<class __main__.X at 007F76CC>
>>> x.__class__.__name__
'X'
>>>


Alex






From neilh at scintilla.org  Tue Dec  5 07:50:26 2000
From: neilh at scintilla.org (Neil Hodgson)
Date: Tue, 05 Dec 2000 12:50:26 GMT
Subject: Python programs (or scripts) as drop targets
References: <hv5k2tg9cdf33j8v333qpkk7jo45ht596n@4ax.com> <90d9hi029hm@news1.newsguy.com> <3A2ACF21.4060406@ActiveState.com> <90fl26022in@news2.newsguy.com> <3A2C73CE.5060506@ActiveState.com> <90icu9023as@news2.newsguy.com>
Message-ID: <mw5X5.5586$xW4.46771@news-server.bigpond.net.au>

Alex Martelli:

> <shudder> if we're talking DDE, there is *another* issue which I
> find more pressing: DDEML is *NOT* usable in professional apps
> (although many apps that call themselves professional do use it).

   Agree completely here - various versions of DDEML have also had some
rather nasty race conditions. I once wrote an application based upon DDEML
only to stumble across these problems. There is the additional burden when
using DDEML of really needing to understand the underlying DDE protocol at
the same time as DDEML's 'simplified' version at the same time to debug. Its
more trouble than just learning DDE.

   Later, a colleague who was familiar with the history of DDEML within MS
said that it was an orphaned project and no one wanted to maintain it.

   Neil





From gregj at pobox.com  Sat Dec  9 01:02:18 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Sat, 09 Dec 2000 06:02:18 GMT
Subject: Python wrapper for C++ STL?
References: <3A31760E.D157EB54@erols.com> <C0gY5.126062$U46.4232859@news1.sttls1.wa.home.com> <90s3fk$j8p$1@nnrp1.deja.com> <fUgY5.126218$U46.4241063@news1.sttls1.wa.home.com> <90s9l9$njd$1@nnrp1.deja.com>
Message-ID: <KVjY5.126920$U46.4261650@news1.sttls1.wa.home.com>

<shindich at my-deja.com> wrote:

> I am not saying that there is any point in wrapping STL for Python. All
> I'm saying is that STL can be used to write C++ extensions that
> implement specific algorithms.

Sure it could, though C++ + STL = C O D E  B L O A T, so using that combo
for Python modules will likely suck up a lot of memory. If you want some
data structure Python doesn't have--sorted associative containers, for
example--you'd be better off writing a module in C to implement just that,
rather than try to press the STL into service where it won't give you any
benefit.

Anyway the original post that started this all asked about writing a wrapper
for the STL, and that's the issue I was responding to.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From amk at mira.erols.com  Fri Dec 15 13:52:58 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 15 Dec 2000 18:52:58 GMT
Subject: comparable shelve function in xml
References: <91dg2q$3a1$1@nnrp1.deja.com>
Message-ID: <slrn93kqt5.1sm.amk@mira.erols.com>

On Fri, 15 Dec 2000 16:15:31 GMT, Stan Lee <slee4812 at my-deja.com> wrote:
>Is there a comparable way to store Python objects in an XML formatted
>file like the shelve module does?  I want to be able to store Python
>objects to a file that's human-readable on both a Windows and Linux
>platform.  Thanks.

There are a few.  Hiding inside Zope there's a ppml.py file which has
much the same functionality as the pickle module.  Hiding inside the
PyXML codebase, there's an xml.marshal package which has support for
WDDX, XML-RPC, and a generic Python marshalling format.  It depends if
you only need simple data types -- integers, strings, that sort of
thing -- or need the full generality of handling objects, recursive
lists, &c.

--amk


From root at [127.0.0.1]  Mon Dec 11 08:22:06 2000
From: root at [127.0.0.1] (deadmeat)
Date: Mon, 11 Dec 2000 13:22:06 GMT
Subject: Pascal -> Python 'record' structures
Message-ID: <2y4Z5.7634$KY1.20696@news1.rivrw1.nsw.optushome.com.au>

Can anyone tell me how to read in file records that were created by Pascal
and C programs? Will I have to write a wrapper and parse the raw data
manually?





From chris at araidesign.com  Tue Dec 19 03:21:37 2000
From: chris at araidesign.com (Chris Arai)
Date: Tue, 19 Dec 2000 00:21:37 -0800
Subject: Need Help sorting alphabetically.
Message-ID: <3A3F1A91.A7ED9182@araidesign.com>

Hi,

This seems like it should be obvious, but it is evading me!

I would like to sort alphabetically (I'm not sure that is the correct
term) so that strings of alpha numerics sort alphabetically instead of
by ASCII order.

eg:
>>> m=['Pg95_100.GIF',
'Pg9_100.GIF','Pg95A_100.gif','Pg9A_100.gif','Pg10_100.gif']
>>> m.sort()
>>> m
['Pg10_100.gif', 'Pg95A_100.gif', 'Pg95_100.GIF', 'Pg9A_100.gif',
'Pg9_100.GIF']


I would like the order to be
['Pg9_100.GIF', 'Pg9A_100.gif', 'Pg10_100.gif, 'Pg95_100.GIF',
Pg95A_100.gif' ]

What is the obvious answer??

Thanks,


--
Chris Arai
Arai Design
chris at araidesign.com
http://www.araidesign.com
707 431 1299
FAX 431 1599

-------------- next part --------------
A non-text attachment was scrubbed...
Name: chris.vcf
Type: text/x-vcard
Size: 224 bytes
Desc: Card for Chris Arai
URL: <http://mail.python.org/pipermail/python-list/attachments/20001219/1bc76912/attachment.vcf>

From grey at despair.rpglink.com  Fri Dec  8 16:03:14 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Fri, 08 Dec 2000 21:03:14 -0000
Subject: GNU tools in a library?
References: <slrn92vrnb.1v5.grey@teleute.rpglink.com> <p6qaea6k5mu.fsf@informatik.hu-berlin.de>
Message-ID: <slrn932iqq.bgc.grey@teleute.rpglink.com>

On 08 Dec 2000 21:41:29 +0100, Martin von Loewis
<loewis at informatik.hu-berlin.de> wrote:
>I believe there is a Linux boot disk that implements a standard shell
>with a small set of utilities as Python functions.

    Nothing more formal, however?

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From sandj.williams at gte.net  Wed Dec 13 12:39:48 2000
From: sandj.williams at gte.net (Steve Williams)
Date: Wed, 13 Dec 2000 17:39:48 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com> <r0pwvd51gga.fsf@black.cs.washington.edu> <9165he$lon$1@panix3.panix.com> <917pvv02clr@news2.newsguy.com> <3A379CB4.405992F8@gte.net> <9187tq02qlf@news2.newsguy.com>
Message-ID: <3A37B57B.ABC2D672@gte.net>

Alex Martelli wrote:

> "Steve Williams" <sandj.williams at gte.net> wrote in message
> news:3A379CB4.405992F8 at gte.net...
> > Alex Martelli wrote:
> >
> > > [lots of snip]
> >
> We don't really understand software development's technical
> underpinnings quite as well as the Romans understood building's,
> but I think that is not the key reason for most software's
> defects; rather, I think most of them come from items that
> stem more from management issues than strictly technical ones.
>

Here's an alternate viewpoint:

http://upside.com/Open_Season/3a3661271.html





From kc5tja at garnet.armored.net  Fri Dec  8 15:52:20 2000
From: kc5tja at garnet.armored.net (Samuel A. Falvo II)
Date: 8 Dec 2000 12:52:20 -0800
Subject: Dare I say it...
References: <0ivv2tk3h07rp31c2mtheqsns6voriqq71@4ax.com> <p6q7l5ak5jx.fsf@informatik.hu-berlin.de>
Message-ID: <slrn932iad.b96.kc5tja@garnet.armored.net>

On 08 Dec 2000 21:43:14 +0100, Martin von Loewis wrote:
>Not sure, yet. What application of that character do you have in mind?
>I.e. what API that you want to call requires a Unicode string as part
>of a structure?

Perhaps some COM stuff on Windows, but I'm not sure.  At any rate, it's
still something that needs to be considered.

On some systems, wchar_t is 16 bits (in particular, Win32).  On others, it's
32-bits (e.g., Linux).  Knowing the width of wchar_t ahead of time is
required to do it manually.

Just because the author of the pack module didn't see a need for it doesn't
mean a need doesn't exist.  :)

-- 
KC5TJA/6, DM13, QRP-L #1447 | Official Channel Saint, *Team Amiga*
Samuel A. Falvo II	    |
Oceanside, CA		    |


From mwh21 at cam.ac.uk  Sun Dec 10 06:45:17 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 10 Dec 2000 11:45:17 +0000
Subject: Bug: Readline getting stuck on Linux and Solaris
References: <lxk8zqwupy2.fsf@ioasun3.epfl.ch> <90m53k$ovh$1@nnrp1.deja.com> <m3u28h48ja.fsf@atrus.jesus.cam.ac.uk> <slrn92u8al.76b.quinn@lira.ugcs.caltech.edu> <m3r93k4kqk.fsf@atrus.jesus.cam.ac.uk> <slrn932eoj.d1j.quinn@ngwee.ugcs.caltech.edu> <m34s0c4ik7.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <m31yvg4i0i.fsf@atrus.jesus.cam.ac.uk>

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

> do the IRIX manpages say anything scary about signal/sigaction
> interaction?

http://www.chemie.fu-berlin.de/cgi-bin/man/sgi_irix?sigaction+2

  The POSIX signal routines (sigaction(2), sigpending(2),
  sigprocmask(2), sigsuspend(2), sigsetjmp(3)), and the 4.3BSD signal
  routines (sigvec(3B), signal(3B), sigblock(3B), sigpause(3B),
  sigsetmask(3B)) must NEVER be used with signal(2) or sigset(2).

That'd be "yes", then.

But I've just d/l-ed readline and it's signal handler is (predictably)
straight out of Preprocessor Hell, so I can't tell what's going on at
all :-(.

M.

-- 
  ... but I guess there are some things that are so gross you just have
  to forget,  or it'll destroy something within you.  perl is the first
  such thing I have known.              -- Erik Naggum, comp.lang.lisp


From cpr at emsoftware.com  Mon Dec  4 14:06:29 2000
From: cpr at emsoftware.com (Chris Ryland)
Date: Mon, 04 Dec 2000 19:06:29 GMT
Subject: Researching Python IDEs, debuggers, code coverage, unit testing and mem management tools
References: <3A2A175C.D144ACA6@hotmail.com> <90fd6f$3vp$1@news1.xs4all.nl>
Message-ID: <90gpvg$dih$1@nnrp1.deja.com>

In article <90fd6f$3vp$1 at news1.xs4all.nl>,
  boud at rempt.xs4all.nl wrote:

> I've just got an article published that compares Pythonworks,
> Pythonwin, Wing IDE and IDLE on InformIT: http://www.informit.com.

That'd be wonderful to read, but informit has a broken user
registration module right now (at least for me after many tries) and
requires registration to read articles. Could you perhaps post it
somewhere else?
--
Chris Ryland * Em Software, Inc. * www.emsoftware.com


Sent via Deja.com http://www.deja.com/
Before you buy.


From DanielK at aracnet.com  Mon Dec 25 16:37:20 2000
From: DanielK at aracnet.com (Daniel Klein)
Date: Mon, 25 Dec 2000 13:37:20 -0800
Subject: __init__ question
References: <EvM16.270$LU6.118037@typhoon.aracnet.com> <9285oj$6rh$1@panix2.panix.com>
Message-ID: <t0P16.273$LU6.119021@typhoon.aracnet.com>

Is this part of the syntax or is this a regular expression? I can't find
this anywhere in the online documentation or in any of the four Python books
I have purchased. :^(

Dan:


"Aahz Maruch" <aahz at panix.com> wrote in message
news:9285oj$6rh$1 at panix2.panix.com...
> In article <EvM16.270$LU6.118037 at typhoon.aracnet.com>,
> Daniel Klein <DanielK at aracnet.com> wrote:
> >
> >    def __init__(self, *args):
> >
> >What is the meaning of the '*args' expression?
>
> It means "accept any number of parameters here"; "args" is a tuple.  For
> example, I could create a print function instead of the blecherous
> redirected print statement as follows:
>
> def print (file=None, *args):
>     if not file:
>         file = sys.stdout
>     for arg in args:
>         file.write("%s " % arg)
>     file.write("\n")
> --
>                       --- Aahz (Copyright 2000 by aahz at pobox.com)
>
> Androgynous poly kinky vanilla queer het    <*>
http://www.rahul.net/aahz/
> Hugs and backrubs -- I break Rule 6
>
> '"Crisp" is a good quality for crackers; less so for pot roast.'  --pnh




From jadestar at idiom.com  Mon Dec 11 07:13:31 2000
From: jadestar at idiom.com (James T. Dennis)
Date: 11 Dec 2000 12:13:31 GMT
Subject: 3rd argument to os.path.walk(): Why?
Message-ID: <912gdb$1oqr$1@news.idiom.com>

  In the _Python_Essential_Reference_, by David Beazley
 (New Riders Publishing, 2000) I notice that os.path.walk()
 takes three arguments which he lists as:

	path, visitfunc, arg

 It notes that the function referenced in visitfunc is called
 with (arg, dirname, namelist) for every directory that is traversed
 by the os.path.walk.

 I don't understand what the "args" is for.

 I can see how to use the other arguments.  Here's a couple of 
 trivial examples:

	import os
	def print_dirnames(a, dir, flist):
		print dir

	def list_directories(a, d, l):
		for each_file in d:
			print "%s/%s" % (d, each_file)

	os.path.walk("/", print_dirnames, '') 
	os.path.walk("/", list_directories, 'x') 

 ... these both work as I expected.  Of course I'm simply
 ignoring the 'args' argument in my functions.  I guess I 
 could write my functions to behave differently based on 
 the contents of the args.  (I can even imagine that as
 useful, though I don't need it for anything right now).

 At first I though that this "args" element was required
 to be a string.  However, I tried calling the os.path.walk() 
 with numeric and anonymous (literal) lists and saw no 
 differences in the output as a result of those.   I guess
 I could call it with a lambda function or a reference to a
 function if I wanted to.  I don't know Python well enough to
 show an example, but I could see where that might be useful.

 However, my question boils down to:

	Why is this third argument required in os.path.walk()?
	


 



From DanielK at aracnet.com  Thu Dec 28 23:25:31 2000
From: DanielK at aracnet.com (Daniel Klein)
Date: Thu, 28 Dec 2000 20:25:31 -0800
Subject: how do i check if a file exists with python?
References: <RlM26.98766$yR4.2639444@news1.rdc1.tx.home.com>
Message-ID: <phU26.31$KD3.4767@typhoon.aracnet.com>

How about something like

def fileExists(f):
    try:
        file = open(f)
    except IOError:
        exists = 0
    else:
        exists = 1
    return exists


Although this is not as elegant as the person who submitted the 'glob'
solution.

Dan



"Scott Hathaway" <slhath at home.com> wrote in message
news:RlM26.98766$yR4.2639444 at news1.rdc1.tx.home.com...
> I have searched Deja, looked in two pyton books, and searched the
python.org
> site trying to find out this simple question.
>
> Can someone please give me the call and the module it lives in?
>
> Thanks,
> Scott
>
>




From arturo_perez at wanadoo.es  Mon Dec 11 19:17:35 2000
From: arturo_perez at wanadoo.es (Arturo Pérez Mulas)
Date: Tue, 12 Dec 2000 00:17:35 GMT
Subject: Fault trees structure
References: <D_IY5.49$17.2785@m2newsread.uni2.es> <9128e902uo3@news2.newsguy.com>
Message-ID: <z8eZ5.249$YZ4.8868@m2newsread.uni2.es>

An introctury fault-tree, including a very brief introduction:

http://www.riskspectrum.com/docs/methods_2.htm

I have to admit that, at this point I have not been able to find any more
"introductory" pieces. However, I include below a schematic description
of the underlying structure for fault-trees; if anyone is interested and teh
discussion proceeds, I will post a more detailed description of the
'quantification' and operations of fault-trees, along with a description
of what they are really used for, and how (fault tree analysis).

I will try to describe the Fault Tree Structure with my own words: a
fault-tree is a tree-like structure (similar to the binary tree), developed
from a unique top-event or top-node. The main characteristics of a
Fault Tree are the following:

- They have only ONE top node

- Each node can have unlimited inputs (opposed to two allowed for binary
trees)

- Nodes can be of any of the following different types:

a) 'Gates' : are nodes that admit further inputs; there are different types
of gates,
            AND, OR, XOR, n/m, NOT but this is not important at this point,
since
            from the tree structure they are all similar (except type n/m
gates which
            must have exactly 'm' inputs)

b) 'Basic Events': or final nodes, which cannot have any more inputs; for
operation
            purposes, a basic event has a numeric 'probability' between 0
and 1, but I
            leave this for latter.

c) 'House Events': special events, similar to the Basic Events, but they do
not have
            a numerical probability, but a "on/off" status flag.

d) 'Links', or 'link gates', are nodes which actually point to other
fault-tree or other
            node in a different fault tree. They are used to crop large
fault-trees in pieces,
            or in larger structures, when several fault-trees are used, they
may share smaller
            pieces, so 'links' are used not to repeat the same structure in
several places.
            Note that usually a link points to the top event of another
fault tree, but
            this is not mandatory, and in some cases links pointing to
intermediate nodes
            (gates) in other fault tree can be found.

- Considering the previous description, the following 'properties' should be
considered
for the different types of nodes (along with a 'name' and 'description' for
all of them):

a) 'Gates':     - type (AND, OR, XOR...)
                    - inputs

b) 'Basic Events':    - probability (float between 0 to 1)
                                - other probability data (not important at
this point), such as uncertainity distributions

c) 'House Events':    - logical state (boolean switch)

d) 'Links':    - none

I have developed a very simple objects hierarchy for this, but I am not very
satisfied with it (I considered other
issues when creating that, based on the sources from which I will obtain the
fault trees, and the formats to which
I have to send them later). If anyone is finally interested, I will post
some info on the structure and we can discuss it.


I would greatly appreciate some discussion here, since this is the first
time I use OOP to model something!

Arturo.














From phd at phd.pp.ru  Thu Dec 28 09:59:47 2000
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 28 Dec 2000 17:59:47 +0300 (MSK)
Subject: PEP 222 draft
In-Reply-To: <3A4B1EA2.E15B4FB7@stroeder.com>
Message-ID: <Pine.LNX.4.30.0012281756290.2265-100000@fep132.fep.ru>

Hi!

On Thu, 28 Dec 2000, Michael [iso-8859-1] Str?der wrote:
> tlownds at my-deja.com wrote:
> >
> > In article <slrn947bj1.1rs.amk at 207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com>,
> >   akuchlin at mems-exchange.org wrote:
> > >     Utility function: build a query string from a list of 2-tuples
> > >
> >
> > Perhaps it would be better to extend urllib.urlencode to handle this
> > case. I think it should be extended to handle lists as values as well:
> >
> > >>> urlencode({'x': ['a', 'b']})
> > 'x=%5b%27a%27,+%27b%27%5d'
> >
> > would be:
> >
> > 'x=a&x=b'
>
> But you can do that with a list of 2-tuples either. Note: The list
> order of dictionary keys is normally not preserved.

   Not so long ago I wrote such function and sent it to Guido (it was the
time before Python 1.6) and GvR rejected my code saying something like "it
is rare need for someone, write it in your own library".
   No proble, I put it in my libraray. Here is the code. Please note, it is
perfectly compatible with current urllib.urlencode:

# obj must be either dict or list of tuples [(k1, v1), (k2, v2), ...] or list of lists [[k1, v1], [k2, v2], ...]
def urlencode(obj):
     l = []
     if type(obj) == type({}):
         obj = obj.items()

     for k, v in obj:
         k = quote_plus(str(k))
         v = quote_plus(str(v))
         l.append(k + '=' + v)

     return string.join(l, '&')

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




From glenfant at equod.com.nospam  Wed Dec  6 12:59:01 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Wed, 6 Dec 2000 18:59:01 +0100
Subject: [OT] XML introduction
References: <d8qX5.227$rZ6.6363@m2newsread.uni2.es>
Message-ID: <90luha$s4m$1@reader1.imaginet.fr>

Arturo,

You'll get better answer posting your question to newsgroup "comp.text.xml".


"Arturo P?rez Mulas" <arturo_perez at wanadoo.es> a ?crit dans le message news:
d8qX5.227$rZ6.6363 at m2newsread.uni2.es...
Hello,

this is kind of off-topic, but not quite. I cannot find any comprehensive
introduction
to XML that gives real information on what is it used for. As far as I could
grasp,
it allows to define structures to store data cretaing your html-like tags,
and rules
to format them on a xml-compliant browser... but I keep reading things about
"developing web sites or applications with xml" and then I get lost.

Anyone any pointer or hint, please?

Thanks

Arturo P?rez



From fredrik at effbot.org  Mon Dec 11 04:16:34 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Mon, 11 Dec 2000 09:16:34 GMT
Subject: Multiple Threads & IRC
References: <3A3415E0.1080607@bellsouth.net>
Message-ID: <SX0Z5.3104$Kd1.386434@newsb.telia.net>

"Wilbur" wrote:
> The server sends data in packets, and it sends whenever it can.  How can
> I get my client to be able to receive data from the server without
> knowing when it will send and what amount of data will be sent?  I only
> know how to tell Python to wait until it receives, say, 1024 bytes.  If
> the server doesn't send any more, how can I tell Python to stop waiting?

http://www.python.org/doc/current/lib/module-asyncore.html

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->





From lizard at mrlizard.com  Sat Dec  2 02:00:37 2000
From: lizard at mrlizard.com (Lizard)
Date: Fri, 01 Dec 2000 23:00:37 -0800
Subject: [OT] Apple IIe help please
In-Reply-To: <20001201212507.A14226@westsidecnc.rh.rit.edu>
Message-ID: <5.0.0.25.2.20001201225938.053fcde0@mrlizard.com>

At 06:25 PM 12/1/2000, D-Man wrote:

>Hi all.  I apologize for the off topic-ness of the message and the 
>cross-post.
>I have a friend who has a program she likes (written in BASIC) on an Apple 
>IIe.
>I have a way to get into the code and list it on the screen.  What I am 
>looking
>for is someone who has had some experience using an Apple IIe who can tell me
>how I can get the listing to be redirected to the printer port.  Also if it is
>possible to get a copy of the program file onto a disk for an IBM compatible
>comptuer (Windows or Linux) that would be great.

Gods, THAT brings back memories! Something like "PRINT #1" or the like? I 
know you specify a port..."IN #1" turned on the printer? Maybe? I'm sorry, 
I'm dredging up memories from 20+ years ago...




From python9999 at my-deja.com  Mon Dec 18 03:25:40 2000
From: python9999 at my-deja.com (python9999 at my-deja.com)
Date: Mon, 18 Dec 2000 08:25:40 GMT
Subject: How to enable html links using XmHTML module??
Message-ID: <91khm2$7as$1@nnrp1.deja.com>

hi..
i am writing an application in pygtk, in which i need to display HTML
screens . i could display them using python's XmHTML module, but i am
not able to understand how to enable the links in the screens. i am
displaying the HTML screens on a gtk frame and i know that for enabling
link it needs a browser...

Can we use the module available for enabling the links??

Are there any functions available for doing so ?

regards,
Priya


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


From rachelp_ny at yahoo.com  Sat Dec 16 18:38:29 2000
From: rachelp_ny at yahoo.com (Rachel)
Date: Sat, 16 Dec 2000 18:38:29 -0500
Subject: latest python release for solaris
Message-ID: <3a3bff01_4@news1.prserv.net>

Is anyone running the most current release?  We're running 1.5.2 -- it's
been pretty stable.  We're having some trouble with the code for SSL
connections -- it works fine with Perl, but we've had some problems with
Python.  My understanding is that the latest release has native SSL support
in it. I wondering if upgrading is worthwhile....




From nicmila at idoox.com  Mon Dec 11 02:12:51 2000
From: nicmila at idoox.com (Miloslav Nic)
Date: Mon, 11 Dec 2000 08:12:51 +0100
Subject: pattern matching
References: <DZQY5.7555$KY1.20484@news1.rivrw1.nsw.optushome.com.au>
Message-ID: <3A347E73.B00368A3@idoox.com>

Zvon Regexp reference was generated with Python using
re module:

http://zvon.org/other/reReference/Output/index.html

You can download sources at :
http://www.zvon.org/index.php?nav_id=3

re module is definitively the way to go

deadmeat wrote:
> 
> Is there another standard library I can use for pattern matching, other than
> fnmatch? I need to be able to check for metacharacters, fnmatch doesn't
> allow me to escape them...

-- 
******************************************
<firstName> Miloslav </firstName>    
<surname>   Nic      </surname>     

<mail>    nicmila at idoox.com    </mail>   
<support> http://www.zvon.org  </support>
<zvonMailingList> 
    http://www.zvon.org/index.php?nav_id=4 
</zvonMailingList>


From be at et.dtu.dk  Wed Dec 20 04:51:55 2000
From: be at et.dtu.dk (Brian Elmegaard)
Date: Wed, 20 Dec 2000 10:51:55 +0100
Subject: Python's scope rules
Message-ID: <3A40813B.C04D21C@et.dtu.dk>

Hi,

I am not a computer scientist, just an engineer knowing how to program in
Fortran, Pascal,... So I after reading some of the python material on the web, I
decided to give it a try and is now using the language some, finding it fun and
easy.

On usenet I have now learned from skilled scientists, that Python has some
deficiencies regarding scope rules, and I am not capable of telling them why it
has not (or that it has). 

In Aaron Watters 'The wwww of python' I have read that Python uses lexical
scoping with a convenient modularity. But, other people tell me it does not.
Some say it uses dynamic scoping, some say it uses it own special
'local-global-builtin' scope. What is right?

The above mainly is a theoretical question. A more practical example which I
agree seems a bit odd is:

Python 1.5.2 (#1, Mar  9 2000, 17:40:34)  [GCC 2.95.2 19991024 (release)] on hp-
uxB
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> x=24
>>> def foo():
...     y=2
...     print x+y
...
>>> def bar():
...     x=11
...     foo()
...
>>> bar()
26
>>>

Well, I can live with it. In Fortran I would always put all the variables to be
passed in the parameter list. Other people say that Fortran is not easy, and
this example shows how stupid python is. It does?

Hopefully not, but probably, this is a faq, but I have not been able to find
anything on the matter.

-- 
Brian Elmegaard (be at et.dtu.dk) http://www.et.dtu.dk/energysystems 
Dept. of Energy Engineering, Technical University of Denmark, 
Phone +45 4525 4169  Fax +45 4593 5215
:) http://www.rk-speed.dk http://fiduso.dk http://sunsite.auc.dk/dk-tug


From stevee at recombinant.demon.co.uk  Wed Dec  6 15:28:50 2000
From: stevee at recombinant.demon.co.uk (Stephen D Evans)
Date: Wed, 06 Dec 2000 20:28:50 +0000
Subject: rotate() in Tkinter
References: <5yuX5.761950$8u4.13165276@news1.rdc1.bc.home.com>
Message-ID: <3A2EA182.51F74B7@recombinant.demon.co.uk>

To rotate an oval convert the oval to a polygon and rotate that. Polygons are
easy to rotate, but this has to be done to the points before calling
create_polygon().

Referring to Tk documentation - The canvas does not support rotation.
'Practical Programming in Tcl and Tk', 3rd edition, 2000, by Brent B. Welsh,
publisher Prentice Hall PTR

This is a simple example.

(There are better methods of creating the points for an oval. Try fast
generation of ellipsoids in 'Graphics Gems V', 1995, edited by Alan W Paeth,
publisher AP Professional)

#
#
import math
import Tkinter


def poly_oval(x0,y0, x1,y1, steps=20, rotation=0):
    """return an oval as coordinates suitable for create_polygon"""

    # x0,y0,x1,y1 are as create_oval

    # rotation is in degrees anti-clockwise, convert to radians
    rotation = rotation * math.pi / 180.0

    # major and minor axes
    a = (x1 - x0) / 2.0
    b = (y1 - y0) / 2.0

    # center
    xc = x0 + a
    yc = y0 + b

    point_list = []

    # create the oval as a list of points
    for i in range(steps):

        # Calculate the angle for this step
        # 360 degrees == 2 pi radians
        theta = (math.pi * 2) * (float(i) / steps)

        x1 = a * math.cos(theta)
        y1 = b * math.sin(theta)

        # rotate x, y
        x = (x1 * math.cos(rotation)) + (y1 * math.sin(rotation))
        y = (y1 * math.cos(rotation)) - (x1 * math.sin(rotation))

        point_list.append(round(x + xc))
        point_list.append(round(y + yc))

    return point_list



import Tkinter
root = Tkinter.Tk()
canvas = Tkinter.Canvas(root, width=400, height=400)

dict = {}
dict['outline'] = 'black'
dict['fill']   = 'yellow'
dict['smooth'] = 'true'

# use a polygon to draw an oval rotated 30 degrees anti-clockwise
apply(canvas.create_polygon, tuple(poly_oval(40,40, 200,300, rotation=30)),
dict)

canvas.pack()
root.mainloop()

-- Stephen D Evans


Dethe Elza wrote:

> I'm trying to create a complex cursor in Tkinter.  I want to have an image
> which I can rotate, change the color of, and move about.  I can build it as
> a vector image, but I use ovals.  I haven't found any way to rotate images,
> vector drawings, or ovals in Tk.
>
> TIA
>
> --Dethe
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20001206/0006828b/attachment.html>

From aleaxit at yahoo.com  Thu Dec 28 12:08:30 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 28 Dec 2000 18:08:30 +0100
Subject: Basic MAPI
References: <91vfsb02cn6@news2.newsguy.com> <92fonu$auu$1@plutonium.compulink.co.uk>
Message-ID: <92ft0m029uo@news1.newsguy.com>

<gbreed at cix.compulink.co.uk> wrote in message
news:92fonu$auu$1 at plutonium.compulink.co.uk...
    [snip]
> > > >>> s=Dispatch("Mapi.Session")
> >
> > There is no creatable coclass (that I ever heard of) with progId
> > 'Mapi.Session'.
>
> There most certainly is.  It comes with Outlook, among other things, and
> I think lives in cdo.dll ("MAPI" is an abbreviation for "CDO" or
> something).  That Dispatch works fine on my machine, so I don't know
> what's going wrong.  Check that Outlook's installed.

Aha -- maybe we have a clue here.  I do have Outlook installed, but,
it's Outlook 2000.  "Migrating Applications to Microsoft Outlook 2000"
by Jim Groves (found on MSDN) does list among the migration concerns:
"Collaboration Data Objects (CDO) [...] now optional".  And indeed,
I have no cdo.dll on my system, nor it it anywhere in my registry.

This is what Groves has to say about CDO versus MAPI:

"""
Support for Collaboration Data Objects
Formerly known as OLE Messaging and Active Messaging, CDO is an object model
that gives
applications more direct access to the Messaging Application Programming
Interface (MAPI) used by Outlook to structure its data storage than does the
Outlook object model.
Programming with CDO is more flexible than MAPI because, unlike MAPI, CDO
does not
require C or C++ as the programming language. Instead, CDO can be accessed
through
Visual Basic and Active Server Pages (ASP), as well as Visual C++? and any
other
development tool that creates Component Object Model (COM) objects.
"""

which I interpret as: CDO offers dual-interface COM access, so it can be
scripted from VB or anything else (Python included), while MAPI is a C/C++
only API (making it peculiar that a CDO object could possibly be named
'MAPI' something, but, whatever).

So, if Outlook 2000 is what you have, still quoting Groves:

"""
If users in your organization have already installed Outlook 2000, it is
easy for each
of them to install the CDO support.

To install CDO support
Click Start, point to Settings, and then click Control Panel.

Double-click the Add/Remove Programs icon.

In the list, click Microsoft Office 2000 or Microsoft Outlook 2000, and then
click Add/Remove.

Click Add or Remove Features.

Click the plus sign (+) next to Microsoft Outlook for Windows to open it.

Click Collaboration Data Objects, and then click Run from My Computer.
"""


Haven't tried this, so I don't know if this will indeed bring to life
a cdo.dll implementing a cocreatable "Mapi.Session" on your computer,
but, if Outlook2000 is what you're trying to work with, it sure seems
worth trying.  (There are other CDO implementations, "CDO for NTS" which
is said to use SMTP and run on NT Server only, and "CDO for Exchange
2000", but they apparently use different prog-id strings to let you
create their respective 'Session' objects).


Alex





From fgeiger at datec.at  Wed Dec 27 01:47:19 2000
From: fgeiger at datec.at (Franz GEIGER)
Date: Wed, 27 Dec 2000 07:47:19 +0100
Subject: A Q concerning map (a comparison to Maple V)
References: <921sil$moe$1@newsreaderg1.core.theplanet.net> <mailman.977570347.5588.python-list@python.org>
Message-ID: <92c41g$n1p$4@newsreaderg1.core.theplanet.net>

>   map(replace, ["abc", "cde"], ["c"]*2, ["C"]*2)
>

That's it, but doesn't look that beautiful.

> whereas with you proposal this would give you:
>
>   ["abc", "CDe"]

No,  with my proposal nothing would happen, because there is no "cd"
anywhere. Be aware, the list is the list replace has to be applied on, not
the strings within it.

> Excuse me if I misinterpreted your proposal, but if the

Excused ;-)


Best regards
Franz GEIGER


"Peter Schneider-Kamp" <peter at schneider-kamp.de> wrote in message
news:mailman.977570347.5588.python-list at python.org...
> Franz GEIGER schrieb:
> >
> > map(replace, ["abc", "cde"], "c", "C");
> > instead of having to create a lambda function.
> > Wouldn't it make sense to have that in Python too? Do there come
> > more such opportunities into your mind?
>
> If I get you right you intend this to mean:
>
>   map(replace, ["abc", "cde"], ["c"]*2, ["C"]*2)
>
> So basically your proposal is to reuse map parameters
> which are no sequences in every iteration over the
> parameters who are, s.t.
>
>   map(f, a_1, .., a_n, b_1, .., b_m)
>
> for f being a (m+n)-ary function, a_1 to a_n being sequence
> parameters and b_1 to b_m non-sequence parameters would yield
>
>   maxlen = max(map(len,[a_1, .., a_n]))
>   map(f, a_1, .., a_n, [b_1]*maxlen, .., [b_m]*maxlen)
>
> But how would you decide if "abc" is a sequence parameter
> or not? To keep your example, what would
>
>   map(replace, ["abc", "cde"], "cd", "CD")
>
> mean? At the moment this gives you:
>
>   ["abC", "cDe"]
>
> whereas with you proposal this would give you:
>
>   ["abc", "CDe"]
>
> Excuse me if I misinterpreted your proposal, but if the
> above is what you mean, I think this a Bad Idea(tm).
>
> merry-xmas-shopping-ly y'rs
> Peter
>










From tlownds at my-deja.com  Sun Dec 24 12:17:39 2000
From: tlownds at my-deja.com (tlownds at my-deja.com)
Date: Sun, 24 Dec 2000 17:17:39 GMT
Subject: PEP 222 draft
References: <slrn947bj1.1rs.amk@207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com>
Message-ID: <925b3g$5pl$1@nnrp1.deja.com>

In article <slrn947bj1.1rs.amk at 207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com>,
  akuchlin at mems-exchange.org wrote:
>     Utility function: build a query string from a list of 2-tuples
>

Perhaps it would be better to extend urllib.urlencode to handle this case. I
think it should be extended to handle lists as values as well:

>>> urlencode({'x': ['a', 'b']})
'x=%5b%27a%27,+%27b%27%5d'

would be:

'x=a&x=b'

>     The cgi.py module would be deprecated.  (XXX A new module or
>     package name hasn't been chosen yet: 'web'?  'cgilib'?)
>

cgilib gets my vote.

>     * An HTML generation module is not part of this PEP.  Several such
>       modules exist, ranging from HTMLgen's purely programming
>       interface to ASP-inspired simple templating to DTML's complex
>       templating.  There's no indication of which templating module to
>       enshrine in the standard library, and that probably means that
>       no module should be so chosen.
>

There are more ASP-like modules out there than other types...

>     Parsing functions: carry over most of the parse* functions from
>     cgi.py
>

It would be really nice to change the main parse-form-input function to
return a dictionary of plain strings instead of the FieldStorage instances.
When a file is uploaded, a FieldStorage instance for that value makes sense.
You've suggested this before, I may have missed it in your PEP.

>
>     # Module-level function
>     def wrapper(func, logfile=sys.stderr):
>         """
>         Calls the function 'func', passing it the arguments
>         (request, response, logfile).  Exceptions are trapped and
>         sent to the file 'logfile'.
>         """
>         # This wrapper will detect if it's being called from the command-line,
>         # and if so, it will run in a debugging mode; name=value pairs
>         # can be entered on standard input to set field values.
>         # (XXX how to do file uploads in this syntax?)
>

Regarding debugging: A command-line debugging ability is very useful! But why
tie that feature to the wrapper feature? You can put it in the
parse-form-input function. Also, allow name=value pairs as command-line
arguments. name=value pairs would still be read even with command-line
arguments, unless a final command-line argument "go" is present.

One debugging feature that would be very useful in a wrapper, is a way to
start pdb either right away or at an exception.

Your wrapper passes (request, response, logfile) - why not pass an object
with .request, .response, and .logfile attributes. Using a tuple now makes it
harder to extend later. A cgi.log-like method on that "context" object would
be useful too!

-Tony Lownds



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


From steve at serf.org  Fri Dec 15 16:32:30 2000
From: steve at serf.org (A SERFer)
Date: 15 Dec 2000 21:32:30 +0000
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <3a36c85e.4409136@News.CIS.DFN.DE> <owen_nospam-1312001131590001@cs204-90.fsmodem.washington.edu> <3A37DD31.C9A5F18C@pacbell.net> <91bhlh$c64u$1@nntp6.u.washington.edu> <3A395197.CB2974D5@home.com> <L_r_5.7453$n51.136546@e420r-atl1.usenetserver.com>
Message-ID: <steve976915950@serf.org>

"Steve Holden" <sholden at holdenweb.com> writes:

>James A. Robertson <jarober at home.com> wrote in message
>news:3A395197.CB2974D5 at home.com...
>> "Russell E. Owen" wrote:
>> >
>>
>> >
>> > My point is, that Cincom's pricing policies look self-destructive to me.
>> > I cannot in good conscience make my observatory rely on a language that
>> > I fear will continue to be a marginal player, at best.
>> >
>>
>> How so?  We went way over our sales targets last year, and things are
>> looking pretty good for the upcoming year.  Cincom has devoted a lot of
>> resource into Smalltalk; we are committed to it.
>>
>Unfortunately, sales targets are arbitrary numbers.  What evidence do you
>have that revenues might have been increased if you had reduced the price?
>I'm sure you can appreciate that if you sold for one-third the price and
>made six times as many sales, profits would be up.  Most sales organizations
>I've worked with don't have a model which shows profits as a function of
>price.
>
>The fact that I've seen two postings on this thread complaining that your
>prices are too high suggests sales would increase with lower pricing.  Of
>course the effect on profit is unknown.
>
>The fact that an academic version is available redeems your policy somewhat,
>but I suspect your target market may be nervous about the company's survival
>because it feels your pricing policies are unrealistic.

Indeed we are!  We keep hearing a circular argument:
1. Cincom are committed to Smalltalk in the long term.
2. We must charge these prices if Smalltalk is to survive within Cincom.

Its like a sort of quantum mechanical uncertainty thing... each statement
on its own makes sense, but put together, they are meaningless.
>
>> > I'm also reluctant to write an application with a "freebie" splash
>> > screen that comes up every time. And I'm reluctant to use free versions
>> > of commercial stuff we cannot afford, because it doesn't support the
>> > company and because their pricing policies are risky -- if they dump the
>> > free version we're in deep trouble.
>>
>> If an asteroid hits the planet, you're in trouble too.  Nothing is
>> certain.  If Sun's hardware sales slip at some point, Java is uncertain
>> as well.  They aren't making money on Java itself.
>>
>Well, that's probably why this discussion is being held on the Python group!
>Clearly Sun's demise would impact a lot of people and upset a lot of plans,
>but they do have a certain financial momentum which means a crash would
>probably be heralded by several years of poor performance.

Having done some research into Java's history, I would guess that Sun are
in fact making money from Java:  one of the motivations was to provide
a common platform and a common set of APIs for internal Sun use as well
as a product.  Even if Java had only succeeded as an internal tool, it
would have cut costs, and so made money for Sun.

>
>Also, Java has a number of compatible compilers, so even if Sun died the
>language would conceivably survive.  It sounds as though the degree of
>SmallTalk compatibility is lower, effectively making the Cincom
>implementation proprietary.

There is no doubt the language would survive.  Its now one of IBM's
primary tools for internal use, and IBM have written their own VMs
(such as the UVM used in VisualAge), and they are developing Java byte
code emulation tools (see the DAISY project).

>
>Sun's revenues are measured in the billions - they became a $1bn company in
>year six, if I remember rightly.  Whereas Cincom's web site says "And during
>the past three decades, our annual sales have grown to $185 million".  Not
>quite so impressive.
>
>Is there NOBODY in the company who suspects this slow growth might be due to
>overpricing?  I'm not denigrating your figures in absolute terms (I achieved
>much less in my own company), but other companies have grown much larger
>revenues in the same period, and thereby become a better long-term bet.
>
>> >
>> > If there were compatible implementations I'd not worry so much about the
>> > stability of one company, but I'm reluctant to hitch our star to Cincom.
>> >
>> > Smalltalk is my favorite language. However, I am the only one at the
>> > observatory who knows it and feels it is worthwhile. As I fight Python's
>> > disgusting GUI I keep wavering, and I may yet give up and switch to
>> > Smalltalk. The worst they can do is fire me. But I love my job.
>> >
>> > -- Russell
>>
>> --
>> James A. Robertson
>> Technical Product Manager (Smalltalk), Cincom
>> jarober at mail.com
>>
>> <Talk Small and Carry a Big Class Library>
>>
>regards
> Steve
>
>
>


From rturpin at my-deja.com  Fri Dec 29 12:42:34 2000
From: rturpin at my-deja.com (rturpin at my-deja.com)
Date: Fri, 29 Dec 2000 17:42:34 GMT
Subject: Python vs. PHP (& Java?)
References: <69BC4C6AE83ED311BC9400805FA7B14AF84D2C@il93exp01.css.mot.com> <mailman.978018791.20357.python-list@python.org> <92fv4a$l2l$1@nnrp1.deja.com> <92gghv02iq1@news2.newsguy.com>
Message-ID: <92iie9$lg3$1@nnrp1.deja.com>

I wrote:
>> In contrast, texts in computer languages are more studied
>> as artifacts. "How did this work?" "Why did it stop
>> working?" "What did the original authors intend?" They,
>> the rascals, are no longer around to explain, take credit,
>> or defend their honor. This kind of comprehension requires
>> a high level of competency in the language.

In article <92gghv02iq1 at news2.newsguy.com>,
  "Alex Martelli" <aleaxit at yahoo.com> wrote:
> Are you implying that this _doesn't_ happen a lot for
> many written texts in natural languages...?  Literary
> analysis is all about such work, after all -- and courts
> and lawyers do a lot of that on laws and constitution
> articles.

Of course. And in both cases, it requires a proficiency
beyond normal, either with language generally, as in the
case of literary critics, or with a technically specialized
variant, as with lawyers. I am not trying to draw too sharp
a distinction between natural language and computer language.
This would be foolish: computer languages ARE just
technically specialized parts of natural language. All I
point out is that the manner in which they are used requires
proficiency and scope that most casual communication does
not. The same is true for other technically specialized
sublingua, including hoary examples such as law and medicine.

Turpin:
>> Knowing one-tenth of Perl is not enough to understand
>> most Perl scripts.

Martell:
> Of course not, but, it IS enough to _write_ a lot of
> somewhat useful scripts ..

Your use of a computer language is fairly limited if
you can only WRITE it. Software engineering requires
the ability to READ. And not just what you once wrote,
but what others have written, including others who have
been using it for years and know all the idioms and
tricks. Even though they are long gone. This is why
software engineering is more like literary analysis or
the study of Constitutional law, than it is like
an after dinner conversation. The interesting and hard
productivity question is not how long it takes a
beginner to complete a task, but how much effort it
requires a group of professionals to maintain a sizable
product, over its lifecycle. (This isn't meant as
criticism of  those who want to use Python for teaching
or personal use. It is a very good language for both
purposes. But my interest and most of my comments are
from the viewpoint of professional use.)

> The complexity of natural language springs in good
> part from the _adaptive_ characteristics (pace Gould...)
> of ambiguity in its use; words serve to hide, as much as
> they serve to show.

That's certainly part of it. There are many social
situations where it is advantageous to say something, but
only so much. It's also the case that in most social
interaction, communication is going on at several levels
simultaneously. A rich language that admits multiple
meaning and implication is very useful. If you want to
look for adaptation, sexual selection may have more to do
with this than natural selection. A silver tongue will
not save you from the lion or bag the deer, but it will
win a heart.

> Nothing "is EXACTLY like" anything else (Korzibsky would
> insist on such firm avoidance of "the IS of identity", but,
> as he doesn't seem to be following this newsgroup regularly,
> I'll have to play his role, I guess). ..

I'm surprised no one brought up the good Count when Clinton
uttered his infamous question.

> Alternatively, of course, great complexity in implementation
> MIGHT be spent towards such targets as optimization, as
> you were interestingly arguing Python should be doing now.

Yep. Keep in mind that only the implementation team has to
tackle the complexity of optimization, whereas the world has
to tackle the complexity of language and semantics.

> But, even if a strongly-optimizing language processor,
> being more fragile, exists more precariously, it IS still
> good to have a simple, clear, transparent "reference
> implementation" to fall back to...

Absolute agreement. More importantly, we poor programmers
need a simple, clear, transparent conceptual model, when
we're trying to puzzle out why some construct behaved as
it did.

> [Perl's] design goals DID apparently include rich,
> exhuberant, redundant abundance -- it seems to have
> matched them closely. ..

That is a most unfortunate goal.

> .. (from your excellent diagnostic-manual on quack
> theories) ..

It's amazing how one post in sci.med fifteen years ago has
taken on a life of its own. I would rather you criticize
my recent proposal for caching attribute and method
references.

BTW, didn't we meet once?

Regards,
Russell


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


From timd at macquarie.com.au  Tue Dec 12 21:38:23 2000
From: timd at macquarie.com.au (Timothy Docker)
Date: 13 Dec 2000 13:38:23 +1100
Subject: Issues with the "McMillan" installer
Message-ID: <m4in1e1f3kw.fsf@macquarie.com.au>


I'm trying to get the "McMillan" installer working on an application
that I have written, and am struggling with several problems.

The application uses python-2.0, wxWindows, and omniORB. 

My first issue is that builtin-modules are not getting resolved
correctly. If I run Simple.py on the following script (test.py)

| import sys, imputil
|
| print sys.path
|
| import time
| import thread
|
| print "hello"

and then use the install script to install a clean directory, when I
run the created executable I get...

| C:\TEMP\rt>test
| ['C:\\TEMP\\rt\\', 'C:\\TEMP', 'C:\\Python20\\Pythonwin', <archive_rt.ZlibArchive instance at 0042FCCC>]
| Traceback (most recent call last):
|   File "<string>", line 5, in ?
|   File "c:\python20\lib\imputil.py", line 91, in _import_hook
|     raise ImportError, 'No module named ' + fqname
| ImportError: No module named time
| 
| C:\TEMP\rt>

Needless to say the original script runs fine. The path as finally set
up looks a bit sus to me. (The Builder.log is attached below). I got
around this problem by changing the script to look like...

| import sys, imputil
| 
| print sys.path
| 
| sys.path.insert(0, imputil.BuiltinImporter())
| 
| import time
| import thread
| 
| print "hello"

But presumably this shouldn't be needed - what's going on here?

After working around this, I then had a problem with the wxwindows
library. The wxwindows library has its c extension (wxc.pyd) inside
the package. After building and running the installer, I end up with 
the file wxwindows.wxc.pyd in the installation directory. This isn't found
correctly when an attempt is made to import it. Renaming it to
'wxc.pyd' made the problem go away, but again a hack. Is this somehow
related to the problem above?

Any tips/pointers would be much appreciated.

Tim Docker

----------------------------------------------------------------------
---- PYZTarget: installzlib.pyz -----
{'__name__': 'installzlib.pyz',
 '_dependencies': [(imputil, c:\python20\lib\imputil.pyc, m), (zlib, q:\pc\zlib.pyd, b), (archive_rt, Q:\pc\installer\support\archive_rt.py, m)],
 'bindepends': [],
 'cfg': <ConfigParser.ConfigParser instance at 00844B8C>,
 'children': [],
 'debug': 0,
 'dependencies': ['installutils'],
 'destdir': '',
 'directories': [],
 'excludes': ['dospath', 'posixpath', 'macpath'],
 'expatterns': [],
 'exstdlib': 0,
 'extypes': [],
 'icon': '',
 'includes': ['installutils'],
 'includetk': 0,
 'misc': [],
 'name': 'installzlib.pyz',
 'packages': [],
 'pathprefix': ['Q:\\pc',
                'Q:\\pc\\installer\\support'],
 'script': [],
 'support': 1,
 'toc': [],
 'trees': [],
 'userunw': 0,
 'zlib': []}
Dependencies of installutils that couldn't be traced:
{'MACFS': {'tempfile': None},
 'ce': {'os': None},
 'dos': {'os': None},
 'mac': {'os': None},
 'macfs': {'tempfile': None},
 'os2': {'os': None},
 'posix': {'os': None},
 'pwd': {'posixpath': None}}
lTOC after expanding 'depends':
[('archive_rt',
  'q:\\pc\\installer\\support\\archive_rt.py',
  'm'),
 ('os', 'c:\\python20\\lib\\os.py', 'm'),
 ('sre_parse', 'c:\\python20\\lib\\sre_parse.py', 'm'),
 ('ntpath', 'c:\\python20\\lib\\ntpath.py', 'm'),
 ('carchive_rt',
  'q:\\pc\\installer\\support\\carchive_rt.py',
  'm'),
 ('re', 'c:\\python20\\lib\\re.py', 'm'),
 ('sre_constants', 'c:\\python20\\lib\\sre_constants.py', 'm'),
 ('UserDict', 'c:\\python20\\lib\\UserDict.py', 'm'),
 ('tempfile', 'c:\\python20\\lib\\tempfile.py', 'm'),
 ('sre', 'c:\\python20\\lib\\sre.py', 'm'),
 ('posixpath', 'c:\\python20\\lib\\posixpath.py', 'm'),
 ('types', 'c:\\python20\\lib\\types.py', 'm'),
 ('sre_compile', 'c:\\python20\\lib\\sre_compile.py', 'm'),
 ('copy', 'c:\\python20\\lib\\copy.py', 'm'),
 ('popen2', 'c:\\python20\\lib\\popen2.py', 'm'),
 ('imputil', 'c:\\python20\\lib\\imputil.py', 'm'),
 ('repr', 'c:\\python20\\lib\\repr.py', 'm'),
 ('dospath', 'c:\\python20\\lib\\dospath.py', 'm'),
 ('macpath', 'c:\\python20\\lib\\macpath.py', 'm'),
 ('string', 'c:\\python20\\lib\\string.py', 'm'),
 ('copy_reg', 'c:\\python20\\lib\\copy_reg.py', 'm'),
 ('stat', 'c:\\python20\\lib\\stat.py', 'm'),
 ('pre', 'c:\\python20\\lib\\pre.py', 'm')]
lTOC after includes, dir, pkgs:
[('archive_rt',
  'q:\\pc\\installer\\support\\archive_rt.py',
  'm'),
 ('os', 'c:\\python20\\lib\\os.py', 'm'),
 ('sre_parse', 'c:\\python20\\lib\\sre_parse.py', 'm'),
 ('ntpath', 'c:\\python20\\lib\\ntpath.py', 'm'),
 ('carchive_rt',
  'q:\\pc\\installer\\support\\carchive_rt.py',
  'm'),
 ('re', 'c:\\python20\\lib\\re.py', 'm'),
 ('sre_constants', 'c:\\python20\\lib\\sre_constants.py', 'm'),
 ('UserDict', 'c:\\python20\\lib\\UserDict.py', 'm'),
 ('tempfile', 'c:\\python20\\lib\\tempfile.py', 'm'),
 ('sre', 'c:\\python20\\lib\\sre.py', 'm'),
 ('posixpath', 'c:\\python20\\lib\\posixpath.py', 'm'),
 ('types', 'c:\\python20\\lib\\types.py', 'm'),
 ('sre_compile', 'c:\\python20\\lib\\sre_compile.py', 'm'),
 ('copy', 'c:\\python20\\lib\\copy.py', 'm'),
 ('popen2', 'c:\\python20\\lib\\popen2.py', 'm'),
 ('imputil', 'c:\\python20\\lib\\imputil.py', 'm'),
 ('repr', 'c:\\python20\\lib\\repr.py', 'm'),
 ('dospath', 'c:\\python20\\lib\\dospath.py', 'm'),
 ('macpath', 'c:\\python20\\lib\\macpath.py', 'm'),
 ('string', 'c:\\python20\\lib\\string.py', 'm'),
 ('copy_reg', 'c:\\python20\\lib\\copy_reg.py', 'm'),
 ('stat', 'c:\\python20\\lib\\stat.py', 'm'),
 ('pre', 'c:\\python20\\lib\\pre.py', 'm'),
 ('installutils',
  'Q:\\pc\\installer\\support\\installutils.py',
  'm')]
Applying the following filters:
[<ModFilter {'dospath': 1}>,
 <ModFilter {'posixpath': 1}>,
 <ModFilter {'macpath': 1}>,
 <ModFilter {'archive_rt': 1}>,
 <ModFilter {'imputil': 1}>,
 <ModFilter {'imputil': 1}>,
 <ExtFilter {'.pyo': 1, '.pyc': 1, '.py': 1}>]
Final Table of Contents for installzlib.pyz:
[('os', 'c:\\python20\\lib\\os.py', 'm'),
 ('sre_parse', 'c:\\python20\\lib\\sre_parse.py', 'm'),
 ('ntpath', 'c:\\python20\\lib\\ntpath.py', 'm'),
 ('carchive_rt',
  'q:\\pc\\installer\\support\\carchive_rt.py',
  'm'),
 ('re', 'c:\\python20\\lib\\re.py', 'm'),
 ('sre_constants', 'c:\\python20\\lib\\sre_constants.py', 'm'),
 ('UserDict', 'c:\\python20\\lib\\UserDict.py', 'm'),
 ('tempfile', 'c:\\python20\\lib\\tempfile.py', 'm'),
 ('sre', 'c:\\python20\\lib\\sre.py', 'm'),
 ('types', 'c:\\python20\\lib\\types.py', 'm'),
 ('sre_compile', 'c:\\python20\\lib\\sre_compile.py', 'm'),
 ('copy', 'c:\\python20\\lib\\copy.py', 'm'),
 ('popen2', 'c:\\python20\\lib\\popen2.py', 'm'),
 ('repr', 'c:\\python20\\lib\\repr.py', 'm'),
 ('string', 'c:\\python20\\lib\\string.py', 'm'),
 ('copy_reg', 'c:\\python20\\lib\\copy_reg.py', 'm'),
 ('stat', 'c:\\python20\\lib\\stat.py', 'm'),
 ('pre', 'c:\\python20\\lib\\pre.py', 'm'),
 ('installutils',
  'Q:\\pc\\installer\\support\\installutils.py',
  'm')]
---- PYZTarget: test.pyz -----
{'__name__': 'test.pyz',
 '_dependencies': [(imputil, c:\python20\lib\imputil.pyc, m), (zlib, q:\pc\zlib.pyd, b), (archive_rt, Q:\pc\installer\support\archive_rt.py, m)],
 'bindepends': [],
 'cfg': <ConfigParser.ConfigParser instance at 00844B8C>,
 'children': [],
 'debug': 0,
 'dependencies': ['test.py'],
 'destdir': '',
 'directories': [],
 'excludes': ['dospath', 'posixpath', 'macpath'],
 'expatterns': [],
 'exstdlib': 0,
 'extypes': [],
 'icon': '',
 'includes': [],
 'includetk': 0,
 'misc': [],
 'name': 'test.pyz',
 'packages': [],
 'pathprefix': ['Q:\\pc',
                'Q:\\pc\\installer\\support'],
 'script': [],
 'support': 1,
 'toc': [],
 'trees': [],
 'userunw': 0,
 'zlib': []}
Dependencies of test.py that couldn't be traced:
{'dos': {'imputil': None},
 'mac': {'imputil': None},
 'os2': {'imputil': None},
 'posix': {'imputil': None}}
lTOC after expanding 'depends':
[('imputil', 'c:\\python20\\lib\\imputil.py', 'm')]
lTOC after includes, dir, pkgs:
[('imputil', 'c:\\python20\\lib\\imputil.py', 'm')]
Applying the following filters:
[<ModFilter {'dospath': 1}>,
 <ModFilter {'posixpath': 1}>,
 <ModFilter {'macpath': 1}>,
 <ModFilter {'archive_rt': 1}>,
 <ModFilter {'imputil': 1}>,
 <ModFilter {'imputil': 1}>,
 <ExtFilter {'.pyo': 1, '.pyc': 1, '.py': 1}>]
Final Table of Contents for test.pyz:
[]
---- ExeTarget: test.exe -----
{'__name__': 'test.exe',
 '_dependencies': [(imputil, c:\python20\lib\imputil.py, m), (archive_rt, q:\pc\installer\support\archive_rt.py, m), (zlib, q:\pc\zlib.pyd, b), (python20.dll, C:\WINNT\System32\python20.dll, b), (carchive_rt, Q:\pc\installer\support\carchive_rt.py, m)],
 'bindepends': [],
 'cfg': <ConfigParser.ConfigParser instance at 00844B8C>,
 'children': ['APPZLIB'],
 'debug': 0,
 'dependencies': [],
 'destdir': '',
 'directories': [],
 'excludes': [],
 'expatterns': [],
 'exstdlib': 0,
 'extypes': [],
 'icon': '',
 'includes': [],
 'includetk': 0,
 'misc': [],
 'name': 'test.exe',
 'packages': [],
 'pathprefix': ['Q:\\pc',
                'Q:\\pc\\installer\\support'],
 'script': ['test.py'],
 'support': 0,
 'toc': [],
 'trees': [],
 'userunw': 0,
 'zlib': ['APPZLIB']}
Checking zlib[0]
merging test.pyz
self.zlib[0]: APPZLIB
built: {'test.pyz': <__main__.PYZTarget instance at 0088BB6C>, 'installzlib.pyz': <__main__.PYZTarget instance at 00DF175C>}
target: <__main__.PYZTarget instance at 0088BB6C>
ltoc after bindepends:
[('test.pyz', 'test.pyz', 'z'),
 ('imputil', 'c:\\python20\\lib\\imputil.pyc', 'm'),
 ('zlib',
  'q:\\pc\\zlib.pyd',
  'b'),
 ('archive_rt',
  'Q:\\pc\\installer\\support\\archive_rt.py',
  'm')]
sections: ['MYSTANDALONE', 'INSTALLZLIB', 'MYINSTALL', 'APPZLIB']
ltoc after misc:
[('test.pyz', 'test.pyz', 'z'),
 ('imputil', 'c:\\python20\\lib\\imputil.pyc', 'm'),
 ('zlib',
  'q:\\pc\\zlib.pyd',
  'b'),
 ('archive_rt',
  'Q:\\pc\\installer\\support\\archive_rt.py',
  'm')]
ltoc after scripts:
[('test.pyz', 'test.pyz', 'z'),
 ('imputil', 'c:\\python20\\lib\\imputil.pyc', 'm'),
 ('zlib',
  'q:\\pc\\zlib.pyd',
  'b'),
 ('archive_rt',
  'Q:\\pc\\installer\\support\\archive_rt.py',
  'm'),
 ('test', 'test.py', 's')]
ltoc after trees:
[('test.pyz', 'test.pyz', 'z'),
 ('imputil', 'c:\\python20\\lib\\imputil.pyc', 'm'),
 ('zlib',
  'q:\\pc\\zlib.pyd',
  'b'),
 ('archive_rt',
  'Q:\\pc\\installer\\support\\archive_rt.py',
  'm'),
 ('test', 'test.py', 's')]
Applying the following filters:
[<TypeFilter {'d': 1}>]
Final Table of Contents for test.exe:
[('test.pyz', 'test.pyz', 'z'),
 ('imputil', 'c:\\python20\\lib\\imputil.pyc', 'm'),
 ('archive_rt',
  'Q:\\pc\\installer\\support\\archive_rt.py',
  'm'),
 ('test', 'test.py', 's')]
---- InstallTarget: Install_test.exe -----
{'__name__': 'Install_test.exe',
 '_dependencies': [(imputil, c:\python20\lib\imputil.py, m), (archive_rt, q:\pc\installer\support\archive_rt.py, m), (zlib, q:\pc\zlib.pyd, b), (python20.dll, C:\WINNT\System32\python20.dll, b), (carchive_rt, Q:\pc\installer\support\carchive_rt.py, m)],
 'bindepends': ['test.py'],
 'cfg': <ConfigParser.ConfigParser instance at 00844B8C>,
 'children': ['INSTALLZLIB', 'MYSTANDALONE'],
 'debug': 0,
 'dependencies': [],
 'destdir': '',
 'directories': [],
 'excludes': ['PyWinTypes20.dll', 'win32api'],
 'expatterns': [],
 'exstdlib': 0,
 'extypes': [],
 'icon': '',
 'includes': [],
 'includetk': 0,
 'misc': ['MYSTANDALONE'],
 'name': 'Install_test.exe',
 'packages': [],
 'pathprefix': ['Q:\\pc',
                'Q:\\pc\\installer\\support'],
 'script': ['gen_install.py'],
 'support': 1,
 'toc': [],
 'trees': [],
 'userunw': 0,
 'zlib': ['INSTALLZLIB']}
Checking zlib[0]
merging installzlib.pyz
self.zlib[0]: INSTALLZLIB
built: {'test.exe': <__main__.ExeTarget instance at 00845B5C>, 'test.pyz': <__main__.PYZTarget instance at 0088BB6C>, 'installzlib.pyz': <__main__.PYZTarget instance at 00DF175C>}
target: <__main__.PYZTarget instance at 00DF175C>
ltoc after bindepends:
[('win32api', 'c:\\python20\\win32\\win32api.pyd', 'b'),
 ('zlib',
  'q:\\pc\\zlib.pyd',
  'b'),
 ('_sre', 'c:\\python20\\dlls\\_sre.pyd', 'b'),
 ('PyWinTypes20.dll', 'C:\\WINNT\\System32\\PyWinTypes20.dll', 'b'),
 ('python20.dll', 'C:\\WINNT\\System32\\python20.dll', 'b'),
 ('installzlib.pyz', 'installzlib.pyz', 'z'),
 ('imputil', 'c:\\python20\\lib\\imputil.pyc', 'm'),
 ('archive_rt',
  'Q:\\pc\\installer\\support\\archive_rt.py',
  'm')]
sections: ['MYSTANDALONE', 'INSTALLZLIB', 'MYINSTALL', 'APPZLIB']
checking MYSTANDALONE
 which is test.exe
ltoc after misc:
[('win32api', 'c:\\python20\\win32\\win32api.pyd', 'b'),
 ('zlib',
  'q:\\pc\\zlib.pyd',
  'b'),
 ('_sre', 'c:\\python20\\dlls\\_sre.pyd', 'b'),
 ('PyWinTypes20.dll', 'C:\\WINNT\\System32\\PyWinTypes20.dll', 'b'),
 ('python20.dll', 'C:\\WINNT\\System32\\python20.dll', 'b'),
 ('installzlib.pyz', 'installzlib.pyz', 'z'),
 ('imputil', 'c:\\python20\\lib\\imputil.pyc', 'm'),
 ('archive_rt',
  'Q:\\pc\\installer\\support\\archive_rt.py',
  'm'),
 ('test.exe', 'test.exe', 'x')]
ltoc after scripts:
[('win32api', 'c:\\python20\\win32\\win32api.pyd', 'b'),
 ('zlib',
  'q:\\pc\\zlib.pyd',
  'b'),
 ('_sre', 'c:\\python20\\dlls\\_sre.pyd', 'b'),
 ('PyWinTypes20.dll', 'C:\\WINNT\\System32\\PyWinTypes20.dll', 'b'),
 ('python20.dll', 'C:\\WINNT\\System32\\python20.dll', 'b'),
 ('installzlib.pyz', 'installzlib.pyz', 'z'),
 ('imputil', 'c:\\python20\\lib\\imputil.pyc', 'm'),
 ('archive_rt',
  'Q:\\pc\\installer\\support\\archive_rt.py',
  'm'),
 ('test.exe', 'test.exe', 'x'),
 ('gen_install', 'gen_install.py', 's')]
ltoc after trees:
[('win32api', 'c:\\python20\\win32\\win32api.pyd', 'b'),
 ('zlib',
  'q:\\pc\\zlib.pyd',
  'b'),
 ('_sre', 'c:\\python20\\dlls\\_sre.pyd', 'b'),
 ('PyWinTypes20.dll', 'C:\\WINNT\\System32\\PyWinTypes20.dll', 'b'),
 ('python20.dll', 'C:\\WINNT\\System32\\python20.dll', 'b'),
 ('installzlib.pyz', 'installzlib.pyz', 'z'),
 ('imputil', 'c:\\python20\\lib\\imputil.pyc', 'm'),
 ('archive_rt',
  'Q:\\pc\\installer\\support\\archive_rt.py',
  'm'),
 ('test.exe', 'test.exe', 'x'),
 ('gen_install', 'gen_install.py', 's')]
Applying the following filters:
[<FileFilter {'PyWinTypes20.dll': 1}>,
 <FileFilter {'win32api': 1}>,
 <TypeFilter {'d': 1}>]
Final Table of Contents for Install_test.exe:
[('zlib',
  'q:\\pc\\zlib.pyd',
  'b'),
 ('_sre', 'c:\\python20\\dlls\\_sre.pyd', 'b'),
 ('python20.dll', 'C:\\WINNT\\System32\\python20.dll', 'b'),
 ('installzlib.pyz', 'installzlib.pyz', 'z'),
 ('imputil', 'c:\\python20\\lib\\imputil.pyc', 'm'),
 ('archive_rt',
  'Q:\\pc\\installer\\support\\archive_rt.py',
  'm'),
 ('test.exe', 'test.exe', 'x'),
 ('gen_install', 'gen_install.py', 's')]


From gregj at pobox.com  Mon Dec 25 05:02:37 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Mon, 25 Dec 2000 10:02:37 GMT
Subject: Some Python 2.1 ideas
References: <mailman.977590932.19335.python-list@python.org> <Cw616.23054$qN6.3308205@typhoon.we.rr.com> <3A452881.7686E717@uswest.net> <%vb16.187233$U46.5801214@news1.sttls1.wa.home.com> <3A464D5D.B541C04E@uswest.net>
Message-ID: <1XE16.190885$U46.5872195@news1.sttls1.wa.home.com>

"Bjorn Pettersen" <pbjorn at uswest.net> wrote:
> This would be a good idea, except it wouldn't fit with the standard Python
idiom
> for reading files:
>
>    while 1:
>       line = f.readline_chomp()
>       if not line: break # would always break

Oops! That shows how far I thought my idea through! And me always talking
about idioms...

I always use rstrip because I never care about trailing whitespace. If I did
have a reason to preserve trailing spaces I would write my own code to strip
the trailing newlines.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From fredrik at effbot.org  Tue Dec 19 11:36:47 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 19 Dec 2000 16:36:47 GMT
Subject: sys.stdout.write()
References: <r3M%5.159$LU6.61501@typhoon.aracnet.com>
Message-ID: <z8M%5.467$AH6.37187@newsc.telia.net>

Daniel Klein wrote:
> In 'Learning Python' by Mark Lutz and David Ascher, on page 77 it makes
> mention of the 'sys.stdout.write' method. However, when I check this out, I
> don't see the 'write' method in the list...
>
> >>> import sys
> >>> dir(sys.stdout)
> ['shell', 'softspace', 'tags']
>
> I must be doing something fundamentally wrong. :^(

dir() shows the contents of a given namespace.

For an instance object, this is the attributes defined in the
instance object, not anything defined by its class.

Instance objects provide a hidden attribute, __class__, which
points to the class object.

If the instance uses single-level inheritance, you can get away
with dir(sys.stdout.__class__).  To get the full inheritance tree,
you need to check the class object's __bases__ attribute as
well (recursively).

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From aleaxit at yahoo.com  Sun Dec 24 03:18:33 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 24 Dec 2000 09:18:33 +0100
Subject: One Python 2.1 idea
References: <mailman.977590932.19335.python-list@python.org> <9237md$orm$1@nnrp1.deja.com>
Message-ID: <924bjt0e3n@news1.newsguy.com>

<rturpin at my-deja.com> wrote in message news:9237md$orm$1 at nnrp1.deja.com...
    [snip]
> scripting languages what location is to real estate. One of
> the nice things about list comprehensions is that it
> improves performance by abstracting an existing language
> mechanism.

Hi Russell, long time no see!  Unfortunately, the 'improves
performance' here is counterfactual.  There was a silly error
in my benchmark (I was timing the same function twice --
thanks to a kind correspondent for pointing this out!-), and
here it is in the correct form, augmented with a simple loop
solution as a reference point...:

import time

def withMap(inList):
    return map(lambda x: x+1, inList)

def withLC(inList):
    return [x+1 for x in inList]

def withLoop(inList):
    l = len(inList)
    result = [None]*l
    for i in range(l):
        result[i] = inList[i]+1
    return result

def stopWatch(function, N=100*1000):
    start = time.clock()
    function(range(N))
    stend = time.clock()
    return stend-start

print "Using map and lambda:     %f" % stopWatch(withMap)
print "Using list-comprehension: %f" % stopWatch(withLC)
print "Using a simple loop:      %f" % stopWatch(withLoop)

D:\PySym\cover>python spe.py
Using map and lambda:     0.891596
Using list-comprehension: 1.051381
Using a simple loop:      0.721732

D:\PySym\cover>python spe.py
Using map and lambda:     0.706179
Using list-comprehension: 1.005335
Using a simple loop:      0.706091

D:\PySym\cover>python spe.py
Using map and lambda:     0.817602
Using list-comprehension: 1.053890
Using a simple loop:      0.703723




So, the list comprehension in this case is SLOWER by over 10%
than the map-and-lambda.  If speed was indeed paramount, one
should thus eschew it -- counterintuitive, but I guess there must
be problems in the allocation/reallocation strategy of the list
comprehension, which appears to reallocate N times for N items
(while map allocates the resulting-list once, I think).

But note that a simple loop with preallocation of result is _at
least_ as fast as map, here -- indeed, it's the fastest approach.
If speed be all, then, avoiding advanced constructs may in some
cases prove the best approach!


Actually, for all-out speed in a higher-level language, Scheme
(with the stalin compiler) or Common Lisp probably cannot be
beat.  There ARE, after all, decades of Lisp experience and many
groups of clever people laboring away at getting such speed.
(If one is willing to forego dynamic typing, O'Caml or other
MLs may be contenders; Dylan may also be of interest; and
Smalltalk also has speedy implementations, if one will pay...).

I prefer Python because I do NOT prize speed in a higher-level
language above other factors -- great practicality at "playing
well with others", clarity, cleanness of syntax, great productivity.

If you do have another scale of values, Python may not be the
best choice for you at this time; specifically, if for some reason
the solutions MUST be single-language ones (you can't or
won't use a lower-level, faster language for that 10% or so of
your code that bottlenecks its speed).  I do not have any such
constraint, so Python + some tiny amount of C++ at the right places
(often already provided in the form of existing modules such
as Numeric, PIL, ...) is exactly right for my needs...


Alex





From arturo_perez at wanadoo.es  Mon Dec 11 10:45:13 2000
From: arturo_perez at wanadoo.es (Arturo Pérez Mulas)
Date: Mon, 11 Dec 2000 15:45:13 GMT
Subject: Fault trees structure
References: <D_IY5.49$17.2785@m2newsread.uni2.es> <9128e902uo3@news2.newsguy.com>
Message-ID: <dE6Z5.270$Pe.4203@m2newsread.uni2.es>

Hello!

> Fault-tree analysis does appear like an interesting subject, but I
> can't easily find a precise model of "a fault-tree" around the web.
> Can you point to a summary description ("fault-tree analysis" gets
> a lot of hits, but I wonder if a "fault-tree object" can be described
> well enough without getting into the depth of the analysis issues)?

I will definitely post some pointers this evening, and add some more
comments myself.

> Discussing how to model something in terms of Python classes &c seems
> to be a rather basic purpose of this forum, so I think it would be
> all right to pursue a discussion here, but you'd have to supply the
> domain-expertise (at least if I'm the other discussant...!).

It would be nice to know how much you know about them, to write some
meaningful
post that is not too simple neither too complex.

My main goal is not in fault-tree analysing (quantification), since this is
a
very complex and computing costly exercise, and a few wonderful tools are
running out there. My goal is to be able to read/write fault trees from
different
formats and manipulate them from my codes. This is the main reason I got
into
python, and for that matter into OOP. Fault trees seem a nice entity to be
attacked
using OOP (at a further stage I think it would be wonderful, not only to
come out
with an object model for FTs -fault trees-, but to set up an xml set of
rules to
store/interchange them... but this is far away from my knowledge, time and
employer's budget!




From mspal at sangria.harvard.edu  Wed Dec  6 18:08:10 2000
From: mspal at sangria.harvard.edu (Michael Spalinski)
Date: 06 Dec 2000 18:08:10 -0500
Subject: wither PyFITS ?
References: <3A2EB29B.EBB3CCB5@noaa.gov>
Message-ID: <m3u28hup0l.fsf@schwinger.harvard.edu>

>>>>> "j" == j vickroy <jim.vickroy at noaa.gov> writes:

    j> Hello,

    j> The following link

    j> http://monet.astro.uiuc.edu/adass98/Proceedings/barrettpa/

    j> claims there is a FITS module (PyFITS) available, but I have not
    j> been able to find it.

    j> Any help would be appreciated.

Google finds this: http://ra.stsci.edu/pyfits/

--M.


From slee4812 at my-deja.com  Fri Dec 15 11:15:31 2000
From: slee4812 at my-deja.com (Stan Lee)
Date: Fri, 15 Dec 2000 16:15:31 GMT
Subject: comparable shelve function in xml
Message-ID: <91dg2q$3a1$1@nnrp1.deja.com>

Hi,
Is there a comparable way to store Python objects in an XML formatted
file like the shelve module does?  I want to be able to store Python
objects to a file that's human-readable on both a Windows and Linux
platform.  Thanks.

-Stan


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


From gb at a3design.de  Wed Dec 20 07:05:55 2000
From: gb at a3design.de (gunter )
Date: Wed, 20 Dec 2000 12:05:55 -0000
Subject: Greetings from the Interpreter!!!
Message-ID: <91q7b3+7a78@eGroups.com>

import string,time,os

print (string.join((lambda x:x[:10]+x[8:])(map(lambda x:
string.center("*" * (lambda x: ((x < 24) * (x - 3)) + 3)(x) ,24),
range(1,28,2))),os.linesep), "Please Wait...")[time.time()<977612400]





From greg at cosc.canterbury.ac.nz  Tue Dec 12 22:50:35 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Wed, 13 Dec 2000 16:50:35 +1300
Subject: [Q] Are Exceptions used that much in practice?
References: <3A35D546.A4731B5B@enteract.com> <914mfd02ibt@news2.newsguy.com>
Message-ID: <3A36F20B.7B766D0@cosc.canterbury.ac.nz>

Alex Martelli wrote:

> def setattr_ok(obj, attr, value):
>     try: setattr(obj, attr, value)
>     except (AttributeError, ValueError): return 0
>     else: return 1

There are dangers in that approach when the exception
being caught is a very common one such as AttributeError.
If the obj is such that calling setattr on it causes some
Python code to be executed, and there is a bug in that
code which causes an AttributeError to be raised, you'll
never know about it.

A similar problem occurs with Python's use of IndexError
to terminate a for-loop.

For that reason, I have mixed feelings about whether
the exception-as-normal-flow-control style of programming
is a good idea. It might be okay if you use a special-purpose
exception class created just for that particular use,
but abusing exceptions which normally indicate bugs is
asking for trouble, in my opinion.

> def open_if_ok2(filepath):
>     try: return open(filepath, 'r')
>     except IOError: return None
> 

A minor gripe here - it would be more useful if Python
had some subclasses of IOError to distinguish between
different reasons for failure, so you can catch them
more selectively. If the reason is "the file doesn't
exist", you might want to do something to make it exist
and then try again. But if the reason is "the file exists
but the user doesn't have permission" or "someone put 
peanut butter in the disk drive", you probably just want 
to let the exception propagate.

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


From rogerha at aol.com  Sun Dec 10 08:36:24 2000
From: rogerha at aol.com (Roger Abbott)
Date: Sun, 10 Dec 2000 13:36:24 GMT
Subject: WWW_URLEcho with dde
References: <3a32b7e6.8685100@24.0.228.33>
Message-ID: <9100sp$a0b$1@nnrp1.deja.com>

In article <3a32b7e6.8685100 at 24.0.228.33>,
  support @ internetdiscovery.com (Mike Clarkson) wrote:
> Has anyone used the dde module in Python to register a URLEcho from
> netscape or Internet Explorer? Looking at the source code to dde I see
> comments about Netscape's use of poke, which indicates the author
> may have been playing with URLEcho.
>
> I'd like to start a dde server and register it with netscape to send
> me an echo when it loads a web page. Has anyone done this?

Yes, in VB. The web site below has a free WebSpy program which does it.
It comes with source code, I am sure you can figure out how it works.
--
Roger Abbott, RHA (Minisystems) Ltd - http://www.rhaminisys.com
DDE Client and Server ActiveX controls for Visual Basic
DDE FAQ and DDE utility tools, browser URL monitor,
Program launcher/setup menu, other freeware and shareware


Sent via Deja.com http://www.deja.com/
Before you buy.


From robin at jessikat.fsnet.co.uk  Thu Dec 14 07:43:03 2000
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Thu, 14 Dec 2000 12:43:03 +0000
Subject: strip weirds
Message-ID: <QI2l1FAXBMO6EwD6@jessikat.demon.co.uk>

Can somebody explain why strip behaves differently with and without an
argument?

>>> from string import split
>>> split(' ')
[]
>>> split(' ',' ')
['', '']
>>> split('  ')
[]
>>> split('  ',' ')
['', '', '']
>>> 

with the default whitespace arg seems as though a run of whitespace is
being treated as a single character.
-- 
Robin Becker


From red5sf at my-deja.com  Sat Dec  9 18:43:18 2000
From: red5sf at my-deja.com (red5sf at my-deja.com)
Date: Sat, 09 Dec 2000 23:43:18 GMT
Subject: ActiveState: PythonWin not using PythonPath for import?
References: <906ki2$s16$1@nnrp1.deja.com> <3A283CDB.8020800@ActiveState.com> <90gjah$7bh$1@nnrp1.deja.com>
Message-ID: <90ug2l$8pp$1@nnrp1.deja.com>

If you edit the Pythonwin path, that should solve your problem.

Christian

In article <90gjah$7bh$1 at nnrp1.deja.com>,
  Bill Seitz <bseitz at my-deja.com> wrote:
> In article <3A283CDB.8020800 at ActiveState.com>,
>   Mark Hammond <MarkH at ActiveState.com> wrote:
> > Bill Seitz wrote:
> >
> > > I made a folder just below the Python root to hold my own code.
> > >
> > > In PythonWin, I edited PythonPath, adding my folder to the end of
> the
> > > value.
> > >
> > > Yet when I use "import aFunction" when "aFunction.py" is within my
> > > folder, I get an error. When I move the file up to the Python root
> > > folder, it works.
> >
> > The "root" PYTHONPATH entry is the core PythonPath, and is only used
> if
> > Python can not locate its core library etc manually.
> >
> > Add your path to one of the subkeys, or just add your own subkey and
> add
> > your value there...
>
> Thanks, that worked (once I quit/relaunched PythonWin). Is this (a)
> recognized outside PythonWin, and (b) documented?
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From aleaxit at yahoo.com  Mon Dec 11 11:59:21 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 11 Dec 2000 17:59:21 +0100
Subject: Microsoft SOAP not compatible with PythonWare's SOAP?
References: <912q8s$c4i$1@news1.xs4all.nl>
Message-ID: <91315o0ou8@news2.newsguy.com>

"Iwan van der Kleyn" <ivdkleyn at xs4all.nl> wrote in message
news:912q8s$c4i$1 at news1.xs4all.nl...
> Hi there,
>
> I'm developing an Intranet-site with Python. Part of the needed
> business-logic is already implemented as Visual Basic (COM) components
> (let's call this a "legacy-sytem" :-). I've investigated several means of
> integrating them into my intranet-application and a low-cost, conventient
> way would be to build a SOAP-interface which I would access with the
> SOAPlib.py from Pyhtonware (http://www.pythonware.com/products/soap/) .
>
> However, during testing I've not been able to retrieve any data from the
> test MS SOAP components. The soaplib seems to working fine, though. Is
this
> the result of "competing" SOAP implementations? Is there any info
availbale
> about this topic? (? haven't found any on the internet)

I haven't run enough tests for this to be much more than hearsay, but
what little I did do, seemed just about to confirm what I was reading
around the net: the early versions of the MS SOAP toolkit had serious
divergences from other reference implementations of SOAP, but things
have been getting better in this regard, and the very latest MS SOAP
SDK beta *is* mostly-interoperable with stuff from IBM and Developmentor.


Alex





From fredrik at effbot.org  Tue Dec 12 14:49:21 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 12 Dec 2000 19:49:21 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <r0pzoi11lhz.fsf@black.cs.washington.edu>
Message-ID: <5jvZ5.3588$Kd1.418770@newsb.telia.net>

Daniel Wood wrote:
> Here is what I wanted to do.  I have some random parameters, lambda
> and beta.  My calculation uses them like so:
>
> plot( lambda, beta )

lambda is a reserved word, so I doubt that's what you're
doing.  Let's call it "spam" instead.

> I'm using pygtk as a widget set.  I have some sliders.  I wanted to
> make one of the sliders control lambda and one control beta.  The
> control should be scaled by some parameter etc, etc.  Now I could have
> a function
>
> def lambdaCB( val ):
>  global lambda
>  lambda = val / constant
>
> but instead I thought that I would have a simple class
>
> class ScaleCB:
>   def __init__( self, variable, scale ):
>     self.variable = variable
>     self.scale = scale
>
>   def __call__( self, val )
>     self.variable = val / self.scale
>
> (then, I can do lots of parameters with nice simple code instead of
> bunches of stupid helper functions.
>
> This obviously didn't work.

(didn't do what you wanted it to do, at least)

How about this one:

    class PlotParams:
        def __init__(self):
            self.spam = 1
            self.beta = 2
        def setspam(self, value):
            self.spam = value
            self.redraw()
        def setbeta(self, value):
            self.beta = value
            self.redraw()
        def redraw(self):
            plot(self.spam, self.beta)

    myplot = PlotParams()

    spamScaleWidget.setcallback(myplot.setspam)
    betaScaleWidget.setcallback(myplot.setbeta)

> I *think* that the "pythonic" sol'n appears to be lambda = [ value ]
> or some such.

Besides being a syntax error, it doesn't look very Pythonic
to me.  If you want to change a named value, the best way
is to make it an attribute of a class.

</F>




From sholden at holdenweb.com  Wed Dec 20 07:59:38 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 20 Dec 2000 07:59:38 -0500
Subject: Wanted: Python White Papers
References: <91oj8v$i0r$1@nnrp1.deja.com>
Message-ID: <F4206.258$WA2.1012@e420r-atl2.usenetserver.com>

While it isn't a "white paper", the following web page details many users of
the python language, some of them quite high-profile:

    http://www.python.org/psa/Users.html

Hope this helps.

regards
 Steve

<noahspurrier at my-deja.com> wrote in message
news:91oj8v$i0r$1 at nnrp1.deja.com...
> Is anyone aware of any white papers to explain the use of Python
> at any big, high-profile companies?
> I'm trying to use a "look-how-these-cool-companies-use-python"
> argument to defend the use of Python for a project. That won't be my
> only argument, of course :-) But currently Python is not winning many
> management converts here based solely on its technical merits.
> Many still say, "What's Python? Never heard of it...". It helps
> a lot when I say "Google uses it!", but it would help more if I
> knew HOW it is used and WHY.
>
> Actually, the only high-profile company that uses Python that I
> can think of is Google. Is anyone familiar with their architecture?
> How does Python fit into their design?
>
> Is there anyone out there from Google that can represent Google's use
> of Python? What are Google's plans for the future of Python in their
> architecture? I heard that they might be switching to Java.
>
> Yours,
> Noah Spurrier
>
>
> Sent via Deja.com
> http://www.deja.com/
>





From 2rkil at my-deja.com  Fri Dec  8 04:28:41 2000
From: 2rkil at my-deja.com (2rkil at my-deja.com)
Date: Fri, 08 Dec 2000 09:28:41 GMT
Subject: Changing image in label
Message-ID: <90q9k6$468$1@nnrp1.deja.com>

Hi.

I'm totally new to, yet amazed by, Python. I'm
trying to
write a program which displays a portrait, asking
the
user who it is. After answering, the picture
should change
to another person, and so on.

Probably simple, but I'm new to event based
programming.
What happens is that the first picture disappears,
and just
the frame (the sizes) of the new one appears.

Here's the code:

#!/usr/bin/python
from Tkinter import *
import string
root = Tk()

def getquestion():
    return "gfx/torkil_grindstein.gif", "Who is
this?", "Torkil Grindstein"

def getquestion2():
    return "gfx/haakon_brugaard.gif", "But who is
this?", "H?kon Brug?rd"

def update(event):
    global photo_widget, question_widget
    photo, question, fasit = getquestion2()
    gif = PhotoImage(file=photo)
    photo_widget["image"] = gif
    question_widget["text"] = question

def main():
    global photo_widget, question_widget,
answer_widget
    photo, question, fasit = getquestion()

    gif = PhotoImage(file=photo)
    photo_widget = Label(root, relief=RAISED,
image = gif)
    photo_widget.pack()

    question_widget = Label(root, text = question)
    question_widget.pack()

    f = Frame(root)
    f.pack()

    answer_widget = Entry(f)
    answer_widget.bind("<Return>", update)
    answer_widget.pack(side=LEFT)

    answer_button = Button(f, text="Answer")
    answer_button.bind("<Button>", update)
    answer_button.pack(side=LEFT)

    root.mainloop()

if __name__ == "__main__":
    main()

-------
If I, however, put a bogus line at the end of
update(), the
program naturally crashes, but then the correct
image is
shown!

Hope anyone can spot my incompetance...

Torkil


Sent via Deja.com http://www.deja.com/
Before you buy.


From fredrik at effbot.org  Fri Dec 15 12:43:40 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 15 Dec 2000 17:43:40 GMT
Subject: compiled or interpreted?
References: <w8jZ5.74702$IP1.2487256@news1.giganews.com> <usnnt3oj4.fsf@cs.uu.nl> <G5Lz4o.EyA@leeds.ac.uk>
Message-ID: <gLs_5.35$u17.4070@newsb.telia.net>

Nick Efford wrote:
> Has the idea of developing a JIT for Python ever been considered?

yes.

> Or is there some technical reason for it not being possible?

no.

brief-ly yrs </F>




From nospam at caltech.edu  Thu Dec 14 01:19:10 2000
From: nospam at caltech.edu (Nathan Gray)
Date: Wed, 13 Dec 2000 22:19:10 -0800
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu> <976153091.257544@clam-ext> <3A30A181.4302F102@caltech.edu> <p6qr93ik7ks.fsf@informatik.hu-berlin.de> <90u5mo$kr@gap.cco.caltech.edu> <3gyY5.2864$O%.449093@newsc.telia.net>
Message-ID: <3A38665E.1036D6BE@caltech.edu>

Fredrik Lundh wrote:

> Nathaniel Gray wrote:
> > So why not provide a pointer instead of just flaming?
>
> some people do, you know.

Indeed.  For the most part this newsgroup is really friendly and mature.

> (or was my first post to this thread lost somewhere
> in the vast usenet space?)
>
> </F>

Not this time.  The fickle gods of the Caltech newsfeed smiled on that
message (unlike so many others).  I liked it too. <wink>

-n8gray
-at
-caltech
-dot
-edu




From jp at ulgo.koti.com.pl  Thu Dec 28 23:20:40 2000
From: jp at ulgo.koti.com.pl (Jacek =?iso-8859-2?Q?Pop=B3awski?=)
Date: 29 Dec 2000 04:20:40 GMT
Subject: tkinter and images - again
Message-ID: <slrn94o48o.a5p.jp@localhost.localdomain>

I played a while with Tkinter and images. Why can't I use jpeg and png
formats? Is it any way to read and show this kind of files in Tkinter?
The only way I see is to use "convert" from ImageMagick, but...

-- 
Changes comes slowly 
But this is what you'll be
Somehow it snows in the summer  
And the sun is freezing me                     "Otherworld" - Ronnie James Dio


From rob at hooft.net  Tue Dec 12 02:17:49 2000
From: rob at hooft.net (Rob Hooft)
Date: 12 Dec 2000 08:17:49 +0100
Subject: 3rd argument to os.path.walk(): Why?
References: <912gdb$1oqr$1@news.idiom.com> <912nj60ccr@news2.newsguy.com>
Message-ID: <m2vgsqw1k2.fsf@hooft.net>

 AM> E.g., say you want to _count_ the files:

 AM> import os

 AM> class Counter:
 AM>   def __init__(self):
 AM>     self.count = 0
 AM>   def inc(self, n=1):
 AM>     self.count += n

 AM> def count_names(count, dir, flist):
 AM>   count.inc(len(flist))

 AM> count = Counter()
 AM> os.path.walk("/", count_names, count)

 AM> print count.count,"files in all"

If you'd do this, why not make "count_names" a method of the Counter
class?

-- 
=====   rob at hooft.net          http://www.hooft.net/people/rob/  =====
=====   R&D, Nonius BV, Delft  http://www.nonius.nl/             =====
===== PGPid 0xFA19277D ========================== Use Linux! =========


From fredrik at effbot.org  Fri Dec 29 16:31:29 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 29 Dec 2000 21:31:29 GMT
Subject: 2001 Enchancement Wishlist
References: <3A4CEACB.220998A0@javanet.com>
Message-ID: <Ro736.2509$Qb7.273135@newsb.telia.net>

Raymond Hettinger wrote:
> 1.  List.append(elem) and List.pop() are great.  Let's add two more
>      methods, List.prepend(elem) and List.leftpop().  The goal is to
>      make queues,deques, and reverse stacks easier, faster, and clearer
>      to implement.

what's wrong with insert(0, elem) and pop(0) ?

</F>




From sbrunning at bigfoot.com  Mon Dec  4 05:05:53 2000
From: sbrunning at bigfoot.com (Simon B.)
Date: Mon, 04 Dec 2000 10:05:53 GMT
Subject: Python mailing list/usenet bridge.
References: <8vt9t4$r0f$1@nnrp1.deja.com> <900365$fp6$1@nnrp1.deja.com> <905t5s$6ef$1@nnrp1.deja.com>
Message-ID: <90fq9v$jpp$1@nnrp1.deja.com>

*Still* nothing. This is getting a bit irritating - deja is horrid.
*Please*, anyone know who I can nag to get this sorted?

--
Simon Brunning
I was thrown out of college for cheating on the metaphysics exam; I
looked into the soul of the boy sitting next to me. - Woody Allen


Sent via Deja.com http://www.deja.com/
Before you buy.


From reichowa at tcfreenet.org  Wed Dec 13 12:07:03 2000
From: reichowa at tcfreenet.org (Aaron Jon Reichow)
Date: Wed, 13 Dec 2000 11:07:03 -0600
Subject: Smalltalk and Python
In-Reply-To: <kupuiw698s.fsf@lasipalatsi.fi>
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <kuzoi15d60.fsf@lasipalatsi.fi> <Pine.BSF.4.10.10012122310510.48551-100000@tcfreenet.org> <kupuiw698s.fsf@lasipalatsi.fi>
Message-ID: <Pine.BSF.4.10.10012131049520.58848-100000@tcfreenet.org>

On 13 Dec 2000, Erno Kuusela wrote:

> In article <Pine.BSF.4.10.10012122310510.48551-100000 at tcfreenet.org>,
> Aaron Jon Reichow <reichowa at tcfreenet.org> writes:
> 
> | On 13 Dec 2000, Erno Kuusela wrote:
> || the thing that struck me as a little foreign was the apparent
> || isolation of the smalltalk image from the rest of the operating system
> || and the filesystem. i want to be able to grep code!
> 
> | While I'm not a proponent of Smalltalk isolationism, grep is a useless
> | tool (more or less), simply an artifact, ye, a symptom of the
> | sickness known as file-based development.
> 
> point was that it's valuable to me to be able to manipulate stuff via
> the filesystem. version control, editors, access control, locking,
> backups, and all the rest of the universe works with files (same
> critique goes for zope).

I think there's definate value of a more unix-ish, script-aimed side of
Smalltalk.  I, personally, would like to have Smalltalk scripts (GNU
Smalltalk is OK, but something more full featured) that I could use in
place of bash and python scripts.  But beyond that, files are largely
obsolete.  Not in practicality, but in how we should represent data.

There's a CVS interface to Squeak, so I don't think that the rest of the
things you listed quite apply.

Aaron

Aaron Reichow                   :: "In essence, Smalltalk is a programming
Twin Ports ACM Vice President   ::  language focused on human beings rather
http://www.d.umn.edu/~reic0024/ ::  than the computer." - Alan Knight




From sholden at holdenweb.com  Fri Dec 22 12:30:52 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 22 Dec 2000 12:30:52 -0500
Subject: Refreshall from Python 1.3
References: <977505346.331315@osiris.esoterica.pt>
Message-ID: <XeM06.7892$3O1.86960@e420r-atl1.usenetserver.com>

Jos? Rui Faustino de Sousa <jrfsousaREMOVE00592424 at esoterica.pt> wrote in
message news:977505346.331315 at osiris.esoterica.pt...
> Hi!
>
> I am learning Python using:
>
> Internet Programming with Python
>
> and
>
> Programming Python
>
> The problem is that both of these books target Python 1.3...
>
> Where can I find some "refreshall" course to put me up to date with
> Python 2.0?
>
> Preferably online... Allready too much $$$ on books... ;-)
>
> Best regards
> Jos? Rui
>
Well, 1.3 and 1.5.2 weren't all that different.  You will probably find that
you can understand the overview referenced in:

    http://www.python.org/2.0/

Hope this helps your wallet!

regards
 Steve






From othello at javanet.com  Wed Dec 13 13:01:27 2000
From: othello at javanet.com (Raymond Hettinger)
Date: Wed, 13 Dec 2000 13:01:27 -0500
Subject: newbie question: python vs. Matlab
References: <3A38800A.6CE1C201@yahoo.com>
Message-ID: <3A37B976.DB2BCA3B@javanet.com>

Based on my experience with both languages, I can tell you that Python
is slower, has a less convenient vector operations, and
lacks the fully developed, expansive book of functions that MatLab
offers (Cleve Moler has thought out every nuance when it comes to round-off
error, optimization, sparse arrays, and algorithmic trade-offs).  NumPy is
much younger and less developed.

That nothwithstanding, you'll love Python.  Building GUIs is much easier;
and there are libraries for an incredible range of non-numeric tasks.
M-files look simple at first, but do not scale up as cleanly as Python
(also, you won't miss meshgrid() one bit).  The best part that you can
easily interface with other apps. It only takes one or two lines to
extract data from the internet or generate email output.  PythonWin
makes it easy to push and pull data from Excel.  In short, Python's
"glue" capabilities and clear syntax far exceed Matlab's abilities.

When I need the best of both worlds, I write everything in Python except
for the intense numeric steps which I delegate to Matlab.  Under windows,
the MatLab engine is easily accessible using  DDE (dynamic data exchange)
interface.

Raymond

P.S. I neglected to mention that Python is free, extendable in C, and has
open source code.


Jinsong Liang wrote:

> Hi every body,
>
> I am a control engineer and familiar with Matlab. It seems to me that
> python is similar to Matlab in some sense. Who can tell me what the
> advantage of python over Matlab is?
>
> Thanks.
>
> Leo



From cerutti at together.net  Fri Dec 15 16:13:09 2000
From: cerutti at together.net (Neil Cerutti)
Date: 15 Dec 2000 21:13:09 GMT
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <tcmay-FCA22F.14073713122000@news.got.net> <918v7a02scn@news1.newsguy.com> <tcmay-04E80C.16181313122000@news.got.net> <91a1ed02a4j@news2.newsguy.com> <3A389B1A.4F33FFD4@cfmu.eurocontrol.be> <91ae8s02llf@news2.newsguy.com>
Message-ID: <slrn93l2lc.5l.cerutti@fiad06.norwich.edu>

Alex Martelli posted:
>[good discussion about squares and rectangels snipped]
>
>But that's implementation, and a well known
>problem with inheriting-from-concrete in that
>realm.  What I'm saying is that the problems with
>inheritance-from-instantiable (fully concrete)
>classes start much earlier than at implementation
>time -- start with conceptual, ontological issues
>regarding OOA and related classification tasks.

This problem arises from the way that our minds work. We like to
classify things, and the classifications are incredibly
complicated. Object Orientation doesn't model our minds very well
due to the problems you discussed. We end up needing imaginary
classes containinag, as practically as we can deduce, the
commonalities in groups of objects we're modelling--for the sake
of inheritance. In reality an ostrich and a blue-jay and a
pengiun are all birds, but the relationships between them don't
*really* go back to a primordial ideal BIRD. They're family
relationships, which I don't know how to model.

Perhaps a new paradigm of "family oriented" programming will crop
up some day?

-- 
Neil Cerutti <cerutti at together.net>


From urner at alumni.princeton.edu  Wed Dec 20 19:56:52 2000
From: urner at alumni.princeton.edu (Kirby Urner)
Date: Wed, 20 Dec 2000 16:56:52 -0800
Subject: Python in "math through programming" curriculum
References: <2g2t3t06r2agrrjmfnr41eu4lc2u39a8qi@4ax.com> <3A3EFF47.67DE69CF@gte.net> <kj624toonmgugtrgrpkea9b01a56hgogam@4ax.com> <3A413A14.54DF58D1@gte.net>
Message-ID: <laj24tcfio6ldfqktnehkck93j7v6iciiv@4ax.com>

"Clifford J. Nelson" <cnelson9 at gte.net> wrote:

>> I went to this website.  It doesn't work for the Mac at all --
>> Windows only.  So what's your beef with Python on the Mac?
>> At least it exists for the Mac.
>
>That's why I don't attach ANY weight to what you write!
>Clean used to be a "Mac first product", and Version 1.3.x
>soon to be updated to Version 2.0 runs on the Mac. You would
>know that it runs on the Mac if you could just read.
>

The latest version with the fancy games library doesn't 
run on the Mac yet (or Linux either).  Older versions 
do, yes, I see that.

I thought maybe you were advocating Clean for kids because 
of the games angle -- what I asked you about.  Apparently 
not so, given the older Clean suits your purposes (what 
do you do with it??).

>> Do you think this emphasis on game programming is better
>> than my "math through programming" approach?  It seems to

So then... what's so attractive about Clean, from your 
personal experience?  I'm just asking.  You haven't said 
anything about that yet (older versions run on the Mac...  
... ... so?).

Python runs on the Mac too (the latest version even).

http://www.python.org/download/download_mac.html
http://www.python.org/doc/current/mac/mac.html


<< repeated text snipped >>

Lets not forget Squeak and C# while we're at it (the former 
actually has some appealing features as a beginner environment 
-- I've played with it some, but not enough, haven't touched 
C# yet).

And, yes, I do make use of Bucky Fuller's stuff in my Pythonic
'Computer Literacy and Numeracy' pages starting at 
http://www.inetarena.com/~pdx4d/ocn/numeracy0.html (most 
curriculum writers still completely ignore his contributions 
-- it's non-standard to integrate it the way I do). 

But I don't see why you're making such a big deal about
that.  I also make use of stuff from Gauss, Fermat, Euler, 
Bell, Descartes and Pascal (sorry about so many guys in 
this context, that'll change)... Euclid and Erastosthenes...

>> My guess is you've done approximately zero using any of
>> the languages under discussion (Python, Clean, Caml).  If
>> I'm wrong, I hope you will show me otherwise.  I don't
>> have a lot of time for completely uninformed discussion.
>> 
>> Kirby
>
>You could inform yourself, if you could only read.
>
>  Cliff Nelson

Read what?  Where's your Clean, OCaml or Python?  You want
me to use a search engine:  "CLIFF NELSON OCAML"?  Didn't
work.  

Does the fact that older versions of Clean run on the Mac 
constitute some big eye-opener?  I'm supposed to be enlightened
now?

You've shown us zilch so far.  Maybe you have some far out 
stuff in Clean to share with us, that you've done on your 
Mac?  That'd be nice.  Wanna post an URL?  Or even answer
my (specific) questions? 

Anyway, I'll stop replying to your posts unless you choose 
to include some substance.  It's fun to bicker like this, 
I have to admit, but I can't justify the time.  My apologies
to comp.lang.python for indulging myself (I don't think 
that other newsgroup is even working -- maybe (and sci.
crypto was a typo -- no such animal)).

Kirby



From jepler at inetnebr.com  Mon Dec 11 08:39:35 2000
From: jepler at inetnebr.com (Jeff Epler)
Date: Mon, 11 Dec 2000 13:39:35 GMT
Subject: How can I use printer port in python?...
References: <90vnol$l5p$1@news1-2.kornet.net>
Message-ID: <slrn939m8n.6bm.jepler@potty.housenet>

On Sun, 10 Dec 2000 20:02:24 +0900, skidrow at nownuri.net
 <skidrow at nownuri.net> wrote:
>As the title says, I was wondering if there's anyway I can
>
>use printer port in python...
>
>

In Unix, if your printer is named /dev/lp1, just use
	f = open("/dev/lp1", mode)
and then look at the contents of the fcntl and FCNTL modules
for ways to perform low-level manipulation of the printer.

If you want to just write output to the printer, use
	f = os.popen("lpr", "w")
and write to f.  You may need to get arguments to lpr, such as
"-p<printer name>".

There is bound to be a way to do it on Windows, but I don't have a clue how.

There is probably no cross-platform way.

Jeff


From reidom1 at yahoo.com  Fri Dec  8 09:14:36 2000
From: reidom1 at yahoo.com (Dominik Reiter)
Date: Fri, 8 Dec 2000 06:14:36 -0800 (PST)
Subject: caller process hangs
Message-ID: <20001208141436.18378.qmail@web1504.mail.yahoo.com>

Hi everybody,

I have embedded python into a C program. From there I
have to call a python - script. Unfortunately,
whenever I do this, the script starts to run, but the
C programm hangs until the script is finished. I need
to have access to both programs at the same time. Is
there any way to do this. I'm running on a SGI machine
with python 1.6 (necessary). The code for the C
program looks like this (shortened):

#include	<aptypes.h>
#include	<appublic.h>
#include	<uipublic.h>
#include	<uierr.h>
#include        <Python.h>
#include        <config.c>

#ifdef R_ROUTINE
#undef R_ROUTINE
#endif

#define	R_ROUTINE	"main"

main (argc,argv)
int	argc ;
char	**argv ;

{
	char	*app_name = "vlo"; 
	char	*rsrc_name = "vlo"; 
	char	*top_menu  = "menu";

Py_SetProgramName(argv[0]);
Py_Initialize();
PyRun_SimpleString("import vlorob");

#undef	R_ROUTINE

and here's the code for the script, which itself calls
a pythonprogram:

import os, string

# set v for Module-directory
v = 'mod'
# set s for startupfile
s = 'somefile.py'

a = os.environ['PYTHONPATH']
b = string.split(a, ':')
for i in b:
	if i[-3:] == 'lib':
		if string.find(i, v) > 0:
			c = i
d = os.path.join(c, s)
os.system('python ' + d)

I've tried to do something with threads, or call a new
Shell, but to no effect.

Any help on this is greatly appreciated

Dominik

__________________________________________________
Do You Yahoo!?
Yahoo! Shopping - Thousands of Stores. Millions of Products.
http://shopping.yahoo.com/



From jdries at mail.com  Sat Dec 30 22:28:42 2000
From: jdries at mail.com (Jan Dries)
Date: Sun, 31 Dec 2000 04:28:42 +0100
Subject: HTML Parser
References: <0FF17853157CD31197E100902784B34B7A1D06@FILESERVER1>
Message-ID: <3A4EA7EA.C3E4CE8E@mail.com>


> "Voitenko, Denis" wrote:
> 
> I am trying to write an HTML parser. I am starting off with a simple
> one like so:
> 
[snip]
> HTMLtags=re.compile('<.*>')

This is presumably the root of your problems. The expression will match
the longest string that starts with a < and ends with a >, including
strings with < and > in the middle. So in your example, the HTMLtags
matches the entire string "<a href=hello.jsp>Hello</a>", because it's
the longest part of your input that starts with < and ends with >.

Try the following instead:

    HTMLtags=re.compile('<[^>]*>')

It will match strings that start with <, end with > and have no > in the
middle. Depending upon the HTML you wish to parse, even this might not
be enough though, because the value of some HTML attributes may contain
a ">".

Regards,
Jan



From paulp at ActiveState.com  Sun Dec 10 16:14:24 2000
From: paulp at ActiveState.com (Paul Prescod)
Date: Sun, 10 Dec 2000 13:14:24 -0800
Subject: Parsing text in python
References: <tYIY5.47$17.2755@m2newsread.uni2.es>
Message-ID: <3A33F230.3B173656@ActiveState.com>

> Arturo P?rez Mulas wrote:
> 
> Hello,
> 
> I was wondering if there is a parsing tool for python similar to
> yacc/lex. I have only come across Plex, but
> seems very limited in comparison with yacc (lacks iterations).

You can look these things up on the "Vaults":

http://www.vex.net/parnassus/apyllo.py?i=137382805

 Paul Prescod



From support at internetdiscovery.com  Sat Dec  9 18:39:56 2000
From: support at internetdiscovery.com (Mike Clarkson)
Date: Sat, 09 Dec 2000 23:39:56 GMT
Subject: Documentation - underscores in \section
References: <Pine.LNX.4.21.0012081707130.29380-100000@saturn.ispol.com> <udvgsuv6jm.fsf@box.home.de>
Message-ID: <3a32a601.4103125@24.0.228.33>

On 09 Dec 2000 00:26:21 +0100, "Jochen =?iso-8859-1?q?K=FCpper?="
<jochen at pc1.uni-duesseldorf.de> wrote:

>>>>>> Gregory Trubetskoy wrote on Fri, 08 Dec 2000 22:15:55 GMT:
>
>Gregory> [Sorry this is perhaps somewhat off-topic]
>
>Yes, there is comp.text.tex as well. I keep it here, as there is no
>LaTeX bug/misbehaviour to see right now. In case you see a
>LaTeX/TeX-problem, please crosspost and set f'up there!

It's on topic here because the problem will likely be specific to
the LaTeX macros Python uses for its documentation.

>Gregory> Has anyone noticed that underscores do not work in \section,
>Gregory> even if preceeded by a \ (backslash).

You'll need a guru. I solved this once years ago for LaTeXinfo, but
I've forgotten how, and it would be different for LaTeX 2e.

>What packages are loaded by these docs? Have you compiled them
>yourself? What LaTeX/TeX versions?

This is important, as is the order in which they were loaded.

Mike.


From arturo_perez at wanadoo.es  Sun Dec 10 05:31:53 2000
From: arturo_perez at wanadoo.es (=?iso-8859-1?Q?Arturo_P=E9rez_Mulas?=)
Date: Sun, 10 Dec 2000 10:31:53 GMT
Subject: Parsing text in python
Message-ID: <tYIY5.47$17.2755@m2newsread.uni2.es>

Hello,

I was wondering if there is a parsing tool for python similar to yacc/lex. I have only come across Plex, but
seems very limited in comparison with yacc (lacks iterations).

Thanks

Arturo.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20001210/c5d1ade3/attachment.html>

From moshez at zadka.site.co.il  Mon Dec 25 12:29:33 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Mon, 25 Dec 2000 19:29:33 +0200 (IST)
Subject: Is there any analog of perl's pack()/unpack() in Python?
In-Reply-To: <9271lj$2b5u$1@news1.simtel.ru>
References: <9271lj$2b5u$1@news1.simtel.ru>
Message-ID: <20001225172933.182B5A84F@darjeeling.zadka.site.co.il>

On Mon, 25 Dec 2000, "Alexander Semenov" <sav at ulmen.mv.ru> wrote:

> Is there any way to read/write file with binary coded numbers in it?

The struct module.
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From tim.one at home.com  Sat Dec 23 00:45:41 2000
From: tim.one at home.com (Tim Peters)
Date: Sat, 23 Dec 2000 00:45:41 -0500
Subject: A Suggestion for Python Colon Syntax
In-Reply-To: <Pine.NXT.4.21.0012230024000.233-100000@localhost.virginia.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEDGIFAA.tim.one@home.com>

[Tim]
> I expect the PEP will be rejected, but that doesn't mean it
> will be.  Guido did surprise me once, in about 1995 <wink>.

[Steven D. Majewski]
> Please save me from searching thru a years worth of python-list
> archives and tell me what that 1995 surprise was!

My mistake:  it was 1994.  That's when Guido released Amrit Prem's code
adding "lambda" in 1.0.0 without breathing a word of it on the mailing list
first.  So I guess he surprised me again in 1995 by leaving it in <wink>.

So what do *you* want for Christmas, Steven?  We have to do *something* to
get you back from that XLISP crap.  How about optional colons, or mandatory
parentheses around 2-letter variable names?

hoping-you'll-settle-for-a-fruit-cake-ly y'rs  - tim




From amk at mira.erols.com  Sat Dec 23 10:00:59 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 23 Dec 2000 15:00:59 GMT
Subject: Why Won't This Work?
References: <mailman.977582107.13540.python-list@python.org>
Message-ID: <slrn949gag.mr.amk@207-172-111-138.s138.tnt1.ann.va.dialup.rcn.com>

On Sat, 23 Dec 2000 06:31:38 -0800, Ben Ocean <lists at TheWebsons.com> wrote:
>and everything appears to be in order. But when I type
>python
>at a command prompt, I get a bash error. What have I not done?

Probably it's very simple; do you have '.', the current working
directory, in your PATH variable?  Try ./python.

--amk


From jp at ulgo.koti.com.pl  Sun Dec 24 00:31:04 2000
From: jp at ulgo.koti.com.pl (Jacek =?iso-8859-2?Q?Pop=B3awski?=)
Date: 24 Dec 2000 05:31:04 GMT
Subject: string split
Message-ID: <slrn94b2go.srm.jp@localhost.localdomain>

I have string:

s="one two <br> three"

splitting it gives:

['one', 'two', '<br>', 'three']

but if string looks like that:

s="one two<br> three"

string.split(s) gives me:

['one', 'two<br>', 'three']

is it possible to split not by space (" ") but by "<" and ">" ?

I need:

['one two','<br>','three']

How to do it in simple way?

-- 
The lover of life's not a sinner 
The ending is just a beginner
The closer you get to the meaning                             "Heaven and Hell" 
The sooner you'll know that you're dreaming                 - Ronnie James Dio


From quinn at dinar.ugcs.caltech.edu  Sat Dec 16 21:12:05 2000
From: quinn at dinar.ugcs.caltech.edu (Quinn Dunkan)
Date: 17 Dec 2000 02:12:05 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <r0pzoi11lhz.fsf@black.cs.washington.edu> <5jvZ5.3588$Kd1.418770@newsb.telia.net> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com> <kuu28971kp.fsf@lasipalatsi.fi> <3A380E4E.8537A0FB@san.rr.com> <91a1vr02b02@news2.newsguy.com> <Co9_5.25465$x6.13756579@news2.rdc2.tx.home.com> <91bbq102mtk@news1.newsguy.com> <rWd_5.26820$x6.14119694@news2.rdc2.tx.home.com> <91d5el0me7@news1.newsguy.com> <rOr_5.31120$x6.15241856@news2.rdc2.tx.home.com>
Message-ID: <slrn93o86s.tu0.quinn@dinar.ugcs.caltech.edu>

On Fri, 15 Dec 2000 16:38:47 GMT, Rainer Deyke <root at rainerdeyke.com> wrote:
>Yes, in-place modification should be supported.  I had something like this
>in mind:
>
>class UserInt:
>  def __init__(self, value):
>    self.data = value
>  def __add__(self, other):
>    if isinstance(other, UserInt):
>      return self.__class__(self.data + other.data)
>    else:
>      return self.__class__(self.data + other)
>  __radd__ = __add__
>  def __iadd__(self, other):
>    if isinstance(other, UserInt):
>      self.data += other.data
>    else:
>      self.data += other
>  # Repeat for all other operations
>
>UserInt as written is more properly UserNumber, since it supports longs ints
>and floats.

Just as a matter of style, would it be superior to add a __coerce__ method
that knows how to coerce ints, longs, and floats into UserInts, or make an
explicit check on each method?  __coerce__ saves a lot of typing, but I'm
generally wary of implicit things like coercion.  When I implemented a class
like this (a vector, actually), I used __coerce__, but after trying to wrap my
brain around the control flow, I was thinking maybe I wouldn't next time, even
though it seems to work fine.

Is this fear unfounded?  Is there a better way?


From paul.moore at uk.origin-it.com  Wed Dec 20 05:50:22 2000
From: paul.moore at uk.origin-it.com (Paul Moore)
Date: Wed, 20 Dec 2000 11:50:22 +0100
Subject: Events in win32com
References: <Y1g=Op6IfWSoJn1ISFtOyuvqwBb3@4ax.com> <FY0=OlZFLy71DtnK62u5TEPlhOts@4ax.com> <3A40060B.10504@ActiveState.com>
Message-ID: <cYxAOmxi9FgNERgBo29zuCl5c2h1@4ax.com>

On Wed, 20 Dec 2000 02:11:29 +0100, Mark Hammond
<MarkH at ActiveState.com> wrote:

>Paul Moore wrote:
>
>> After scanning Deja, I found an article describing DispatchWithEvents.
>> The example in the documentation (OnVisible) works fine. But the
>> OnQuit event doesn't seem to work :-( Basically, the event routine is
>> never fired. Is this because the event routine has a ByRef "Cancel"
>> parameter? Or is it something else?
>
>OnQuit works for me.  test\testMSOfficeEvents.py traps this.
>

Grrr. I'm having problems getting events fired from any of the MS
applications (IE, Excel), either in VBScript or in Python. Looks like
it's something stupid I am doing... We'll ignore that for now...

On the other hand, if I use a 3rd party control (WScriptEx) I get
events quite happily in both VBS and Python. But I tried the "return"
trick to set a ByRef parameter, which is shown in
test\testMSOfficeEvents.py, but it doesn't seem to work. I have

    from win32com.client import *

    class WexEvents:
        def OnTick(self, rem, wake):
            print "Tick...", rem, wake
            if rem < 5000:
                # wake = 1 # doesn't work...
                return 1 # Goes into the only ByRef parameter...

    wex = DispatchWithEvents("WshToolbox.WScriptEx", WexEvents)

    print "Zzzzz"
    wex.Sleep(10000,1000)
    print "Hello!"

Here, the WScriptEx Sleep method takes a time to sleep and a time
between ticks, and fires an OnTick event on each tick. The OnTick
event takes a "Remaining Time" (ByVal) parameter, and a "Wake Up"
(ByRef) parameter. My understanding of the example is that returning
1, should set the wake parameter to 1, as it is the only ByRef
parameter and the OnTick event is defined as a sub. But it doesn't
seem to work.

Am I doing something wrong, or is this not expected to work?

Paul



From amk at mira.erols.com  Fri Dec 22 13:30:41 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 22 Dec 2000 18:30:41 GMT
Subject: ANNOUNCE: soundex.py
References: <mailman.977338109.32467.python-list@python.org> <3A4119C3.2E5E54E1@uswest.net> <mailman.977501169.22862.python-list@python.org>
Message-ID: <slrn94787k.1q7.amk@207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com>

Does this mean that Modules/soundex.c can be dropped from the Python
CVS tree?

--amk



From edcjones at erols.com  Fri Dec  8 19:00:14 2000
From: edcjones at erols.com (Edward C. Jones)
Date: Fri, 08 Dec 2000 19:00:14 -0500
Subject: Python wrapper for C++ STL?
Message-ID: <3A31760E.D157EB54@erols.com>

Has anyone written a Python wrapper for the C++ Standard Template
Library?

Ed Jones




From cjensen at bioeng.ucsd.edu  Tue Dec 12 18:19:36 2000
From: cjensen at bioeng.ucsd.edu (Curtis Jensen)
Date: Tue, 12 Dec 2000 15:19:36 -0800
Subject: Python-like compiled language
References: <913jr3$dmv$1@sylvester.vcn.bc.ca> <3A35835B.AAA92005@bioeng.ucsd.edu> <kuzoi178kg.fsf@lasipalatsi.fi>
Message-ID: <3A36B288.FD0524DF@bioeng.ucsd.edu>

Erno Kuusela wrote:
> 
> In article <3A35835B.AAA92005 at bioeng.ucsd.edu>, Curtis Jensen
> <cjensen at bioeng.ucsd.edu> writes:
> 
> | The closest compiled languege to Python is probaby C/C++.
> 
> huh? those are about a million times lower level than python,
> with manual memory management and pointers and all...
> c++ is also about a million times more complex and pretty ugly
> with all the backwards-compatibility baggage.
> 
>   -- erno

Yes, C/C++ are much lower level.  Though the question was "what is the
most Python-like compiled language?"  If you don't agree with C/C++ then
what is closer.  I thought about including Java.  However, Java realy
isn't realy what I'd call compiled.  It was made to be platform
independant, and binaries for java aren't realy distributed (other than
for wintel and sometimes linux).  Also, Java is pretty slow, and the
questioner was looking for speed.  Also, syntacticly, C may be more like
Python, but it may be too low-level.  C++ has some nicer memory
managment and output routines.  So I decided on C/C++ as the closest.

There are a few posts disagreeing.  However, none of them posted an
alternative.  C/C++ is the closest I could think of without going into
some obscure languages, like some of the scientific calculator
languages, or Oberon, or ML.

If C/C++ isn't the closest, then what is?

-- 
Curtis Jensen
cjensen at bioeng.ucsd.edu
http://www-bioeng.ucsd.edu/~cjensen/
FAX (425) 740-1451


From joconnor at cybermesa.com  Thu Dec 21 16:03:27 2000
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Thu, 21 Dec 2000 21:03:27 +0000
Subject: 2 question from a python newbie!
 
Message-ID: <3A42701F.99D71915@cybermesa.com>

> > 1). Can python scripts be compiled to 'stand alone' .exe files?

What with executable programmings shipping with a myriad of DLL or share
library files, I'm not sure why making an 'executable' is that big an
issue 

>P.S. so python is really used as a scripting language and nothing more?

No.  It makes a good scripting language, but twith the various UI and DB
interfaces it has, it's used for a lot more

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

"God himself plays the bass strings first when He tunes the soul"



From andreas at andreas-jung.com  Mon Dec 25 05:06:48 2000
From: andreas at andreas-jung.com (Andreas Jung)
Date: Mon, 25 Dec 2000 11:06:48 +0100
Subject: Why is Python slow? (was Re: One Python 2.1 idea)
In-Reply-To: <slrn94db6a.rc.neelk@alum.mit.edu>; from neelk@alum.mit.edu on Mon, Dec 25, 2000 at 01:06:48AM +0000
References: <mailman.977590932.19335.python-list@python.org> <9237md$orm$1@nnrp1.deja.com> <924bjt0e3n@news1.newsguy.com> <m38zp6nk8k.fsf@localhost.localdomain> <925mpe$jjg$1@panix6.panix.com> <slrn94db6a.rc.neelk@alum.mit.edu>
Message-ID: <20001225110648.A30627@yetix.sz-sb.de>

On Mon, Dec 25, 2000 at 01:06:48AM +0000, Neelakantan Krishnaswami wrote:
> The Python code was:
> 
> def loop(n):
>     t = time.time()
>     while n != 0:
>         n = n - 1
>     return time.time() - t
> 
> def loop2(n):
>     t = time.time()
>     for i in xrange(n):
>         pass
>     return time.time() - t

This code is not very useful for real-world benchmarks. We have build several
large systems with Python and I have seen FEW situations where Python is REALLY
slow. As a universal scripting language Python is fast enough for most purposes. So
it is not really a good idea to measure the speed of Python with such a useless
benchmark.

Andreas




From ggabor at sopron.hu  Wed Dec  6 12:29:25 2000
From: ggabor at sopron.hu (Gabor Gludovatz)
Date: Wed, 6 Dec 2000 18:29:25 +0100 (CET)
Subject: parsing and searching big text files
Message-ID: <Pine.LNX.4.21.0012061820200.7564-100000@ns.sopron.hu>

Hello,

I have a big text file which contains 3 variable lenght columns with
names. I have to search this file for names in either column and have to
show to other 2 names, and I have to do this very fast!

The text file is about 3 megs long. Does someone know a method to do this
quickly?

Which is the faster, should I read to whole text file into the memory and
parse and search it there, or should I read it from the disk line by line
and parse it..?
the last seemed to be very slow.

Here is an example line from this text file:
Foo Bar:John Doe:Bill

If I look for, for example, John Doe, the function should return Foo Bar
and Bill and all the other records which contain John Doe.

Again: I have to do it very fast.


-- 
 Gabor Gludovatz <ggabor at sopron.hu> http://www.sopron.hu/~ggabor/





From sholden at holdenweb.com  Fri Dec 22 08:31:41 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 22 Dec 2000 08:31:41 -0500
Subject: ?Why is 0e-6 invalid?
References: <3A428C7C.5CFBB181@email.sps.mot.com> <mailman.977460545.25728.python-list@python.org> <Pine.LNX.4.10.10012220922540.14068-100000@kev.ioc.ee>
Message-ID: <FKI06.6065$3O1.66350@e420r-atl1.usenetserver.com>

Pearu Peterson <pearu at cens.ioc.ee> wrote in message
news:Pine.LNX.4.10.10012220922540.14068-100000 at kev.ioc.ee...
>
>
> On Thu, 21 Dec 2000, D-Man wrote:
>
> > AFIAK scientific notation is [1-9]e-?[0-9]+
> > (a 1 thru 9 followed by e followed by optional negative sign followed
> > by any sequence of 0 thru 9)
>
> Scientific notation in Python is a bit more involved:
>
> [+-]?((\d+[.]\d*|[.]\d+|[1-9]+)[eE][+-]?\d+ # may not be a complete
pattern
>
> that allows to represent number `0' also.
>
> Though, I don't understand why 0e-6 is invalid in Python, indeed.
>
> Pearu
>
>
In section 2.4.5 of the (Python 2.0) documentation it states quite clearly
that an "exponentfloat" must begin with a non-zero digit.  You will find
that

    01e-6

is also illegal syntactically.  In full: "Note that the integer part of a
floating point number cannot look like an octal integer, though the exponent
may look like an octal literal but will always be interpreted using radix
10. For example, "1e010" is legal, while "07.1" is a syntax error."

regards
 Steve






From peter at engcorp.com  Tue Dec 12 23:35:01 2000
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 12 Dec 2000 23:35:01 -0500
Subject: compiled or interpreted?
References: <IHBZ5.126133$DG3.2517319@news2.giganews.com>
Message-ID: <3A36FC75.A26BF23E@engcorp.com>

Matt Frey wrote:
> 
> Is Python a compiled language that can be run on a platform like c++, or is
> it interpreted like java, or is it in a class of its own?

If you didn't like the previous four excellent answers, how about this
one?

  No.


:-)


From fredrik at effbot.org  Wed Dec 13 02:56:52 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 13 Dec 2000 07:56:52 GMT
Subject: running a pipe of commands from python
References: <yv24s08pz0c.fsf@lionsp093.lion-ag.de>
Message-ID: <8ZFZ5.3701$Kd1.429028@newsb.telia.net>

Harald Kirsch wrote:
> In Tcl I just write
>
>    exec doThis | doThat |toMore | finishUp >output <input 2>errors
>
> In the Python reference manual I find os.exec(), os.fork(), popen2()
> etc, but none of these is as easy to use as Tcl's exec. What is
> Python's equivalent (and I am sure there must be somehting)?

you don't say what output, input and errors are, so it's
not obvious what you're looking after.

if they're filenames, use:

    os.system("doThis | doThat >output ...")

if they're file handles, use os.popen3:

    input, output, errors = os.popen3("doThis | doThat")

for a more general (but unix-only) solution, see:
http://www.python.org/doc/current/lib/module-pipes.html

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->





From cjensen at bioeng.ucsd.edu  Mon Dec 11 20:46:03 2000
From: cjensen at bioeng.ucsd.edu (Curtis Jensen)
Date: Mon, 11 Dec 2000 17:46:03 -0800
Subject: Python-like compiled language
References: <913jr3$dmv$1@sylvester.vcn.bc.ca>
Message-ID: <3A35835B.AAA92005@bioeng.ucsd.edu>

Alexander Lawhead wrote:
> 
> This may seem like an odd question, but what is the most Python-like
> compiled language? I'm a new programmer and have been learning on Python.
> I'm trying to build some command-line driven programs to do some image
> processing and am trying to find a Python-like compiled language that
> would be faster than Python+NumPy. Ideally I guess I should learn C and
> combine Python + C extensions. For the moment, I'd like to start on a new
> compiled language, preferably one that's oriented towards math/science
> (hopefully with built-in or standard matrix libraries). I've fallen in
> love with Python+NumPy's matrix capabilities and am far more interested in
> a language that comes close in this regard. Object oriented capability is
> not necessary.
> 
> So far, I've been pointed to F, a subset of Fortran 90. Is there anything
> else out there? Wide usage is not necessary, but free availability under
> Linux or WinXX is!
> 
> Thanks,
> 
> Alexander

The closest compiled languege to Python is probaby C/C++. 

If you're looking for speed, your best bet is to use a mathamtical
library (like Numpy).  As was mentioned in a previous post, it's almost
always a small section that is time intensive.  So, you can use Python
for the top layer of whatever you're doing, then call one of the library
routines.  If NumPy isn't fast enough for you, there is LaPack, or
odepack, or rksuite, or Nag, can't think of any others right now.  Most
are Fortran.  Unfortunatley, not all these are free.  I think you can
get LaPack lite for free.  Odepack is free and it's from llnl.  Nag is
NOT free, it is very expensive.  Travis Oliphant has a python-lapack
wrapper (http://oliphant.netpedia.net/).  We have some of Nag wrapped
(email me if you're interested).  Also, I can make a wrapper for odepack
if you need.

Even if you do this, there are other tricks you can do with sparse
matricies, and such things.  The above libraries have functionality for
sparse maticies.  How you set up the memory can be a factor too.  NumPy
will create arrays that are non-continous in memory.  This can affect
performance too. 

In summery, there are a lot of factors when it comes to speed. 
Sometimes it's hard to tell which factors will mater, and which won't.  
Good luck.

-- 
Curtis Jensen
cjensen at bioeng.ucsd.edu
http://www-bioeng.ucsd.edu/~cjensen/
FAX (425) 740-1451


From emile at fenx.com  Fri Dec 22 11:13:58 2000
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 22 Dec 2000 08:13:58 -0800
Subject: Newbie Question
References: <05m64tkqev0aavea80ttjse1sj2gd447hg@4ax.com> <hqs64tc5jjndav3p6nvmu67t7hqaf5go00@4ax.com>
Message-ID: <91vuj5$5n7is$1@ID-11957.news.dfncis.de>

I see what you mean.  I've built a test that appears to
work, but I don't immediately see the material difference
between this and what your code does.

import os

class Class:
    def func(self, root, files):
        print self
        print root
        print files

arg = Class()
root = r'C:\etvwin'

os.path.walk(root, Class.func, arg)

HTH,

--

Emile van Sebille
emile at fenx.com
-------------------


"Bob W." <removethis_kranki2 at earthlink.net> wrote in message
news:hqs64tc5jjndav3p6nvmu67t7hqaf5go00 at 4ax.com...
> I tried using "self" in the os.path.walk and I get an
error saying
> that there 3 expected arguments, but got 4 in ntpath.py,
line 264.
> That was my preferred way of simulating a C++ private
method.
> Thanks for any help provided!
>
> On Fri, 22 Dec 2000 13:44:23 GMT, Bob W.
> <removethis_kranki2 at earthlink.net> wrote:
>
> >I am trying to write the following class, but it fails in
the Run()
> >method reference to __CheckFile and I am not sure how to
get it to
> >work.  Thanks in advance for your help!
> >
> >class RmwExpireCheck:
> >    def __init__( self, szFileName, szDir ):
> >        self.szFileName = szFileName
> >        FileStatList = os.stat( szFileName )
> >        if FileStatList:
> >            self.iCheckDate = FileStatList[8]
> >        self.szDir = szDir
> >        self.szFileList = []
> >
> >    def __getitem__(self, i):
> >        line = self.szFileList[i]
> >        if line:
> >            return line
> >        else:
> >            raise IndexError
> >
> >    def __CheckFile( self, szDirName, szFileNames ):
> >        for szFileName in szFileNames:
> >            szPathName = os.path.join( szDirName,
szFileName )
> >            if os.path.isfile( szPathName ):
> >                iFileTime = os.stat( szPathName )[8]
> >                if iFileTime > self.iCheckDate:
> >                    self.szFileList.append( szPathName )
> >
> >    def Run( self ):
> >        os.path.walk( self.szDir,
RmwExpireCheck.__CheckFile, self )
> >
> >
>




From keisari_ at hotmail.com  Thu Dec  7 09:59:16 2000
From: keisari_ at hotmail.com (joonas)
Date: Thu, 07 Dec 2000 16:59:16 +0200
Subject: Printing modules contents
References: <3A2F84CC.3056C755@hotmail.com> <m3itow4cmi.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <3A2FA5C4.CC479A19@hotmail.com>

I tried this but it didnt worked.

<clip>
import sys
list = dir(sys)
for x in list:
  print x," = ",sys.x
</clip>

Joonas.


Michael Hudson wrote:
> 
> joonas <keisari_ at hotmail.com> writes:
> 
> > How can I make a script that would print modules contents like below,
> >
> > sys
> >   sys.__name__ = "sys"
> >   sys.stdout = "<PyShell.PseudoFile instance at 7fd530>"
> >   sys.winver = "1.5"
> >   sys.etc = "spam"
> 
> Think "dir" & "repr".  Shouldn't be hard.
> 
> Cheers,
> M.
> 
> --
> 39. Re graphics:  A picture is worth 10K  words - but only those
>     to describe the picture. Hardly any sets of 10K words can be
>     adequately described with pictures.
>   -- Alan Perlis, http://www.cs.yale.edu/homes/perlis-alan/quotes.html


From sholden at holdenweb.com  Sun Dec 17 14:07:41 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 17 Dec 2000 14:07:41 -0500
Subject: writing a file on my web site
References: <1D793E572F1CCC75.4177A7CD821DE2C4.6A38056B7E83389E@lp.airnews.net> <91ie42$shk$1@bob.news.rcn.net> <6BBCEDE25C4DD0D8.067F4BB38B52531A.4846B6E09FEA5EB2@lp.airnews.net>
Message-ID: <Kc8%5.29804$n51.589705@e420r-atl1.usenetserver.com>

Danger, danger, Will Robinson!  See below...

ron jeffries <ronjeffries at acm.org> wrote in message
news:6BBCEDE25C4DD0D8.067F4BB38B52531A.4846B6E09FEA5EB2 at lp.airnews.net...
> On Sun, 17 Dec 2000 08:12:58 -0500, "Satheesh Babu" <vsbabu at erols.com>
> wrote:
>
> >With FTP tools, you probably have write permissions. But
> >others might not have write permissions to that file. You'll
> >need that so that web server process has permissions to
> >write to it.
> >
> >chmod 666 should do it. Security alert - change the mode
> >only to your roncount.txt
>
> Thanks, I'll ask my ISP to change it. I don't seem to be able to do it
> from here ...

It is VERY UNSAFE to allow the cgi-bin directory to be writable by the web
server.  Probably the reason you cannot currently do it is because your web
service provider has set your server process to run with the identity of a
user with ownership of few resources and limited read abilities.

Imagine if you allow your web processes to write in cgi-bin.  In theory,
anyone who can get a file installed in that directory can put code in your
web which they can run at will.  Not terribly safe...

Try finding some other way to achieve what you want.

regards
 Steve





From philshinn at mediaone.net  Fri Dec 15 03:23:06 2000
From: philshinn at mediaone.net (Phil Shinn)
Date: Fri, 15 Dec 2000 08:23:06 GMT
Subject: stewpid newbie tricks
Message-ID: <3A39D4F9.648D8DBA@mediaone.net>

Problem: I can execute my runSpider.bat file in NT from D:\Program
files\python\mine\spider
using command line interface just fine - here's runspider.bat:

python "D:\Program files\python\mine\spider\mySpider.py"

Here's mySpider.py:

import urllib
urllib.urlretrieve('http://www.time.org','time.html')

this works nicely & copies the html page at www.time.org to D:\Program
files\python\mine\spider.

My problem is when I try to schedule this to run on a daily basis by
using the NT command line AT:

at 16:00 /Every:m,t,w,th,f,s,su "d:program
files\python\mine\spider\runSpider.bat"

nothing seems to happen, i.e., scheduler does not seem to kick in, no
file gets created, not sure what happened.

Clues for the clueless?

thanks



From pyth at relax.dk  Mon Dec 11 06:52:21 2000
From: pyth at relax.dk (pyth at relax.dk)
Date: Mon, 11 Dec 2000 12:52:21 +0100 (CET)
Subject: variable variables
In-Reply-To: <LIEMKHFIDGKFBCHAGODBEEADDLAA.badassgeek@earthlink.net>
Message-ID: <Pine.LNX.4.21.0012111250310.5044-100000@relax.dk>

Allright! thanx :o)
I hadn't thought of that :o)
I'm still a python newbie, and didn't use that kinda expressions in php
:o)
Thanx again :o)


> You could:
> 
> exec "print test%d" % (var1)
> 
> > What i was thinking of, was something like this:
> > 
> > test1 = 'test text 1';
> > test2 = 'test text 2';
> > 
> > var1 = 1
> > var2 = 2
> > 
> > print test{var1};
> > 
> > This should print: 'test text 1'
> > This example doesn't work, but is close how you would do it in php :o)
> > 

/Dan Larsen




From neilh at scintilla.org  Thu Dec 28 05:45:31 2000
From: neilh at scintilla.org (Neil Hodgson)
Date: Thu, 28 Dec 2000 10:45:31 GMT
Subject: [Pythonwin] Pywin encoding (again)
References: <0sNX5.168$M47.3490@m2newsread.uni2.es> <3A301090.3060304@ActiveState.com> <92f2t6$vji$1@nnrp1.deja.com>
Message-ID: <fRE26.35949$xW4.283280@news-server.bigpond.net.au>

jj:
> is there some workaround to write cp1250 scripts with pywin?

   You may want to turn off Unicode processing in PythonWin. Open
Pythonwin\pywin\scintilla\view.py, find the OnInitialUpdate method and
comment out the two lines under "# Enable Unicode if we can". Save the file,
shut down Pythonwin and restart it and you may prefer the results.

   Neil





From jp at ulgo.koti.com.pl  Wed Dec 27 12:14:34 2000
From: jp at ulgo.koti.com.pl (Jacek =?iso-8859-2?Q?Pop=B3awski?=)
Date: 27 Dec 2000 17:14:34 GMT
Subject: regex, split and ()
Message-ID: <slrn94k8rq.18t.jp@localhost.localdomain>

I think I understand how regex split works:

>>> d=re.compile(r'(aa)')
>>> a="bbaacceeaadd"
>>> d.split(a)
['bb', 'aa', 'ccee', 'aa', 'dd']

I want to use this regex:

<[^<>]*(".*")?>

so:

>>> w=re.compile(r'(<[^<>]*(".*")?>)')
>>> s="<html> one <br> two <img src=\"<blah\"> three </html>"
>>> w.split(s)
['', '<html>', None, ' one ', '<br>', None, ' two ', '<img src="<blah">',
'"<blah"', ' three ', '</html>', None, '']

works good, but why it double everything? probably becouse I used second (),
how to fix it?

-- 
Changes comes slowly 
But this is what you'll be
Somehow it snows in the summer  
And the sun is freezing me                     "Otherworld" - Ronnie James Dio


From mjais at web.de  Mon Dec  4 17:37:56 2000
From: mjais at web.de (markus jais)
Date: Mon, 4 Dec 2000 22:37:56 +0000
Subject: python and Mysql/Postgresql
Message-ID: <00120422375600.01422@eagle>

hello,

I am quite a newbie and I want to use
Python with Mysql and Postgresql.

I saw, that there are serveral modules, but I don't know
which one to choose.

can you tell me, which modules for Mysql and Postgresql
work best??
(and where to get it)

Is there something for Python like DBI is for Perl????

thanks a lot

markus



From qrczak at knm.org.pl  Fri Dec 29 15:02:40 2000
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 29 Dec 2000 20:02:40 GMT
Subject: 2001 Enchancement Wishlist
References: <3A4CEACB.220998A0@javanet.com>
Message-ID: <slrn94prf0.2q0.qrczak@qrnik.zagroda>

Fri, 29 Dec 2000 14:49:31 -0500, Raymond Hettinger <othello at javanet.com> pisze:

> 1.  List.append(elem) and List.pop() are great.  Let's add two more
>      methods, List.prepend(elem) and List.leftpop().

I would call the latter List.shift(). This name is used with this
meaning at least in sh, Ruby and Perl.

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


From owen at astroNOJNK.washington.edu.invalid  Tue Dec 12 13:26:55 2000
From: owen at astroNOJNK.washington.edu.invalid (Russell E. Owen)
Date: Tue, 12 Dec 2000 10:26:55 -0800
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com>
Message-ID: <915qlh$89fa$1@nntp6.u.washington.edu>

In article <3A2FFB8B.2030408 at zoologic.com>, Panu Viljamaa 
<panu at zoologic.com> wrote:

>For the needs of a project, I'm looking into the language Python.
>
>It seems to have many attributes similar to Smalltalk  in general such 
>as dynamic typing, OO with the keyword 'self' and bytecodes. I used to 
>think it was some kind of improved Perl. But by skimming a book I 
>discovered the following quite impressive list of features:
>
>1. Dynamically typed, object oriented
>2. Simple syntax. Concise but readable code. Blocks  (not 
>block-closures) require no braces around them since they are indicated 
>by indentation. Local variables need not be declared.
>3. Cross platform (with GUI), yet well integrated to Windows (COM, 
>D-COM, DDLs, C++,  TCP-IP, ODBC)
>4. 'Single dialect'  . There's no Python-V, Python-VW, Python-MT, 
>Python-X. There is just one Python with a single lineage of released 
>versions..
>5. Web-enabled, http (server included), ftp, news, pop3 etc.
>6. Modules and namespaces. Default arguments.
>7. Innovative gimmicks such as indexing the last element of array with 
>-1, the next to last by -2 etc.
>8. Reflective capabilities
>9. 'Business orientation', compared to say 'Squeak'  (which I dare say 
>feels a academic )
>10. FREE with a growing community of contributors
>
>So I was wondering if any of you Smalltalkers had tried Python and could 
>compare the experience to working with Smalltalk .

I have used both and like both very much.

I think of Python as a very advanced scripting language. Much cleaner 
and more regular than Perl, not as powerful as Smalltalk. It is a great 
language and should appeal to many Smalltalkers for smaller tasks.

- It has an impressive collection of features, and they are well 
implemented. Classes, exception handling, name spaces, collections, 
reflexive programming and networking are all built in and all cleanly 
implemented.
- It is free and runs on unix, Mac, PC
- It is well documented -- lots of good books out there.
- It tends to favor simplicity. Generally this is a good thing, though 
occasionally it is overdone. But if you find Perl or C++ too cluttered 
(as I do) then you definitely should give it a look.
- Nice function/method calling capabilities (positional parameters, 
named parameters, defaults).
- Development is overseen by one person, Giudo VanRossum, whose goal is 
to keep it a simple and usable language. (I very much wish Squeak had 
this kind of driving force behind it; it's a wonderful language by a 
brilliant team, but making a clean, usable language does not seem to be 
their main focus.)
- It has "normal" procedural flow control structures, which is one of 
the few things I miss in Smalltalk. That is my personal opinion, not an 
attempt to anger anybody.

On the other hand, it is not as powerful, clean and regular as 
smalltalk. Limitations include:
- The language is somewhat cluttered and irregular because it has 
procedures, functions and methods. Hence arbitrary decisions had to be 
made in some areas as to how to implement things, which can lead to some 
hunting through the manuals.
- It has lambda functions, but getting late binding behavior is too 
difficult (though doable).
- Only two scopes for variables: global (to a text file) or local (to a 
function/method).
- There are no class methods, though they can be faked reasonably well 
with modules.
- Basic built in types (e.g. integers, floats and lists) are not 
classes, in the sense that they do not have class methods.
- No boolean class; uses C's icky everything-is-a-logical idea.
- I personally find not having to declare variables a negative (too 
dangerous), though it's reasonable for simple scripting languages
- garbage collection is reference counted, though now enhanced in Python 
2.0 so it can handle most cases of circular references
- Collection classes are good, but no match for Smalltalk. (But then 
what language comes close to ST in this area?). It does have 
dictionaries and lists, both of which are great. But that's about it. It 
also has "tuple", a minor and annoying (to me) variant of a list that is 
immutable (more or less) and so can be used as a key in dictionaries.
- Mediocre GUI support (a mediocre interface to Tk and a different 
interface to wxWindows that isn't available on my platform--Mac)
- Mediocre IDE. I hope this will be improved over time, along with the 
GUI support.
- Rather ugly implementation of "package" (multi-file modules). This is 
one of the few areas where I find the implementation to be ugly and 
hastily done; generally it's a very clean language.
- Like most languages, it uses text files to contain sets of code. I 
really miss Smalltalk's more integrated approach.

It has a few minor annoyances, such as:
- One must explicity include the "self" parameter in method parameter 
lists. This is pointless (from a user's perspective, I'm sure there are 
implementation reasons) and much too easy to forget.
- Required colons at the beginning of sub-blocks are much too easy to 
forget. But easily fixed.
Having said that, the language rather stands out because it doesn't have 
more of these.

It also has some interesting oddities, such as:
- you can add a method or field to a class or instance of a class at any 
time; interesting but dangerous
- Code indentation is significant. This tends to annoy folks at first, 
but almost everybody comes to find it reasonable or good after a short 
time. It shocked me at first, but I like it just fine.

It's an absolutely stunning language for free, and although not as fast 
as Perl, still quite reasonable for basic work. Every Smalltalker who 
needs a smaller-scale language for small projects should look at it, but 
don't expect it to be a substitute for Smalltalk.

-- Russell

Having said that Python is not really a great substitute for Smalltalk, 
I will admit that I am working on a reasonable scale GUI project in 
Python because even though I adore ST I have doubts about ST's long-term 
viability (cincom seems to be back to ParcPlace's old tricks on insane 
pricing) and it'd be an incredibly hard sell to my management. I may be 
setting up for failure, due to Python's mediocre GUI, but I'm having a 
go at it.


From m.faassen at vet.uu.nl  Sat Dec 30 22:52:46 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 31 Dec 2000 03:52:46 GMT
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <slrn94shbe.150.grey@teleute.rpglink.com> <X8s36.2731$Qb7.305339@newsb.telia.net> <slrn94smds.150.grey@teleute.rpglink.com> <FOt36.17$vd1.2224@e420r-atl2.usenetserver.com> <slrn94t1vj.150.grey@teleute.rpglink.com> <3A4E95FD.D6C737A7@engcorp.com>
Message-ID: <92maie$cfo$3@newshost.accu.uu.nl>

Peter Hansen <peter at engcorp.com> wrote:
[snip]
> I suspect some of us would even argue that the conditional in 'do/until'
> can be difficult for some people to get right and that forcing the use of
> 'while/if/break' may actually increase the likelihood the resulting code
> will be correct the first time, and be interpreted correctly during
> maintenance.  My guess is that for this precise reason the alternative
> loop construct was removed from/never added to Python in the first place.

I suspect (and seem to recall hearing about this before) the reason is
more prosaic; how does one indent the darn thing? :)

do:
   ..
until x > 25

does not fit a pythonic pattern for statements.

do:
   ..
   until x > 25

doesn't either, unless 'until' is a new statement. But to fit the Python
pattern, it'd need to be usable at least anywhere within the loop, and
in that case it's just a synonym for:

   if x > 25: break

which we already have.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From tlownds at my-deja.com  Sun Dec 24 12:17:40 2000
From: tlownds at my-deja.com (tlownds at my-deja.com)
Date: Sun, 24 Dec 2000 17:17:40 GMT
Subject: PEP 222 draft
References: <slrn947bj1.1rs.amk@207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com>
Message-ID: <925b3i$5pm$1@nnrp1.deja.com>

In article <slrn947bj1.1rs.amk at 207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com>,
  akuchlin at mems-exchange.org wrote:
>     Utility function: build a query string from a list of 2-tuples
>

Perhaps it would be better to extend urllib.urlencode to handle this case. I
think it should be extended to handle lists as values as well:

>>> urlencode({'x': ['a', 'b']})
'x=%5b%27a%27,+%27b%27%5d'

would be:

'x=a&x=b'

>     The cgi.py module would be deprecated.  (XXX A new module or
>     package name hasn't been chosen yet: 'web'?  'cgilib'?)
>

cgilib gets my vote.

>     * An HTML generation module is not part of this PEP.  Several such
>       modules exist, ranging from HTMLgen's purely programming
>       interface to ASP-inspired simple templating to DTML's complex
>       templating.  There's no indication of which templating module to
>       enshrine in the standard library, and that probably means that
>       no module should be so chosen.
>

There are more ASP-like modules out there than other types...

>     Parsing functions: carry over most of the parse* functions from
>     cgi.py
>

It would be really nice to change the main parse-form-input function to
return a dictionary of plain strings instead of the FieldStorage instances.
When a file is uploaded, a FieldStorage instance for that value makes sense.
You've suggested this before, I may have missed it in your PEP.

>
>     # Module-level function
>     def wrapper(func, logfile=sys.stderr):
>         """
>         Calls the function 'func', passing it the arguments
>         (request, response, logfile).  Exceptions are trapped and
>         sent to the file 'logfile'.
>         """
>         # This wrapper will detect if it's being called from the command-line,
>         # and if so, it will run in a debugging mode; name=value pairs
>         # can be entered on standard input to set field values.
>         # (XXX how to do file uploads in this syntax?)
>

Regarding debugging: A command-line debugging ability is very useful! But why
tie that feature to the wrapper feature? You can put it in the
parse-form-input function. Also, allow name=value pairs as command-line
arguments. name=value pairs would still be read even with command-line
arguments, unless a final command-line argument "go" is present.

One debugging feature that would be very useful in a wrapper, is a way to
start pdb either right away or at an exception.

Your wrapper passes (request, response, logfile) - why not pass an object
with .request, .response, and .logfile attributes. Using a tuple now makes it
harder to extend later. A cgi.log-like method on that "context" object would
be useful too!

-Tony Lownds



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


From andreas at andreas-jung.com  Fri Dec 22 04:14:13 2000
From: andreas at andreas-jung.com (Andreas Jung)
Date: Fri, 22 Dec 2000 10:14:13 +0100
Subject: How use urllib.urlopen()?
In-Reply-To: <9012313F4gustaflalgonetse@194.213.69.148>; from gustafl@algonet.se on Fri, Dec 22, 2000 at 05:16:42AM +0000
References: <9012313F4gustaflalgonetse@194.213.69.148>
Message-ID: <20001222101413.B22256@yetix.sz-sb.de>

On Fri, Dec 22, 2000 at 05:16:42AM +0000, Gustaf Liljegren wrote:
> Can't get the urlopen() function to work. I have a script on a UNIX web 
> server that should go to another webpage when the user clicks a button. To 
> isolate the problem, I made a test script:
> 
> #!/usr/local/bin/python
> 
> import urllib
> u = urllib.urlopen("http://www.yahoo.com")
> print u.read()

This should work (and works on my machine) !
 
> When running this I get HTTP 500 - Internal server error. :-( 

Maybe a temporary problem at Yahoo...try another web site.

Andreas



From timd at macquarie.com.au  Wed Dec 27 17:02:53 2000
From: timd at macquarie.com.au (Timothy Docker)
Date: 28 Dec 2000 09:02:53 +1100
Subject: newbie : array initialization, anyone ?
References: <3a4a1352.9818548@news.t-online.de> <wgr26.1691$JG.213419@news.shore.net>
Message-ID: <m4i3df94j5e.fsf@macquarie.com.au>

arcege at shore.net writes:

> This is in the FAQ (#4.50).  You will want to create each inner dimension
> individually:
> 
> >>> multidim = [ None ] * 3
> >>> for i in range(3):
> ...   multidim[i] = [ 0 ] * 10
> ...
> >>>

or with list comprehensions in 2.0:...

>>> x = [ [0 for i in range(4)] for j in range(3) ]
>>> x
[[0, 0, 0, 0], [0, 0, 0, 0], [0, 0, 0, 0]]
>>> x[1][1] = 1
>>> x
[[0, 0, 0, 0], [0, 1, 0, 0], [0, 0, 0, 0]]
>>>


Tim


From nospam at no.spam  Wed Dec 13 19:52:23 2000
From: nospam at no.spam (Carsten Geckeler)
Date: Thu, 14 Dec 2000 01:52:23 +0100 (MET)
Subject: unwrapping argument tuples
In-Reply-To: <3A37D870.F3D01433@kla-tencor.com>
Message-ID: <Pine.LNX.4.10.10012140151270.1095-100000@erde.kosmos.all>

On 13 Dec 2000, Prashanth Mundkur wrote:

> is there a way to unwrap a tuple argument to a function and
> match it to the parameters?
> 
> that is, if
> 
> >>>lst = [1, 2, 3]
> 
> >>>def func (p1,p2,p3):
>        return p1+p2+p3
> 
> then I want to apply func somewhat as
> 
> >>>func(lst)
> 6
> 
> such that the elements of lst are unwrapped and and matched to p1, p2,
> p3
> 
> tx for any tips.
> 
> --prashanth

Well, just try:

>>> apply(func, lst)
6

Cheers, Carsten
-- 
Carsten Geckeler:  carsten dot geckeler at gmx dot de
To get proper email-address replace `dot' and `at' by the corresponding symbols.




From grey at despair.rpglink.com  Sat Dec 23 09:01:07 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Sat, 23 Dec 2000 14:01:07 -0000
Subject: coding convention conversion
References: <mailman.977516469.1157.python-list@python.org> <920v4s$oqj$2@taliesin.netcom.net.uk>
Message-ID: <slrn949c13.40l.grey@teleute.rpglink.com>

On Sat, 23 Dec 2000 01:29:00 +0000, Garry Knight <garryknight at bigfoot.com>
wrote:
>--------8X-cut here-----------------------------------------------
>#!/usr/bin/perl
>
>while (<>) {
>  if (/^\s*def\s+\w+/) {
>    s/([A-Z])/_\l\1/g;
>  }
>  print;
>}
>--------8X-cut here-----------------------------------------------

    Anyone else find it ironic that someone is solving a problem using Perl in
a Python group when the comparable Python would be an excellent exercise to
perform?  :)

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From MarkH at ActiveState.com  Fri Dec  8 01:25:23 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Fri, 08 Dec 2000 06:25:23 GMT
Subject: Implementing IPersistStorage using PythonCOM
References: <t308m58pkecra3@corp.supernews.com>
Message-ID: <3A307D8C.10909@ActiveState.com>

Edward Blakes wrote:

> CATID_PerspectiveAttributeTypeServer = pythoncom.MakeIID("{a guid}")
> CLSID_PythonSample = pythoncom.MakeIID("{a guid}")

This is probably a problem - it means a new GUID will be created each 
time it is run - definitely not what you want.  Paste generated GUIDs 
directly into the code.

Also - you _really_ should get my book ;-)  The discussion on debugging 
is what you need the most.

Adding this:

     import win32com.server.dispatcher
     useDispatcher = win32com.server.dispatcher.DefaultDebugDispatcher

Then changing your 2 "wrap" calls to:

     aServer = 
win32com.server.util.wrap(origObj,pythoncom.IID_IPersistStorage, 
useDispatcher=useDispatcher)

     # Test persistence
     aLockBytes = 
win32com.server.util.wrap(LockBytes(),pythoncom.IID_ILockBytes,useDispatcher=useDispatcher)

(note the new param - set "useDispatcher=None" when you are no longer 
debugging.

Then running you code results in the following to the "Trace collector":

in _InvokeEx_ with SetSize 0 1 (0,) None None
Traceback (most recent call last):
   File "f:\src\pythonex\com\win32com\server\dispatcher.py", line 58, in 
_InvokeEx_
     return self.policy._InvokeEx_(dispid, lcid, wFlags, args, kwargs, 
serviceProvider)
   File "f:\src\pythonex\com\win32com\server\policy.py", line 314, in 
_InvokeEx_
     return self._invokeex_(dispid, lcid, wFlags, args, kwargs, 
serviceProvider)
   File "f:\src\pythonex\com\win32com\server\policy.py", line 502, in 
_invokeex_
     return apply(func, args)
   File "F:\temp\delme.py", line 78, in SetSize
     return S_OK
NameError: name 'S_OK' is not defined


Which should give you a good indication of your problem!

Mark.



From grante at visi.com  Wed Dec 20 17:40:30 2000
From: grante at visi.com (Grant Edwards)
Date: Wed, 20 Dec 2000 22:40:30 GMT
Subject: Python in embedded systems?
References: <_w706.11237$TU6.1069571@ptah.visi.com> <mailman.977345156.5111.python-list@python.org>
Message-ID: <yza06.11263$TU6.1080153@ptah.visi.com>

In article <mailman.977345156.5111.python-list at python.org>, Trent Mick wrote:
>On Wed, Dec 20, 2000 at 07:12:58PM +0000, Grant Edwards wrote:
>> Has anybody heard of a Python interpreter being used in an
>> embedded system to provide user scripting capabilities?
>
>Don't know if it helps for your situation but I believe that the empeg car
>MP3 player embeds Python for driving its UI:
>
>http://www.empeg.com/

The empeg is running Linux, so I guess it's not surprising that
they can run Python.  I'm running eCos (an RTOS w/o virtual
memory and possibly without a filesystem) so there is an
extremely limited set of library calls available.

-- 
Grant Edwards                   grante             Yow!  I hope the
                                  at               "Eurythmics" practice birth
                               visi.com            control...


From aleaxit at yahoo.com  Thu Dec 14 05:14:38 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 14 Dec 2000 11:14:38 +0100
Subject: xreadlines (was Re: while true: !!!)
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <slrn93g37o.9tj.jepler@potty.housenet> <f3%Z5.146409$U46.4768601@news1.sttls1.wa.home.com>
Message-ID: <91a6it02flt@news2.newsguy.com>

"Greg Jorgensen" <gregj at pobox.com> wrote in message
news:f3%Z5.146409$U46.4768601 at news1.sttls1.wa.home.com...
    [snip]
> There is the fileinput module, which also solves the "ugly infinite loop"
> problem that seems to bother some people:

Or rather, it WOULD solve it, *IF*...:


> import fileinput
>
> for line in fileinput.input(filename):
>     print line
>
> The fileinput module can iterate over a single file, a list of files, or

...if it was able to iterate over already-opened file-like objects, rather
than having to open actual files itself, it WOULD conceptually solve it
all.  I wonder how hard it would be to generalize fileinput in this way --
probably not very (the in-place-rewriting option would be incompatible
with having already-opened file-like objects in the list, of course).

> stdin. I've used fileinput to go through big lists of files (10,000+ email
> messages) and it works great. It doesn't appear to do any buffering
> itself--it uses file.readline() to read the files.

If this was a performance problem, it could of course also be fixed
in a future fileinput version without changing code that uses it (again,
in-place-rewriting would probably have to inhibit the optimization,
although that isn't entirely clear).


Alex





From sandj.williams at gte.net  Wed Dec 20 11:59:09 2000
From: sandj.williams at gte.net (Steve Williams)
Date: Wed, 20 Dec 2000 16:59:09 GMT
Subject: real life examples
References: <u48q19.cth.ln@adsl.bapnt>
Message-ID: <3A40E67B.537317B@gte.net>

huber wrote:

> i try to teach python to my son (13 years old)!
> but instead of stupid for loops, classes nobody ever needs and so on .....
> are there some good real life examples a child could be interested in ?
>
> thanks
> udo

This toy parser is always fun to play with.  My version of an electric train.

It's not very Pythonic, but then I've written it in COBOL, Basic, FORTRAN, 360
assembler, Signetics 2650 ML, . . .

I got it from an ACM Surveys article in about 1969.

#=======================================
#Kids!  Be the first on your block to design your own language!
#Implement boolean data types!
#Create a 'while forever do' statement!

#=======================================
def digit(strByte):
  global show
  if show > 3:
    print "applying digit test to %s" % (strIn[p:p + l])
  if strByte >= '0' and strByte <= '9':
    return 1
  else:
    return 0

#========================================
def alpha(strByte):
  global show
  if show > 3:
    print "applying alpha test to %s" % (strIn[p:p + l])
  if strByte >= 'A' and strByte <= 'Z':
    return 1
  else:
    return 0

#========================================
def match(strTerminal):
  global strIn,p,show,clss
  l = len(strTerminal)
  if p + l - 1 > len(strIn): return 0
  if show > 3:
    print "Matching %s to %s, p: %d, l: %d" % (strIn[p:p + l],strTerminal,p,l)

  if strIn[p:p + l] == strTerminal:
    p = p + l
    return 1
  return 0

#========================================
def parse(goal):
  global strIn,p,show
  r = 0
  rules = lang[goal]
  while r < len(rules):
    if show > 1 and p < len(strIn):
      print 'Testing %s at %s' % (repr(rules[r]),strIn[p])

    if p >= len(strIn) - 1 and rules[r][0] == '<EOF>':
      test = 'true'
    elif lang.has_key(rules[r][0]):
      test = parse(rules[r][0])
    elif clss.has_key(rules[r][0]):
      if clss[rules[r][0]](strIn[p]):
        p = p + 1
        test = 'true'
      else:
        test = 'false'
    elif match(rules[r][0]):
        test = 'true'
    else:
      test = 'false'

    if test == 'true':
      if show > 0:
        print "match %s" % rules[r][0]
      action = rules[r][1]
    else:
      action = rules[r][2]

    if action == 'pass':
      return 'true'
    elif action == 'fail':
      return 'false'
    elif action == 'next':
        r = r + 1
    else:
      r = r + int(action)

#========================================
lang = {}

lang["<PRG>"] = (("<SAE>","next","fail"),("<EOF>","pass","fail"))

lang["<SAE>"] =
(("<AOP>","next","fail"),("+","+2","next"),("-","next","pass"),("<AOP>","-2","fail"))

lang["<AOP>"] =
(("<MOP>","next","fail"),("*","+2","next"),("/","next","pass"),("<MOP>","-2","fail"))

lang["<MOP>"] = (("<SGN>","+0","next"),
("<NUM>","pass","next"),("<VAR>","pass","next"), \
  ("(","next","fail"),("<SAE>","next","fail"),(")","pass","fail"))

lang["<SGN>"] = (("+","pass","next"),("-","pass","fail"))

lang["<NUM>"] = (("<GE1>","pass","next"),("<LT1>","pass","fail"))

lang["<GE1>"] =
(("<DIG>","next","fail"),("<DIG>","+0","next"),("<LT1>","pass","pass"))

lang["<LT1>"] =
((".","next","fail"),("<DIG>","next","fail"),("<DIG>","+0","pass"))

clss = {}
clss["<DIG>"] = digit
clss["<VAR>"] = alpha

strIn = "((A)+(A*B)*-23.6)"
p = 0
goal = "<PRG>"
show = 3

print strIn, goal, show
print parse(goal)





From moshez at zadka.site.co.il  Mon Dec 18 23:20:21 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Tue, 19 Dec 2000 06:20:21 +0200 (IST)
Subject: float to string conversion
In-Reply-To: <91lmt3$4b8$1@nnrp1.deja.com>
References: <91lmt3$4b8$1@nnrp1.deja.com>
Message-ID: <20001219042021.62C67A84F@darjeeling.zadka.site.co.il>

On Mon, 18 Dec 2000 19:00:54 GMT, deragon at 8d.com wrote:

>   I cannot figure out how to convert a float to a string (obviously,
>   I am new to python).

The ``str'' function will most types to strings.

str(1) --> "1"
str(1.0) --> '1.0'

But don't forget the formatting operator:

"here's a float: %s\n" % 1.0 --> "here's a float: 1.0\n"
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From kuncej at mail.conservation.state.mo.us  Tue Dec 19 10:22:03 2000
From: kuncej at mail.conservation.state.mo.us (Jeff Kunce)
Date: Tue, 19 Dec 2000 09:22:03 -0600
Subject: xml-rpc module question
References: <dmC%5.75166$yR4.2151642@news1.rdc1.tx.home.com>
Message-ID: <j7L%5.879$VF3.1528@news.more.net>

> I have taken the xml-rpc module and added the following to the call area:
>
> def CubeIt(x):
>    return x**3
>
> def call(self, method, params):
>    if method == 'CubeIt':
>    return CubeIt(params)
>

This is a wild guess, but did you define CubeIt() at the
global level of the module, or was it indented the same
as call() - making it a method of the RequestHandler class?

If the former, I'm not sure what the problem is.

If the latter, try:

    def CubeIt(self, x):
        return x**3

    def call(self, method, params):
        if method == 'CubeIt':
        return self.CubeIt(params)


  --Jeff





From paulp at ActiveState.com  Fri Dec  8 17:11:43 2000
From: paulp at ActiveState.com (Paul Prescod)
Date: Fri, 08 Dec 2000 14:11:43 -0800
Subject: Help! Installing MatPy for ActivePython2.0 Win2000
References: <3A3131D0.82E6AEDF@magnet.fsu.edu>
Message-ID: <3A315C9F.93A5462F@ActiveState.com>

Khaled Khairy wrote:
> 
> If anybody could advise me.
> I am new to Python . I need it for scientific computations;
> manipulations of matrices and linear Algebra problem solving.
> How can I get MatPy to work (with the Cephes extention) on ActivePython
> 2.0 under Win2000 ? Is there another matrix package I could use?
> Please help! Thanks

If MatPy does not ship with Windows binaries they you will to make those
yourself or get someone to make them for you. You could also look at
NumPy which is already compiled into a Windows binary.

 Paul Prescod



From anton at vredegoor.doge.nl  Sun Dec 17 12:14:44 2000
From: anton at vredegoor.doge.nl (Anton Vredegoor)
Date: Sun, 17 Dec 2000 18:14:44 +0100
Subject: Changing IDLE's stdout color at runtime
Message-ID: <91isi7$hup$1@news.hccnet.nl>

Hello,

I am looking for something like:

def SetIdleStdoutColor(color):
	pass  # this function needs defining

print 'hello world'
SetIdleStdoutColor('green') 
print 'hello again'
SetIdleStdoutColor('defaultcolor')

Has anybody got an idea how to do this?

Anton.



From othello at javanet.com  Mon Dec  4 11:53:51 2000
From: othello at javanet.com (Raymond Hettinger)
Date: Mon, 04 Dec 2000 11:53:51 -0500
Subject: Calculating factorial
References: <3a2a9496.647995@uutiset.nic.fi>
Message-ID: <3A2BCC1F.F308E511@javanet.com>

I like your version just fine.  It's simple, clear, and it work's.
Do consider changing line 2 to:  result = 1L so you can
handle fact(50).

If you want to make something hard out of something simple, try:
 def fact(n): return reduce( lambda x,y:x*y, range(2,n+1), 1L )

Raymond

Olli Rajala wrote:

> Hi!
> I've learned a few time Python and programming.  I wrote a function
> that calculates a factorial of N and I was wondering if it's the best
> (I think it's not. =) way to do that. I'm glad to hear your comments.
>
> <script>
>
> def factorial(f):
>     result = 1
>     for number in range(1, f + 1):
>         result = result * number
>     return result
>
> print factorial(10)
>
> </script>
>
> Ps. Sorry my poor English.
> --
> Olli Rajala
> A Finnish student who loves Linux.



From MarkH at ActiveState.com  Tue Dec  5 20:08:42 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Wed, 06 Dec 2000 01:08:42 GMT
Subject: Global object in ASP (calling other scripting languages in ASP pages)
References: <90judk$edp$1@troll.powertech.no>
Message-ID: <3A2D904F.1040905@ActiveState.com>

Syver Enstad wrote:

> If you qualify your function calls with ScriptingNamespace you get the
> string of the foreign function instead of calling it, but at least the
> function is visible to Python. If I could manage to call it too I'd be very
> pleased.

I guess that would be a JS implemted function?

Try printing the repr() of your object - you may find it is not a string 
at all, but itself a COM object.  You may be able to just call this object!

Try writing the test function in a different language to see if that helps.

Mark.



From alan.gauld at gssec.bt.co.uk  Mon Dec  4 09:54:47 2000
From: alan.gauld at gssec.bt.co.uk (Alan Gauld)
Date: Mon, 04 Dec 2000 14:54:47 +0000
Subject: good newbie books?
References: <D_jU5.25243$xK.887693@news20.bellglobal.com> <uvmU5.21820$L26.6509113@typhoon.southeast.rr.com>
Message-ID: <3A2BB037.697FA5BF@gssec.bt.co.uk>

Don Tuttle wrote:
> > I just started Python and it is my first programming language.
> > Instead of constantly posting to this group for help I'm going to buy a
> > book.
> > Can anyone recommend a Python book that covers everything in a
> > newbieized way?

First can I strongly recommend joining the python-tutor mailing
list 
- sign up via python.org web site. It is specifically designed 
for beginners and you'll get a lot more questions etc at your 
level than on c.l.python.


> Hi Georges.  Unfortunately there are not any absolute beginners book, good
> or bad, in Python.  It's a hole that, IMO, desperately needs to be filled.

My new book:

"Learn to program Using Python"

is aimed at complete beginners and should be available next week
(according to Amazon! They don't tell me these things I'm only 
the author....:-)

Ivan Lanningham's "Teach Yourself Python in 24 hours" is also 
aimed at beginners but takes a slightly faster pace than mine.
It covers things like GUI programming more thoroughly but 
has less explanation of concepts.

You pays your money and you takes your choice!

Alan G.

-- 
=================================================
This post represents the views of the author 
and does not necessarily accurately represent 
the views of BT.


From dsh8290 at rit.edu  Thu Dec 14 08:37:52 2000
From: dsh8290 at rit.edu (D-Man)
Date: Thu, 14 Dec 2000 08:37:52 -0500
Subject: Import Error: no module named Java
In-Reply-To: <918ed8$vfq$1@nnrp1.deja.com>; from chicago_george@my-deja.com on Wed, Dec 13, 2000 at 13:16:17 -0500
References: <918ed8$vfq$1@nnrp1.deja.com>
Message-ID: <20001214083752.G21300@westsidecnc.rh.rit.edu>

Hmm, I don't think the Python interpreter provides that kind of access to the
java modules, but I could be wrong.  Check to make sure you have a *Python*
module named 'java'.

-D

On Wed, 13 Dec 2000 13:16:17 chicago_george at my-deja.com wrote:
 | I'm using JPython 1.1 on Java ver 1.1
 | Does anyone have any idea why I get the error msg
 | 
 | ImportError: no module named java
 | 
 | I get that when I try doing something like;
 | >>> from java import awt
 | 
 | WHY??!!  Anyone.. please??
 | 
 | Thanks!
 | George
 | 
 | 
 | Sent via Deja.com
 | http://www.deja.com/
 | -- 
 | http://www.python.org/mailman/listinfo/python-list
 | 




From gregj at pobox.com  Sat Dec 30 23:57:55 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Sun, 31 Dec 2000 04:57:55 GMT
Subject: HTML Parser
References: <mailman.978224102.28797.python-list@python.org>
Message-ID: <n1z36.208621$U46.6407059@news1.sttls1.wa.home.com>

"Voitenko, Denis" <dvoitenko at qode.com> wrote:

> I am trying to write an HTML parser.

This has been done--look at the htmllib and sgmllib modules.

> I am starting off with a simple one
> like so:
>
> # html_parser.py
> import re
> import string
>
> newline=re.compile('\n')
> HTMLtags=re.compile('<.*>')

.* will match as many characters as possible, including (in your case) < and
>. You want this pattern, which will match as few characters as possible
surrounded by < and >:

    HTMLtags = re.compile('<.*?>')

You can split using a literal character instead of a regular expression:

    line = lines.split('\n')

The readlines() method the file object will save you the trouble, but you
don't need to split the input into lines at all if you just want to find the
HTML tags.

Here's my version:

# simple html tag processor

import sys
import re

rx = re.compile('(<.*?>)', re.MULTILINE)

# HTML text will come from a file.read()
html = '<html>\n<head>\n\t<title>Page Title</title>\n</head>\n<body>Hello,
World!</body>\n</html>\n'

# split the text into tags and stuff between tags
# the re.split() creates empty list elements for adjacent matches--those can
be ignored
# uppercase anything inside <..> and output the converted text

for s in rx.split(html):
    if s == '':            # re.split() artifact
        continue
    elif (s[0] == '<') and (s[-1] == '>'):    # <tag>
        sys.stdout.write(s.upper())
    else:          # everything else
        sys.stdout.write(s)

--
Greg Jorgensen
PDXperts
Portland, Oregon, USA
gregj at pobox.com





From Blair.Lowe at compeng.net  Mon Dec 18 03:19:51 2000
From: Blair.Lowe at compeng.net (Blair Lowe)
Date: Mon, 18 Dec 2000 01:19:51 -0700
Subject: while true: !!!
In-Reply-To:  <B43D149A9AB2D411971300B0D03D7E8B0FBCBB@natasha.auslabs.avaya.com>
References:  <B43D149A9AB2D411971300B0D03D7E8B0FBCBB@natasha.auslabs.avaya.com>
Message-ID: <v04220801b663769e1800@[192.168.11.33]>

At 11:03 +1100 2000/12/18, Delaney, Timothy wrote:
>This is precisely the reason this thread exists ...
>
>The idiom is actually
>
>while 1:
>
>	do something
>
>	if condition:
>		break
>
>	do something else
>
>Basically, it's for dealing with situations where you can't tell at the
>beginning or end of a loop if the loop is completed - only somewhere in the
>middle. The canonical example is


This is exactly what a python program should be NOT doing. "while 1" 
is one of the most cryptic statements in Computer Science. other 
languages are known for this stuff, but I do not know why python 
doesn't auto define true and false, so that we don't have to write 
unreadable stuff.

One of the points of this language is to DECRYPT some of the poor 
coding styles out there. The indentation rules are a good example.

While "while 1" (pardon the pun) is a standard fixture that anyone 
with any experience knows what is going on, the whole point is to 
make all our code readable so that we, and others can maintain it and 
reuse it easily.

Imagine the newbie looking at this statement: while 1 ... hmmm must 
be the first condition of a while loop ????

It is cheap to define true and false once, or (my favourite) define 
ever (;;). Although the definition is cryptic, the resulting code is 
very readable. That is what we want.

I find that some python programs have very little documentation yet 
they are simple enough to read that a SMALL reduction in 
documentation is possible unless we start writing perl and C in 
python.

>  >
>  > PS.  I haven't been following most of this thread, so forgive me if
>  > this has already come up.
>  >

ditto,

TTYL,
Blair.
Computer Engineering Inc. http://www.compeng.net
Phone: 780 499 5687 (9 - 5 MST) Fax:   780 435 0693 (24 Hours)



From robin at jessikat.fsnet.co.uk  Sat Dec  9 21:13:24 2000
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sun, 10 Dec 2000 02:13:24 +0000
Subject: sgmllib nit
References: <eyErVLAEHlM6EwA4@jessikat.fsnet.co.uk>
 <EAsY5.2808$O%.435892@newsc.telia.net>
Message-ID: <JagDRIAEbuM6EwxY@jessikat.fsnet.co.uk>

In article <EAsY5.2808$O%.435892 at newsc.telia.net>, Fredrik Lundh
<fredrik at effbot.org> writes
>Robin Becker wrote:
>> The current version of sgmllib doesn't allow the detection of attributes
>> without values.
>
>as stated in the documentation, the current version of sgmllib
>only handles enough SGML to be able to parse HTML.
>
>in HTML, attributes without values (minimized form) are treated
>as boolean attributes.  boolean attributes can have one of three
>formats:
>
>    attribute="attribute" (full form: true)
>    attribute (minimized form: true)
>    (not present: false)
>
>if you need to tell the difference between "true" and "true", you're
>clearly not working with HTML (or anything compatible with HTML).
>
></F>
>
>
I realise that it does that! Unfortunately Zope's dtml isn't HTML.

Which parser should one use to analyse dtml which can contain tags like
<dtml-var standard_html_header>?

The above analysis won't work easily if I have <tag X> as then I'm
presented with attribute 'x' having value 'X'.

I merely want to copy the input to the output whilst making some trivial
changes to the URL's in <A href=URL> style tags. I guess I need to write
my own parser that doesn't make unnecessary assumptions.
-- 
Robin Becker


From n8gray at nospam.nospam  Thu Dec 21 17:09:32 2000
From: n8gray at nospam.nospam (Nathaniel Gray)
Date: Thu, 21 Dec 2000 14:09:32 -0800
Subject: A Suggestion for Python Colon Syntax
References: <91tc8h$2pp$1@hecate.umd.edu>
Message-ID: <91tvd8$2fm@gap.cco.caltech.edu>

William Djaja Tjokroaminata wrote:

> Hi all,
> 
> I have been using Python for some time, and I have a suggestion for the
> Python Colon (:) syntax.
> 
> [snip suggestion for removing the colon (what an evocative sentence!)]
> 
> Any comments, anyone?

I understand this problem of forgetting to care for your colons, but 
perhaps seeking to change the language is the wrong path.  Although "colon 
neglect" is a common problem, I don't think you'll find a lot of support 
for purging the colon, particularly from Guido.  Without Guido's support 
there's just no hope of changing Python's syntax.

What you might consider instead is customizing your text editor.  Write a 
macro that gets executed every time you hit enter that does the following 
with the previous line:
1.  If the first word is not {if, for, def, etc...} then return
2.  Ignore all comments for the following three steps
3.  If there is a colon on the line already, return
4.  If the last character on the line is "\" then return
5.  If the parens, brackets, and braces don't balance then return
6.  Insert a colon after the last non-comment character

Note that this is just a first stab at the problem and it definitely won't 
always work right (e.g. multi-line strings), but it will cure 90% of colon 
neglect cases without a Python parser or a costly hospital visit.  
Extending it to work for multi-line "colon statements" is left as an 
excercise for the reader.

If you're stuck with an editor that you can't or don't know how to tweak, 
try NEdit (http://www.nedit.org) instead.  It has a nice Python mode and is 
imminently scriptable.

maybe-they-shoulda-called-it-the-appendix-instead-ly y'rs  -  n8

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



From Bill.Scherer at VerizonWireless.com  Wed Dec 27 09:52:43 2000
From: Bill.Scherer at VerizonWireless.com (Bill Scherer)
Date: Wed, 27 Dec 2000 09:52:43 -0500
Subject: Controlling the Parallel Port
References: <mvuj4tkmn22gbth1a6t8fag39902qeeu9l@4ax.com>
Message-ID: <3A4A023B.632CAAC2@VerizonWireless.com>

Simon

There is a parallel port library for linux at

http://www.circlemud.org/~jelson/software/parapin/

that is simple to SWIG into a Python module.
I can send you the .i file for this if you want.

I haven't really used it, but it loads and reads the port status just
fine.


-Bill

Simon Faulkner wrote:
> 
> Is it easy to control (read/write) the Parallel Port in linux using
> Python?
> 
> I am hoping to use it to drive some simple electronics
> 
> (Usual disclaimer understood - If it smokes then you probably got the
> wiring wrong!)
> 
> Simon
> Simon Faulkner
> --
> http://www.python.org/mailman/listinfo/python-list



From sifu at atnet.at  Wed Dec 13 08:14:02 2000
From: sifu at atnet.at (Siegmund Fuhringer)
Date: Wed, 13 Dec 2000 14:14:02 +0100
Subject: I dont remember the name of the project...
In-Reply-To: <917s2u$ibr$1@merlin.rembrandtstr.bocholt.de>; from magnus.heino@rivermen.se on Wed, Dec 13, 2000 at 02:03:39PM +0000
References: <917s2u$ibr$1@merlin.rembrandtstr.bocholt.de>
Message-ID: <20001213141402.B2067@office.atnet.at>

On Wed, Dec 13, 2000 at 02:03:39PM +0000, Magnus Heino wrote:
> 
> 
> Hi.
> 
> A few days ago, I visited the website of a project that provided a 
> framework like CORBA, but much more lightweight, and written in python.
> 
> I dont remember the name of the project, anyone know what I may be refering 
> to??

hmm... maybe:

http://www.xs4all.nl/~irmen/ap/pyro.html
or:
http://casbah.org/Scarab/

bye sifu



> 
> /Magnus
> -- 
> http://www.python.org/mailman/listinfo/python-list

-- 
siegmund fuhringer

at-net /------
------/ a-team

jabber-id: sifu at jabber.at
--



From db3l at fitlinxx.com  Wed Dec 20 15:53:14 2000
From: db3l at fitlinxx.com (David Bolen)
Date: 20 Dec 2000 15:53:14 -0500
Subject: Catching Database Exceptions (Sybase)
References: <3A40EEBF.3262040F@ix.netcom.com>
Message-ID: <ulmtahl1h.fsf@ctwd0143.fitlinxx.com>

Thomas Gagne <tgagne at ix.netcom.com> writes:

> And miraculously, even with the table missing, it ran the "else".
> 
> I must be missing something significant here.  Any help?

You might try the same operation via some other tool (e.g., ISQL or
equivalent).  It's my understanding from some of the DB folks here
that depending on the type of error, errors within stored procedures
may sometimes be simply absorbed and not reported - you just get back
an empty result.  I know I've run some stored procedures that had bad
references to fields that didn't exist and didn't actually get a
database error from the execution, just no result.

And to answer your follow-up post question, the Warning and Error
classes that the DB API specifies all exceptions must be subclassed
from are themselves DB API-specific exceptions (both are subclasses of
the standard Python exception StandardError).

So in your exception clauses try using "<module>.Warning" instead of
just "Warning" (where <module> is your DB API compliant module).

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From magnus.heino at rivermen.se  Fri Dec  8 13:55:09 2000
From: magnus.heino at rivermen.se (Magnus Heino)
Date: Fri, 8 Dec 2000 19:55:09 +0100
Subject: xml.dom
References: <90q292$k9k$1@merlin.rembrandtstr.bocholt.de> <3A3100FD.D6C09C1D@ogbuji.net>
Message-ID: <90raqe$2f6$1@merlin.rembrandtstr.bocholt.de>

> Magnus Heino wrote:
> 
> > How do I build an xmldocument using something in xml.dom?
> 
> Here's an example:
> 
> from xml.dom import implementation

Ah, great. Thanks. That helped a lot. I'm getting confused by the different 
packages.
 
> Does that help, or do you need any other info?
> 
> Big improvements in the docs are already in the works.

Maybe you could tell me the differences and relations between the xml 
included in Python2.0, PyXML and Fourthoughs distro?

Seems to me that the Python2.0 is a subset of PyXML (because it wasnt 
finished in time for the release maybe?) and that PyXML includes things 
from Fourthught, but not of all of it.. for some reason.

Am I getting it all wrong? :)

Are there any good books to recommend in this area? 

/Magnus


From dag at orion.no  Wed Dec  6 03:29:05 2000
From: dag at orion.no (Dag Sunde)
Date: Wed, 06 Dec 2000 08:29:05 GMT
Subject: win32evtlogutil: How does it works ?
References: <3A2D556F.6240C661@club-internet.fr> <3A2D8223.4070408@ActiveState.com>
Message-ID: <lNmX5.963$K6.104400@juliett.dax.net>

The dll-part of the EventLog _is_ a little bit strange... :-)

You see... The number you give when logging an event is supposed
to reference an id of a string-recource in the dll you registered

So you have to create a resource-file with all the strings(messages)
you need, include it into a dll, and then reference those strings
by their id from the outside. This is done for nationalization-
purposes, I think...

Dag.

"Mark Hammond" <MarkH at ActiveState.com> wrote in message
news:3A2D8223.4070408 at ActiveState.com...
> G P wrote:
>
> > Any ideas ?
>
> Not really - but you are not alone.  I have the same problem, and havent
> been able to track it down.  The PythonService support for the event log
> works fine, and I can't see the difference between the 2 :-(
>
> I am sure it is just a registry setting, but I can't see what it is...
>
> Mark.
>




From aleaxit at yahoo.com  Mon Dec 18 04:12:42 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 18 Dec 2000 10:12:42 +0100
Subject: adding methods on the fly (was Re: Smalltalk and Python)
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <tcmay-FCA22F.14073713122000@news.got.net> <918v7a02scn@news1.newsguy.com> <tcmay-04E80C.16181313122000@news.got.net> <91a1ed02a4j@news2.newsguy.com> <3A389B1A.4F33FFD4@cfmu.eurocontrol.be> <91ae8s02llf@news2.newsguy.com> <3A397778.2E6D84D3@cosc.canterbury.ac.nz> <91c7r4$7a0$1@slb1.atl.mindspring.net> <3A3D8815.801E773D@cosc.canterbury.ac.nz>
Message-ID: <91kkfc01v06@news1.newsguy.com>

[Note: we've already been kindly asked to stop posting
this thread to fj.comp.lang.ruby, using comp.lang.ruby
instead, so I edited the newsgroups line accordingly]

"Greg Ewing" <greg at cosc.canterbury.ac.nz> wrote in message
news:3A3D8815.801E773D at cosc.canterbury.ac.nz...
> Andrew Dalke wrote:
> >
> > >>> bed = make_a(Furniture, Surface, Scenery)
>
> But how do I give it methods?

Most easily, actually -- in any of several handy ways.


>   bed(Furniture, Surface, Scenery):
>
>     def lie_on(self):
>       print "It's so comfy, you fall instantly" \
>         " asleep, and dream of a golden key sitting" \
>         " on a shelf in the pantry."

Using my version of make_a, i.e., to recap:


import new

def make_a(*classes, **attribs):
    return new.classobj('blob', classes, attribs)()

class Furniture:
     def comfy(self): return 1

class Surface:
    def genus(self): return 0  # Imagine a spherical cow...

class Scenery:
    def description(self): return "boring"

bed = make_a(Furniture, Surface, Scenery)


you have a range of choices, e.g., the most explicit:

def anewmethodjustforbed(self):
    print "elegant and comfy"
bed.lie_on = new.instancemethod(anewmethodjustforbed,
    bed, bed.__class__)


or, a tad less explicitly but perhaps more simply:

bed.__class__.lie_on = anewmethodjustforbed

which takes advantage of the implicit function->method
coercion [which happens when a reference to a function
is bound to a *class*'s attribute -- not an _instance_
attribute], and of course of the fact that instance 'bed'
is the sole instance of its class, so, in order to affect
that instance's behavior, we're at liberty to tweak its
class's:-).

It's just an issue of preferred style.  If one is using
the new module to build the class for object 'bed', I
think it might be more consistent to keep using the same
module also for building new methods to assign on the
fly; if one is using Andrew's original suggestion and
building that class less explicitly, then the implicit
coercion based approach might be better.


Note that my version of function make_a takes keyword
arguments and turns them into class-attributes for
the new class, so, if you already know what methods
you want to add to 'bed' as you generate it, another
alternative is to take advantage of that:

bed = make_a(Furniture, Surface, Scenery,
        lie_on=anewmethodjustforbed)


Of course you'll want to factor things appropriately.
If you have many methods that just print out stuff
while ignoring self, or using self just for read
access to instance data attributes, etc, you might
build a function encapsulating that:

def printout(self, message):
    print message % self.__dict__

and use it to build all on-the-fly methods of this
kind:

bed = make_a(Furniture, Surface, Scenery,
        lie_on=lambda self: printout(self, "whatever"))

or you might prefer a closure-factory approach,
rather than the lambda approach:

def printout_maker(message):
    def printout(self, message=message):
        print message % self.__dict__
    return printout

bed = make_a(Furniture, Surface, Scenery,
        lie_on=printout_maker("whatever"))

or, of course, the closure-factory could also use
module new rather than the implicit approach I've
taken here (which strictly parallel's Andrew's
original suggestion for make_a, except that here
we're using a local function rather than a local
class as in his case).


I think that some one out of these approaches, or
some variant thereof, will probably meet all kinds
of "how do I add a method" requirements.


Alex





From moshez at zadka.site.co.il  Tue Dec 26 10:42:18 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Tue, 26 Dec 2000 17:42:18 +0200 (IST)
Subject: classes (was Re: Same again please for OOP)
In-Reply-To: <928m5l02iuo@news1.newsguy.com>
References: <928m5l02iuo@news1.newsguy.com>, <928569$6ckoi$1@ID-11957.news.dfncis.de>, <92708d0g4g@news1.newsguy.com>, <mailman.977623149.8184.python-list@python.org> <9257mf0sl1@news1.newsguy.com> <roy-E5B925.13324224122000@news.panix.com> <925khi019hs@news1.newsguy.com> <kubsu1t81x.fsf@lasipalatsi.fi> <mailman.977739488.28224.python-list@python.org> <mailman.977785029.26926.python-list@python.org>
Message-ID: <20001226154218.E87F4A84F@darjeeling.zadka.site.co.il>

On Tue, 26 Dec 2000, "Alex Martelli" <aleaxit at yahoo.com> wrote:

> That would be handy, and I like 'Boost Python' approach to that --
> naming conventions rather than complex correspondences (see
> www.boost.org, but i guess C++-loathers would rather ignore it:-).

In spite of your warning, I went to the site, but couldn't figure
out the relevance. Can you summarize what is the Boost Python approach?
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From erno-news at erno.iki.fi  Tue Dec 12 13:39:05 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 12 Dec 2000 20:39:05 +0200
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl>
Message-ID: <ku4s098oxi.fsf@lasipalatsi.fi>

In article <slrn93cotl.80c.scarblac at pino.selwerd.nl>,
scarblac at pino.selwerd.nl (Remco Gerlich) writes:

| Integers, longs, complex numbers, strings, functions, tuples, types, floats
| and None are immutable. Most of these are obvious; the number 3 is the
| number 3, why would you change that?!

it is easy to think that when you know how things work in python.
but, you could just as easily say, "a list containing the numbers 1, 2 and
3 is a list of 1, 2 and 3, why would you change that?!"
mutable integer or string objects are certainly not an impossible idea.

  -- erno


From news at dorb.com  Sun Dec 10 13:03:42 2000
From: news at dorb.com (Darrell)
Date: Sun, 10 Dec 2000 18:03:42 GMT
Subject: reloading modules question
References: <90scdf$hpe$1@troll.powertech.no>
Message-ID: <2APY5.12112$hp.1409369@typhoon.nyroc.rr.com>

I haven't made extensive use of reload, and want to hear how it goes for
you.
Although killing Python would clean up memory leaks and the like.
I assume IIS would restart it, maybe??

# Hacked from the win32 demos
import win32api, win32pdhutil, win32con#, sys

def killProcSelf():
    myPid= win32api.GetCurrentProcessId()
    handle = win32api.OpenProcess(win32con.PROCESS_TERMINATE, 0,myPid)
    win32api.TerminateProcess(handle,0)

if __name__ == '__main__':
    killProcSelf()
    raw_input("Failed to kill self")

"Syver Enstad" wrote:
> Is it safe to loop through sys.modules.items()
> reloading each module, as long as I catch the ImportError?
>





From erno-news at erno.iki.fi  Sun Dec 24 13:43:14 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 24 Dec 2000 20:43:14 +0200
Subject: Python Database Interfaces - Any Standard?
References: <RUo16.204045$RP.37382354@news1.rdc1.md.home.com>
Message-ID: <kuk88pu0cd.fsf@lasipalatsi.fi>

>>>>> "Brian" == Brian Knox <brianknox at home.com> writes:

| One huge question I have about Python is: is there any sort of
| standard for database access with Python? 

yes, db-api, see <URL:http://www.python.org/topics/database/>.

| I've been coding Perl DBI apps as long as I've been coding perl, and
| the DBI module is going to be sorely missed by me. Are all of the
| Python database interfaces database specific, or is there an
| abstraction layer similiar to DBI?
| Is there any one DB module for Python that has been adopted as a
| standard by the community?

i don't know anything about dbi. db-api is what the name says,
a standard api that different modules implement.

it's ofcourse not possible to have a single module cover all databases
since the various databases have different protocols for talking to
them (modulo odbc/jdbc). well, unless the single module contained code
for every conceivable database, but that would hardly be sensible.

  -- erno




From guido at python.org  Fri Dec 15 15:11:45 2000
From: guido at python.org (Guido van Rossum)
Date: Fri, 15 Dec 2000 15:11:45 -0500
Subject: no current thread warning
Message-ID: <200012152011.PAA03302@cj20424-a.reston1.va.home.com>

Charles Crain:

> I, like many before me, and writing a python C module that wraps a library
> that issues callbacks from a thread created from within C.  I have, of
> necessity, become acquainted with all the
> PyThreadState/PyEval_AcquireThread(), etc. stuff in order to do this without
> crashing the interpreter.
> 
> However, I am disturbed by the following warning, which is issued whenever
> one makes certain calls into the threading module from a callback within
> python:
> 
> "currentThread(): no current thread for XXXX", where XXXX (I presume) is a
> thread id.
> 
> Near as I can tell, this relates to the statement in the documentation for
> the threading.currentThread() function, which states something to the effect
> of "for threads not created by the threading module, a dummy thread with
> limited capabilities will be created."
> 
> A have a couple questions regarding this:
> 1.  Exactly how are the capabilities limited?  Are there any gotchas I
> should be on the lookout for?
> 2.  Why is the warning printed out unconditionally?  IMHO, it should only be
> printed out if the error is potentially fatal, and/or amounts to an asserion
> failure within the interpreter.  If I interpret the dox correctly, this
> message could at best be considered an informational, and should therefore
> be able to be turned off.  It is only cosmetic, but in considering whether
> to commercially release products we have written in Python, it is an issue.

Guilty as charged.  That warning was merely a debugging aid that stuck
around by mistake!  You can safely ignore it or comment it out; I'll
remove it from the next release.

The limitations of dummy threads are explained in a comment in
threading.py:

# Dummy thread class to represent threads not started here.
# These aren't garbage collected when they die,
# nor can they be waited for.
# Their purpose is to return *something* from currentThread().
# They are marked as daemon threads so we won't wait for them
# when we exit (conform previous semantics).

--Guido van Rossum (home page: http://www.python.org/~guido/)



From dag at orion.no  Thu Dec  7 03:51:11 2000
From: dag at orion.no (Dag Sunde)
Date: Thu, 07 Dec 2000 08:51:11 GMT
Subject: win32evtlogutil: How does it works ?
References: <3A2D556F.6240C661@club-internet.fr> <3A2D8223.4070408@ActiveState.com> <lNmX5.963$K6.104400@juliett.dax.net> <3A2E2CD7.2090602@ActiveState.com>
Message-ID: <3cIX5.1329$K6.146134@juliett.dax.net>

Ah!

I was out of line here... (Never used the python module)
I just remembered doing it in VB once, and had that
problem...

Dag.

"Mark Hammond" <MarkH at ActiveState.com> wrote in message
news:3A2E2CD7.2090602 at ActiveState.com...
> Dag Sunde wrote:
>
> > The dll-part of the EventLog _is_ a little bit strange... :-)
> >
> > You see... The number you give when logging an event is supposed
> > to reference an id of a string-recource in the dll you registered
> >
> > So you have to create a resource-file with all the strings(messages)
> > you need, include it into a dll, and then reference those strings
> > by their id from the outside. This is done for nationalization-
> > purposes, I think...
>
> Yes - but win32evtlog.pyd _should_ be such a DLL.  It should magically
> register itself as a valid source of such messages.
>
> All of the messages it comtains are generic - eg, often simply of the
> form "%1" - meaning that they are pretty pointless from an
> internationalization point-of-view - but they _should_ still work (and
> have been known to) for the simple use they were intended.
>
> Mark.
>




From dsh8290 at rit.edu  Fri Dec 22 16:55:27 2000
From: dsh8290 at rit.edu (D-Man)
Date: Fri, 22 Dec 2000 16:55:27 -0500
Subject: coding convention conversion
In-Reply-To: <20001222122015.A7987@llnl.gov>; from trotts@llnl.gov on Fri, Dec 22, 2000 at 12:20:15PM -0800
References: <20001222122015.A7987@llnl.gov>
Message-ID: <20001222165527.C25976@westsidecnc.rh.rit.edu>

This shouldn't be too hard with a python script and some regexes with
some more string processing in between (or maybe multiple pass).

Everything you want to change is bounded by "def" on one side and "("
on the other.  Then replace the capitals with _ followed by the lower
case version.

A quick (partial) solution:

re = re.compile( "def ([a-z]+)([A-Z])([a-z]*)\(" )

for line in source.readlines() :
	match = re.match( line )
	if ( match ) :
		res = "def " + re.group( 0 ) + "_" + \
			string.lower( re.group( 1 ) + re.group( 2 ) + "("


This isn't complete and won't have the arguments in the result (nor
preserve indentation before the "def").

I think the re.sub() function will do more of what you want it to.


HTH,
-D


On Fri, Dec 22, 2000 at 12:20:15PM -0800, Issac Trotts wrote:
> Does anyone know of a utility to change between the coding conventions 
> of
> 
> def fooBarBaz(): pass
> 
> and 
> 
> def foo_bar_baz(): pass
> 
> It is a pain to do this by hand.
> 
> Thanks,
> Issac



From fig at monitor.net  Sat Dec  9 19:03:25 2000
From: fig at monitor.net (Stephen R. Figgins)
Date: Sat, 09 Dec 2000 16:03:25 -0800
Subject: corba, soap, com, etc...
References: <pKsY5.58762$yR4.1496678@news1.rdc1.tx.home.com> <90tph1$o21$1@nnrp1.deja.com> <acwY5.58954$yR4.1507182@news1.rdc1.tx.home.com>
Message-ID: <3A32C84D.96BB5B9A@monitor.net>

For some more on xml-rpc and Python you could read Dave Warner's
article on the O'Reilly Network. 

  http://www.oreillynet.com/pub/a/python/2000/11/22/xmlrpcclient.html

In a followup to this article Warner will give information on setting
up a simple xml-rpc server with Python.  We should publish that
article in a couple of weeks. 

-Stephen R. Figgins


From aleaxit at yahoo.com  Wed Dec  6 06:03:50 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 6 Dec 2000 12:03:50 +0100
Subject: "inherited" keyword in Python?
References: <3A2B9EFA.3A64C13B@schlund.de> <90gevs02uu1@news2.newsguy.com> <m2elznipbl.fsf@mycroft.actrix.gen.nz> <90ievt025bl@news2.newsguy.com> <m266kzhua6.fsf@mycroft.actrix.gen.nz> <90jlrk012u7@news1.newsguy.com> <m23dg2ibes.fsf@mycroft.actrix.gen.nz>
Message-ID: <90l6ea026g1@news2.newsguy.com>

"Paul Foley" <see at below> wrote in message
news:m23dg2ibes.fsf at mycroft.actrix.gen.nz...
> On Tue, 5 Dec 2000 22:11:08 +0100, Alex Martelli wrote:
>
> > To clarify: Python's rule (as applied by, e.g., getattr) is anything
> > but broken, but it doesn't easily let you return (for a given class X
>
> Python's rule works for Python, sure -- but Python doesn't provide any
> way to call a superclass method without knowing exactly what you want
> to call (else all this would be unnecessary!)

It's not a built-in feature, but we have just proven it can be
built up (the main thing missing being a way to find out 'in
what class am I right now' -- that one needs to be made explicit).


> > that inherits from A, B, C, ...) a single class Z that will ensure that
> > getattr(Z,whatever) is the same as getattr(X,whatever) for any
> > whatever that is not a key in X.__dict__.
>
> I'm not sure that's all that's needed, though.

What else would be needed?


> >> that won't work.  I thought super() was just supposed to call the
> >> appropriate same-named method of a parent class.  Anything else can be
> >> done through the usual self.whatever, surely?
>
> > Not really.  To translate literally a Java class which has methods X
> > (which calls the base's Y) and Y (which calls the base's X method),
> > I'd need a Python's super() to be like Java's super (returning a class
>
> I've never bothered with Java.  How can you write a method X which
> calls the base's Y without explicitly calling the base's Y?  And if
> you have to do it explicitly, that'll work in Python, too.

    super.Y(anarg, another)

will call the Y method of the (single) superclass, just as surely as

    self.Y(anarg, another)

will call it on the (current) class.  Both of these calls work in Java
(and the different-syntax but same-semantic equivalents in Smalltalk,
I think) from ANY method X of a certain object 'self'.

This is easy in a single-inheritance language such as Smalltalk or
Java (or C#, or VB7, ...), of course.


> > object reference) and not like what I name 'supercall'.  self.X and
> > self.Y just won't do for this.
>
> It seems to me, what you said above about Java notwithstanding, that
> the only time it's OK to call X where X is a method in a base class,
> and where self.__class__ also has a method named X (so that self.X()
> will do the wrong thing), is from inside self.X.

No way!  The set of implementations of a given class's methods
makes up ONE tightly-woven net of mutual dependencies.  It's
perfectly all right _within such a set_ for a method X of class
K to know everything about how another method Y of K is implemented,
and depend in its implementation on K's Y's implementation details,
for example to bypass an override that K's Y implementation
performs on some inherited-Y.

For example, consider two methods, a somewhat typical getter/setter
pair -- in C++ notation:

    virtual AnObj& getObj() throws(MissingObj);
    virtual AnObj& setObj(AnObj&) throws(ObjSettingFailed);

Specifically: in the baseclass B, getObj() is defined to return
(a reference to) the AnObj that was last set, raising an exception
if no setting was ever performed; setObj() is defined to set a
new AnObj (raising an exception if that somehow fails) and return
a reference to the same to allow chaining, typical use:

    myContainer.setObj(LocateObj()).tweakDetails(anoption, another);

meant to have the same semantics as:

    myContainer.setObj(LocateObj());
    myContainer.getObj().tweakDetails(anoption, another);

but with tighter expression.  All right so far?


Now, in derived class K, we need to tweak the semantics a little
bit: if NO setting was ever performed, we want getObj, before it
fails, to make a last-ditch attempt to set a suitable AnObj by
other means; similar means to be used in setObj if the setting
as requested failed.  The natural way to express this, switching
to Python, would be something like:

    def setObj(self, anObj):
        try: return B.setObj(self, anObj)
        except ObjSettingFailed:
            return B.setObj(self, self.LastDitchAttempt())

    def getObj(self):
        try: return B.getObj(self)
        except MissingObj:
            try: return B.setObj(self, self.LastDitchAttempt())
            except ObjSettingFailed:
                raise MissingObj

If in this getObj we *HAD* to use the setObj *from this
very class*, rather than the superclass's, we would get
subtly wrong semantics: *TWO* retries (two calls to
LastDitchAttempt) if the first setting fails.  We need
to call the *superclass*'s setObj from within the
*derived class*'s getObj, to get maximum simplicity.

If each method was only allowed to call-on-superclass
*the same method as itself*, we'd have to complicate
things with TWO levels of inheritance -- one to override
just getObj, then another one to override just setObj --
and in slightly harder cases, where the need to call-
superclass-implementation-of-other-method is mutual,
we'd be slightly out of luck.

Actually, this first sketch at a solution could be
usefully refactored (IMHO) to:

    def __lastDitch(self, ExceptionToTranslate=None):
        try: return B.setObj(self, self.LastDitchAttempt())
        except ExceptionToTranslate: raise MissingObj

    def setObj(self, anObj):
        try: return B.setObj(self, anObj)
        except ObjSettingFailed:
            return self.__lastDitch()

    def getObj(self):
        try: return B.getObj(self)
        except MissingObj:
            return self.__lastDitch(ObjSettingFailed)

but, even here, it's crucial for __lastDitch to be able
to call the superclass's implementation of setObj; indeed,
here we might well get a _loop_ if method __lastDitch was
only allowed to call self.setObj!

Allowing such useful elementary refactorings is another
important motivation for having ANY method of an object
call the superclass's implementation of that method, by
the way.  Suppose that 'by nature' your only calls to
super.Foo where from inside self.Foo; still, some of
those calls can occur within a repeated fragment of code,
and of course you might well want to 'factor them out'
into a little method of their own.  I can hardly stand
a sequence of 3-4 statements repeated 3 or 4 times when
it clearly should be a function/method... having copies
of it strewn around can well make maintenance needlessly
more difficult (when you miss one of the repetitions and
thus fail to tweak it -- if you only have one 'repetition',
called from wherever it's needed, no risk of that then:-).


> >> > But, how often is it going to be satisfactory to call 'just
> >> > the one' (of several) inherited-method of a certain name, in
> >> > a multiple-inheritance context?  Just as often we may want
> >>
> >> It isn't (at least, not to me), which is why I avoided doing it that
>
> > But you did do that -- you call just one of the several 'inherited'
> > implementations of the method, in the code below.
>
> Any one call to super() only calls one method, naturally.  But if
> everyone calls super() _all_ the methods eventually get called.  Under
> "Python rules", only the methods in D, B and A can ever run -- the one
> in C never does, unless called explicitly with C.foo(self) from D; and
> in that case, it's called out of order -- you either get D, B, A, C or
> D, C, B, A, or, more likely D, B, A, C, A or D, C, A, B, A, with A.foo
> getting called twice, since both B and C inherit A.  This sucks.  I
> get D, B, C, A, which is the right order (B.foo and C.foo are both
> called before their common superclass's A.foo, B.foo is called before
> C.foo since B is to the left of C in D.__bases__) and A.foo is only
> called once.

Sorry, I really don't get this.  B and C don't inherit from each
other either way, so why should a call to super(B) inside class B
ever invoke anything within C, or vice versa?  Remember we said
we'd provide the classobject as an explicit argument to super --
method foo in class B will thus say 'super(B)'.  How do you get
from this to C...?

D-B-A-C is not "out of order" -- it is THE order (depth-first,
left-first) that attributes are looked up in Python.  Yeah, I've
gathered that you dislike it, but, personally, I find it easy
to explain and understand, easily predictable, and simple.


> >> way.  Having figured out how to get the frame info, here's an
> >> implementation.  No need to pass in the object and method name:
>
> > I'd still want to pass them, at least optionally, to handle the
> > above case where I want to call a _different_ inherited method.
>
> I'm working on the assumption that if you want to do that, you're
> doing something wrong.

I've provided one example where this is not so.  I can provide
a zillion more, for example through the refactoring idea.  I
believe your assumption is just plain wrong.


> >> If you just call super(__thisclass) with no args other than the class,
> >> it passes on the args the calling method got; or you can specify other
> >> args explicitly [yes, I know this prevents you calling with no args if
> >> super's caller has args]
>
> > As, once again, convenience gets in the way of simplicity and
>
> Maybe.  This is an attempt to copy the CALL-NEXT-METHOD functionality
> of CLOS; in CLOS, all the methods will have similar argument lists,
> because they're all part of one generic function, not independent
> members of various classes.  Just chop out the 5 lines in the middle
> that copy the caller's arglist if you think it's better not to do
> this.

CLOS (or Dylan) generic-function/specific-methods setting has lots
of interest, but trying to shoehorn Python into that mold is, IMHO,
an unfruitful attempt.


> > generality.  Oh well:-).  "Tantum convenientia potuit suadere
> > malorum", as Lucretius might write if he lived today!-)
>
> [Am I supposed to be one of them?  And shouldn't "malorum" be dative?]

I'm definitely not accusing you to be 'an evil'!-)  I'm just
jokingly paraphrasing well-known verse 101 of Lucretius' "De Rerum
Naturae", cfr e.g.:
http://www.informalmusic.com/DRN/
(and, yes, I assure you it's correct in the genitive case here:-).
[I had no particular reason to pick Lucretius, but, on reflection,
I think one _could_ make a case for simplicity of programming in
the Epicurean mold.  "Catastematic Python", here we come...!-).


Alex





From henrik.martensson at frontconsulting.se  Fri Dec 22 06:10:11 2000
From: henrik.martensson at frontconsulting.se (Henrik Mårtensson)
Date: Fri, 22 Dec 2000 11:10:11 GMT
Subject: Q: Basic MAPI 
Message-ID: <nEG06.1008$Qb7.95625@newsb.telia.net>

Hi

I'm trying to run a basic MAPI initialisation, but I get stuck with the
dispatcher.

Below I have pasted the code I'm running. Is there something that you have
to do
to make the MAPI interface available to you? I'm not running this in a MS
Exchange
office environment.

Henrik

>>> from win32com.client import Dispatch
>>> from win32com.mapi import mapi
>>> s=Dispatch("Mapi.Session")
Traceback (innermost last):
  File "<interactive input>", line 1, in ?
  File "D:\Program Files\Python\win32com\client\__init__.py", line 13, in
Dispatch
    dispatch, userName =
dynamic._GetGoodDispatchAndUserName(dispatch,userName,clsctx)
  File "D:\Program Files\Python\win32com\client\dynamic.py", line 67, in
_GetGoodDispatchAndUserName
    return (_GetGoodDispatch(IDispatch, clsctx), userName)
  File "D:\Program Files\Python\win32com\client\dynamic.py", line 58, in
_GetGoodDispatch
    IDispatch = pythoncom.CoCreateInstance(IDispatch, None, clsctx,
pythoncom.IID_IDispatch)
com_error: (-2147221005, 'Invalid class string', None, None)
>>>




From SBrunning at trisystems.co.uk  Wed Dec  6 04:37:42 2000
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Wed, 6 Dec 2000 09:37:42 -0000 
Subject: Sort
Message-ID: <31575A892FF6D1118F5800600846864D5B1551@intrepid>

> From:	Alfred [SMTP:scj at mcom.mcom.fr]
> I'm trying to sort a List of List, on the third col, but I don't find how
> to
> do this.
 
See the sorting mini-HOWTO, at
<http://www.python.org/doc/howto/sorting/sorting.html>.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From tim.one at home.com  Sat Dec 23 00:15:25 2000
From: tim.one at home.com (Tim Peters)
Date: Sat, 23 Dec 2000 00:15:25 -0500
Subject: <> vs != was Re: A Suggestion for Python Colon Syntax
In-Reply-To: <2kM06.120$Lm2.7518@news.clarkson.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEDEIFAA.tim.one@home.com>

[Brad Clements]
> Horrors, I just checked some of my code and find that I use both !=
> and <> in the same module!

Dear Lord, man!  We can only pray that your msg expires from Guido's
newsfeed before he returns.  If not ... well, in the spirit of the season
you should be sure to tell your family how much you loved them.

> I don't know why.. Python is so easy to write for, I just use whatever
> is easier to type, so if my hands are low on the keyboard, out
> comes <> .. hands higher up then !=
>
> I think I'll stick with <>, but then I'll have to *think* about
> the syntax, rather than the algorithm.. something I normally don't
> have to do in Python..

Best guess is that "<>" will go away in P3K.  I jokingly told Guido it was
clearly wrong for an inequality operator to look so symmetric, and he forgot
to laugh <wink>.

> What a lovely language.

matched-only-by-your-lovely-post-ly y'rs  - tim




From arhiv at freemail.org.mk  Fri Dec  8 13:30:47 2000
From: arhiv at freemail.org.mk (Damjan)
Date: Fri, 8 Dec 2000 19:30:47 +0100
Subject: A patch for gdbmmodule
Message-ID: <20001208193047.A338@localhost.localdomain>

Hi everybody. According the info page for gdbm:
...
The option GDBM_FAST is now
obsolete, since `gdbm' defaults to no-sync mode.
...
So I made some modifications to the gdbmmodule to reflect this, and to allow
opening a database unlocked. The new options to gdbm.open(...) are 's' to open
the db in SYNC mode and 'u' to open the db unlocked. There is also a need to
define lock() and unlock methods to the gdbm object, but I can't find a way to
unlock the gdbm file.

Here's the patch (against a Python2.0 gdbmmodule.c):

*** gdbmmodule.c.old	Mon Dec  4 13:34:59 2000
--- gdbmmodule.c	Mon Dec  4 13:34:15 2000
***************
*** 454,461 ****
                          "Flags should be one of 'r', 'w', 'c' or 'n'");
          return NULL;
      }
!     if (flags[1] == 'f')
!         iflags |= GDBM_FAST;
      return newdbmobject(name, iflags, mode);
  }
  
--- 454,463 ----
                          "Flags should be one of 'r', 'w', 'c' or 'n'");
          return NULL;
      }
! 	if (flags[1] != '\0') {
! 		if (flags[1] == 's' || flags[2] == 's') iflags |= GDBM_SYNC;
! 		if (flags[1] == 'u' || flags[2] == 'u') iflags |= GDBM_NOLOCK;
! 	}
      return newdbmobject(name, iflags, mode);
  }
  


Now, what do you think and whom should I send this patch?
(and please Cc: the reply to me also)

-- 
Damjan Georgievski		|           ?????? ???????????
Skopje, Macedonia		|           ??????, ??????????



From hamish_lawson at yahoo.co.uk  Thu Dec 28 08:22:54 2000
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: Thu, 28 Dec 2000 13:22:54 GMT
Subject: newbie : array initialization, anyone ?
References: <3a4a1352.9818548@news.t-online.de> <92dlvu$tm1$1@nnrp1.deja.com>
Message-ID: <92ferd$7ke$1@nnrp1.deja.com>

Oops, careless editing in my last post resulted in this:

    multidim = [([0] * 10)[:] for i in range(3)]

The [:] is of course unnecessary there, as [0] * 10 is evaluated each
time:

    multidim = [[0] * 10 for i in range(3)]

Hamish Lawson


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


From yelled at yahoo.com  Tue Dec 12 16:09:13 2000
From: yelled at yahoo.com (yelled at yahoo.com)
Date: Tue, 12 Dec 2000 21:09:13 GMT
Subject: _oleobj_ and dir
References: <910lpp$p2g$1@nnrp1.deja.com> <910v3j01os2@news2.newsguy.com>
Message-ID: <91645k$38l$1@nnrp1.deja.com>

Thanks much Alex. Exactely what I was looking for!

Hoon,
In article <910v3j01os2 at news2.newsguy.com>,
  "Alex Martelli" <aleaxit at yahoo.com> wrote:
> <yelled at yahoo.com> wrote in message
news:910lpp$p2g$1 at nnrp1.deja.com...
> > Hi,
> >
> >  I am trying to get available functions and attributes of COM obj:
>
> Two steps: first, you must ensure that the Python wrappers for
> the typelibrary you're using are on -- the simplest way may be
> to use:
> adoConn = win32com.client.gencache.EnsureDispatch('ADODb.connection')
> rather than:
> adoConn = win32com.client.Dispatch('ADODb.connection')
>
> Once that is done, dir(adoConn.__class__) will give you, among
> other things, the COM methods of the class:
>
> >>> print dir(adoConn.__class__)

snip..
> Alex
>
>


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


From kragen at dnaco.net  Sun Dec 31 00:02:52 2000
From: kragen at dnaco.net (Kragen Sitaker)
Date: Sun, 31 Dec 2000 05:02:52 GMT
Subject: HTML Parser
References: <mailman.978224102.28797.python-list@python.org> <0gw36.45$9s1.9126@e420r-atl2.usenetserver.com> <qnkito1p9jj.fsf@arbutus.physics.mcmaster.ca>
Message-ID: <06z36.123$9s1.19200@e420r-atl2.usenetserver.com>

In article <qnkito1p9jj.fsf at arbutus.physics.mcmaster.ca>,
David M. Cooke <cookedm at physics.mcmaster.ca> wrote:
>At some point, kragen at dnaco.net (Kragen Sitaker) wrote:
>> In a string like "x<a>b<c>d", this will match "<a>b<c>", because the .*
>> matches "a>b<c".  This explains your problem.
>> 
>> Fixing it is harder.
>
>Not that hard: use the pattern '<.*?>'.

Well, he wants to upcase his tag names; this will still match the
entire attribute name and all attribute values, so his URLs will get
upcased.  This is bad, and fixing it *is* harder.

-- 
<kragen at pobox.com>       Kragen Sitaker     <http://www.pobox.com/~kragen/>
Perilous to all of us are the devices of an art deeper than we possess
ourselves.
       -- Gandalf the White [J.R.R. Tolkien, "The Two Towers", Bk 3, Ch. XI]




From nickatvideosystemdotcodotuk  Mon Dec 18 06:22:51 2000
From: nickatvideosystemdotcodotuk (Nick Trout)
Date: Mon, 18 Dec 2000 11:22:51 -0000
Subject: what is a closure?
References: <3a39033b@news.xtml.co.uk> <S18_5.3929$Kd1.455131@newsb.telia.net> <8nck3tocsg8kp1u2g1du6nd3lt9gfeckr2@4ax.com> <91dmf7$efb$1@news.udel.edu>
Message-ID: <3a3df3bb@news.xtml.co.uk>

:-D

So "what is a closure" and "does Python have first class functions" are 2
difficult and controversial questions to ask! :-)



"Terry Reedy" <tjreedy at udel.edu> wrote in message
news:91dmf7$efb$1 at news.udel.edu...
>
> "Steve Horne" <sh at ttsoftware.co.uk> wrote in message
> news:8nck3tocsg8kp1u2g1du6nd3lt9gfeckr2 at 4ax.com...
> > I was reading a book about developing compilers recently (Modern
> > Compiler Design, by Dick Grune, Henri Bal, Ceriel Jacobs and Koen
> > Langendoen), and something that was presented in it was a 'closure
> > algorithm'. This took a collection of data and some related
> > expressions, and repeatedly evaluated the expressions using both
> > original and previously derived data until all the data it was
> > possible to derive was fully derived.
> >
> > The thing that made it a closure algorithm seemed to be the fact that
> > the expressions were applied repeatedly on an accumulating set of data
> > until no further data could be derived - the closure being the
> > complete collection of data.
>
> This sounds like 'transitive closure', often denoted by '*', which is
> different from function closure.
>
>
>




From jan.dries at dcube-resource.be  Sun Dec 24 07:05:20 2000
From: jan.dries at dcube-resource.be (Jan Dries)
Date: Sun, 24 Dec 2000 13:05:20 +0100
Subject: Same again please for OOP
References: <200012231813.eBNIDoh20571@ZOOstation.cc>
Message-ID: <3A45E680.7241D2CF@mail.com>


Rob Brown-Bayliss wrote:
> 
> Like wise,  I am trying to get my head around OOP, but all I can find
> are basic howto make a class in C++
> 
> I understand how to make classes in both Python and C++.  What I want is
> help in deciding waht would be a good class and what would not,
> preferably in a pythjon contxt, but a generic context would be wonderful
> as well.
> 

To some extent, only experience can really help you in deciding what
would be a good class and what not. And as they say: true judgement
comes from experience; experience comes from poor judgement. There's no
cookbook for OOP, no 10 golden rules that you can blindly follow.
Still, there's a lot of material out there with the experience of other
people. And if I would have to recommend just one, I'd suggest you try
"Object Oriented Software Construction" by Bertrand Meyer, Prentice
Hall, ISBN 0-13-629155-4. It doesn't once mention Python, but when it
comes to OO, it is, in my oppinion, one of the better books out there.

Regards,
Jan



From cobrien at Radix.Net  Wed Dec 27 14:40:46 2000
From: cobrien at Radix.Net (Cary O'Brien)
Date: 27 Dec 2000 14:40:46 -0500
Subject: sniffing btw my web browser and a server
References: <92aush$rhc$1@news.nuri.net>
Message-ID: <92dgju$60f$1@saltmine.radix.net>

In article <92aush$rhc$1 at news.nuri.net>,
June Kim <junaftnoon at nospamplzyahoo.com> wrote:
>I'm doing some sort of web client programming, and have to observe the
>messages that my browser and web-servers transmit, being tired of guessing
>what the messages'd be and looking up the references and the specifications.
>
>Anyone has done it in Python or knows how it could be done?
>
>
>Best regards,
>
>June
>
>ps. using TCP/IP packet sniffers made for mere hacking seems too primitive
>or low-level in this case.
>

Take a look at "ethereal".  It is a (GTK?) graphical packet sniffer that
unpacks many protocols.  You can use it to snapshot an HTTP session and use
the "TCP Session" option to see exactly what is going on.

   http://www.ethereal.com

Slick.  Hours of fun on a busy network.

-- cary




From quinn at euro.ugcs.caltech.edu  Sun Dec 17 06:57:13 2000
From: quinn at euro.ugcs.caltech.edu (Quinn Dunkan)
Date: 17 Dec 2000 11:57:13 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <r0pzoi11lhz.fsf@black.cs.washington.edu> <5jvZ5.3588$Kd1.418770@newsb.telia.net> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com> <kuu28971kp.fsf@lasipalatsi.fi> <3A380E4E.8537A0FB@san.rr.com> <91a1vr02b02@news2.newsguy.com> <Co9_5.25465$x6.13756579@news2.rdc2.tx.home.com> <91bbq102mtk@news1.newsguy.com> <rWd_5.26820$x6.14119694@news2.rdc2.tx.home.com> <91d5el0me7@news1.newsguy.com> <rOr_5.31120$x6.15241856@news2.rdc2.tx.home.com> <slrn93o7j4.tu0.quinn@dinar.ugcs.caltech.edu>
Message-ID: <slrn93pago.8tf.quinn@euro.ugcs.caltech.edu>

On 17 Dec 2000 02:01:08 GMT, Quinn Dunkan <quinn at dinar.ugcs.caltech.edu> wrote:
[ 6 times ]

Argh, sorry about the multiple posts.

Note to self: when newsreader says "Post failed, try again? (y/n) " what it
*really* means is "Post succeeded, make a fool of yourself on Usenet? (y/n) ".


From claird at starbase.neosoft.com  Tue Dec  5 11:21:55 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 5 Dec 2000 10:21:55 -0600
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Dec  4)
References: <8C419A8040EAFDCB.1774103991763D9E.2F034C54E78FF82C@lp.airnews.net> <90iqs8$n6k$1@newsreaderm1.core.theplanet.net>
Message-ID: <FF300EF48DEFEA95.7BB307DE8FFF7426.C113AE2086024325@lp.airnews.net>

In article <90iqs8$n6k$1 at newsreaderm1.core.theplanet.net>,
Franz GEIGER <fgeiger at datec.at> wrote:
>>     Alex Martelli shows a clever way to overload a method based on
>>     number of arguments.
>>         http://deja.com/=dnc/getdoc.xp?AN=975947338
>
>This link leads to "Invalid document request". How is the right one?
			.
			.
			.
I don't know--but 
<URL:http://deja.com/=dnc/getdoc.xp?AN=697365234>
and
<URL:http://deja.com/=dnc/getdoc.xp?AN=664267725>
certainly make for good reading.
-- 

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


From ngps at madcap.dyndns.org  Mon Dec 25 10:44:16 2000
From: ngps at madcap.dyndns.org (Ng Pheng Siong)
Date: 25 Dec 2000 15:44:16 GMT
Subject: registry tweaking program
References: <9250ek$ueo$1@nnrp1.deja.com>
Message-ID: <927q0g$j55$1@coco.singnet.com.sg>

According to  <kevrjohn at my-deja.com>:
>    I would like to write a program that will automatically modify
> certain parts of the registry in win98.

A better (heh!) program will be to diff the registry images before
and after your or other programs' (e.g., installers) modifications.

Merry Christmas!

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



From aahz at panix.com  Mon Dec 11 09:20:21 2000
From: aahz at panix.com (Aahz Maruch)
Date: 11 Dec 2000 06:20:21 -0800
Subject: Newbie Thread question
References: <3A3453BD.B9A54981@ihug.co.nz>
Message-ID: <912nr5$6h2$1@panix3.panix.com>

In article <3A3453BD.B9A54981 at ihug.co.nz>,
Matthew  <matthewm at ihug.co.nz> wrote:
>
>I've got thread stuff working by sub-classing threading.Thread, but
>haven't be able to figure out how to use them by calling a runnable
>target. Something like:
>
>dog1 = threading.Thread(None, Dog().bark())

I'm suspecting that you need a little more practice with the Python
class system.  The way one would normally do this (assuming that Dog is
the name of your subclass) is like this:

class Dog(threading.Thread):
    def run(self):
        self.bark()

dog1 = Dog()
dog1.start()

>I checked the faq and a few other places and there seems to little to no
>docs to get one started - any urls i've missed.

Try http://starship.python.net/crew/aahz/
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

What if there were no rhetorical questions?  --Aahz


From grisha at ispol.com  Sun Dec 31 01:31:13 2000
From: grisha at ispol.com (Gregory Trubetskoy)
Date: Sun, 31 Dec 2000 06:31:13 GMT
Subject: tftplib ?
Message-ID: <Pine.LNX.4.21.0012310129140.9173-100000@saturn.ispol.com>

Before I write my own - 

is there by a chance a tftplib.py or something that implements tftp in
Python out there?

Thanks!

--
  Gregory (Grisha) Trubetskoy
       grisha at ispol.com



From fredrik at effbot.org  Sun Dec 31 05:57:28 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sun, 31 Dec 2000 10:57:28 GMT
Subject: array of objects
References: <92n1s7$tq9$1@nnrp1.deja.com>
Message-ID: <siE36.2779$Qb7.318673@newsb.telia.net>

Anders Olme wrote:
> How do I create a array of object from my own classes?
>
> If I have a class like this:
>
> class test:
> def dosomething(self):
> return 1234
>
> obj= test()
>
> obj= [test(),test()]
>
> The last line works but is there a more elegant way?

Not if you need exactly two elements ;-)

To create a list containing N instances, you can use a
list comprehension:

    obj = [test() for x in range(N)]

which is short for

    obj = []
    for x in range(N):
        obj.append(test())

If the test class is "immutable" (that is, if it cannot be modified
in place), you can also use "sequence multiplication":

    obj = [test()] * N

this creates a list with N references to the same instance.

</F>




From jerry_spicklemire at my-deja.com  Thu Dec 21 10:58:07 2000
From: jerry_spicklemire at my-deja.com (jerry_spicklemire at my-deja.com)
Date: Thu, 21 Dec 2000 15:58:07 GMT
Subject: while true: !!!
References: <3A4024E5.98843CFC@cosc.canterbury.ac.nz> <91qadb$1di$1@plutonium.compulink.co.uk> <3A41539E.DF6977D6@cosc.canterbury.ac.nz>
Message-ID: <91t9aa$9sa$1@nnrp1.deja.com>

  Greg Ewing wrote:
> Hmmm, let's try that out. Is [[]] true or false?
> It's a list containing a list containing nothing.
> Since [] is false, a list containing nothing must be
> nothing. So [[]] is a list containing nothing,
> which is nothing. Ah, it must be false! :-)
>
> Nothing-is-weird-stuff,

Maybe I was just lucky, but the first time I encountered this concept,
the documentation at hand refered to "empty". That's easy for me to
grasp. If I have a sack with an empty jar inside, the jar is indeed
empty, but the sack is certainly not!

Later,
Jerry S.


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


From johann at physics.berkeley.edu  Wed Dec 13 14:58:18 2000
From: johann at physics.berkeley.edu (Johann Hibschman)
Date: 13 Dec 2000 11:58:18 -0800
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <kuzoi15d60.fsf@lasipalatsi.fi> <Pine.BSF.4.10.10012122310510.48551-100000@tcfreenet.org>
Message-ID: <mtaea0nlet.fsf@astron.berkeley.edu>

Aaron Jon Reichow writes:

> According to some benchmarks I've seen, Squeak is faster. ;P
> (although it doesn't feel like it sometimes, with Morphic)

Has the speed of Squeak improved in the past six months or so?  I
downloaded it last spring, just to play with a SmallTalk, and found it
slow, ugly, and rather clunky.  Not to mention willing to eat all the
memory on my machine.

Perhaps I wasn't doing it right, but I was not impressed.  I'm not
sure what I can do with smalltalk that I can't do with a good Common
Lisp environment, but I've just never quite "got" smalltalk.

Anyway, should I download the current Squeak and try again?

-- 
Johann Hibschman                           johann at physics.berkeley.edu


From tseaver at starbase.neosoft.com  Sun Dec 10 18:40:44 2000
From: tseaver at starbase.neosoft.com (Tres Seaver)
Date: 10 Dec 2000 17:40:44 -0600
Subject: Programming Habits in Python
References: <cspU5.82459$U46.2617014@news1.sttls1.wa.home.com> <slrn92tv0u.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com> <udn1e8k0qi.fsf@box.home.de> <slrn937caf.h4.amk@207-172-146-162.s162.tnt3.ann.va.dialup.rcn.com>
Message-ID: <682DDF8D24ABBCEB.7AA1428334F7B3A8.A771CF7758416C27@lp.airnews.net>

In article <slrn937caf.h4.amk at 207-172-146-162.s162.tnt3.ann.va.dialup.rcn.com>,
A.M. Kuchling <akuchlin at mems-exchange.org> wrote:

>In the back of my mind I had a quotation, which I cannot track down on
>the Web.  The Principia uses geometrical methods throughout, even
>where analytical methods would have been much simpler, and someone
>(Halley, perhaps) said that reading the Principia was like walking
>through an arsenal of giant weapons of a past age, and that one is
>amazed that anyone could wield them at all.

I don't think it could be Halley;  he was actively involved (as the
then secretary of the Royal Society) in getting the Principia published.
It sounds a bit like an analogy my memory wants to attribute to
Bertrand Russell:  "the amazing thing about a waltzing bear is not
how well he waltzes, but that he waltzes at all."
-- 
===============================================================
Tres Seaver                                tseaver at digicool.com
Digital Creations     "Zope Dealers"       http://www.zope.org


From sandj.williams at gte.net  Wed Dec 13 23:11:42 2000
From: sandj.williams at gte.net (Steve Williams)
Date: Thu, 14 Dec 2000 04:11:42 GMT
Subject: Python for REAL Projects ?
References: <3a380d1e.409238@bendigo.news.origin.net.au>
Message-ID: <3A384992.2A2A861F@gte.net>

dragon at hitech.net.au wrote:

> Hello Python People !
> I have been looking at, and trying out Python (under Win NT) and
> although Python SEEMS interesting can it "cut the mustad" on a real
> client project ??
>
> Curently, I develop using DELPHI + Interbase SQL ( or Borland BDE
> tables) because , in the main, DELPHI offers "mostly" easy to setup
> forms and access to database tables ( not perfect but usable).
>

[snip]

You won't find anything as nice as the Delphi IDE in the Python world.  Or
anywhere else, for that matter.  It is the gold standard, IMHO.

But, as you probably know, real Delphi programmers bypass the IDE and
write the code directly in the editor.  That's one of the lovely things
about Delphi--the IDE/editor duality.

Instead of Tkinter, look at wxPython.  Some of the demos are quite amazing
and the code looks a lot like Delphi code.

You don't have to use mySQL.  Interbase is a [categorically] superior
product.  Having said that, I must confess I've never tried to access
Interbase via mxODBC on non-Windows platforms.




From robin at alldunn.com  Fri Dec 15 11:57:55 2000
From: robin at alldunn.com (Robin Dunn)
Date: Fri, 15 Dec 2000 08:57:55 -0800
Subject: Compiling Python 2.0 with Berkeley DB 3.0.55?
References: <3A39DF89.D76B6415@schlund.de>
Message-ID: <u4s_5.168$SKV4.36635003@news.randori.com>

"Carsten Gaebler" <cg at schlund.de> wrote in message
news:3A39DF89.D76B6415 at schlund.de...
> Hi there!
>
> In our company we have a Linux server with the Berkeley DB 3.0.55
> installed and some DB-3.0.55 files. I compiled Python 2.0 out of the box
> on that machine and the dbhash module worked fine for the DB files. But
> on my own workstation I get the error message "Invalid argument" from
> dbhash.open() when I try to open a DB-3.0.55 file. I compiled Berkeley
> DB 3.0.55 and replaced the db*.h files in /usr/include/ and libdb.a in
> /usr/lib/ but it just won't work. Any hints?
>

Is this is on an existing file created on the other system?  Can you create
and use new files with dbhash.open()?  If so the file may have become
corrupted or something.  Try running db_verify (one of the utilities built
when you compiled DB) on the file and see what it tells you.

P.S. In case you are interested, the next generation of Berkeley DB wrappers
is available at http://pybsddb.sourceforge.net/


--
Robin Dunn
Software Craftsman
robin at AllDunn.com
http://wxPython.org     Java give you jitters?
http://wxPROs.com        Relax with wxPython!





From aleaxit at yahoo.com  Mon Dec 18 15:16:19 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 18 Dec 2000 21:16:19 +0100
Subject: while true: !!!
References: <B43D149A9AB2D411971300B0D03D7E8B0FBCBB@natasha.auslabs.avaya.com> <v04220801b663769e1800@[192.168.11.33]> <mailman.977147642.18708.python-list@python.org> <91lclt02qr3@news1.newsguy.com> <3A3E4C55.9C785306@san.rr.com>
Message-ID: <91lrgs0236n@news2.newsguy.com>

"Darren New" <dnew at san.rr.com> wrote in message
news:3A3E4C55.9C785306 at san.rr.com...
    [snip]
> > Few languages equate 'success' (not raising an exception, for
> > languages where 'exception' means something) with 'a boolean
> > result' (Icon is rather oriented that way, but that's because
> > it _backtracks_ over 'insuccess' -- a very different setup
> > than Python's).
>
> Actually, it's not *quite* like that. "if" in Icon is true for *any*
result.
> It's only the "no result" that invokes the else part. A failure/exception
in
> Icon is more like partial functions in math, where it's possible there are
> inputs which generate no results at all.  One of the cleaner ways of
> handling it I've seen.

If we want to get into Icon details...: the general case is a
*sequence* of results, if/when the expression gets resumed
(the backtracking I mentioned -- but also explicitly, in
'every') to generate more.  'no _more_ results' (which holds
from the start for an empty result-sequence) is if's 'false'
meaning (also while's, of course).

It's cool semantics, but if you've ever tried to explain to a
programming beginner the difference between every and
while, you might be slightly less enthusiastic about it:-).


Alex





From mwh21 at cam.ac.uk  Tue Dec 12 18:53:22 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 12 Dec 2000 23:53:22 +0000
Subject: Bug: Readline getting stuck on Linux and Solaris
References: <lxk8zqwupy2.fsf@ioasun3.epfl.ch> <90m53k$ovh$1@nnrp1.deja.com> <m3u28h48ja.fsf@atrus.jesus.cam.ac.uk> <916bum$a9k$1@nnrp1.deja.com>
Message-ID: <m3d7ex2o3x.fsf@atrus.jesus.cam.ac.uk>

sragsdale at my-deja.com writes:

> In article <m3u28h48ja.fsf at atrus.jesus.cam.ac.uk>,
>   Michael Hudson <mwh21 at cam.ac.uk> wrote:
> > Hmm.  This *used* to happen on Linux (with slightly different
> > symptoms; control-C would kill the interpreter, not hang it), but the
> > bug got fixed.  A couple of things to try: does control-C do what it
> > should when you're not sitting at the ">>> " prompt?  E.g. if you type
> >
> > >>> while 1:
> > ...    pass
> > ...
> >
> > and whack control-C, what happens?
> 
> control-C behaves normally in that case.  Python2 is able to exit from a
> while loop via control-C without problems.  However if you hit control-C
> *again* (after you've exited to the '>>>' prompt) you're back in the
> locked-up mode.

Oh goody.  Can you kill -SEGV or something the python process and post
(or send me, if it's really huge) the backtrace?  Also, it might be an
idea to do look at the stack trace a few times to check that we don't
have a memory scribbling type bug.

If you're really feeling keen, you could send me the config.h your
readline made when you configured it...

> Haven't tried building without threads, but needless to say this
> shouldn't happen even with threads.

Well, obviously not, but when mystery-bug hunting all information is
good.  Anyway, from what Quinn said, I bet your problem would go away
without threads.

Fun fun fun.

Cheers,
M.

-- 
  I have a cat, so I know that when she digs her very sharp claws into
  my chest or stomach it's really a sign of affection, but I don't see
  any reason for programming languages to show affection with pain.
                                        -- Erik Naggum, comp.lang.lisp


From sh at ttsoftware.co.uk  Mon Dec 18 05:40:04 2000
From: sh at ttsoftware.co.uk (Steve Horne)
Date: Mon, 18 Dec 2000 10:40:04 +0000
Subject: what is a closure?
References: <3a39033b@news.xtml.co.uk> <S18_5.3929$Kd1.455131@newsb.telia.net> <8nck3tocsg8kp1u2g1du6nd3lt9gfeckr2@4ax.com> <91dkvu$7mk$1@nnrp1.deja.com>
Message-ID: <h3qr3tg0antlh0umqj86ml55k0uruids7t@4ax.com>

Thanks to everyone who replied. I now get the idea that there are at
least *three* things that can be called a closure.

-- 
Steve Horne
Home : steve at lurking.demon.co.uk
Work : sh at ttsoftware.co.uk


From claird at starbase.neosoft.com  Wed Dec 27 06:03:12 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 27 Dec 2000 05:03:12 -0600
Subject: Python vs. PHP (& Java?)
References: <LOBBJMHOIFOLFIPFANHJMEMHCFAA.beno@TheWebsons.com> <mailman.977912052.14706.python-list@python.org>
Message-ID: <17B4EBB7FF259C84.AB8AC2B072F97FFB.6CF7A7DEF4942B87@lp.airnews.net>

In article <mailman.977912052.14706.python-list at python.org>,
Thomas Wouters  <thomas at xs4all.net> wrote:
			.
		[many other true
		things]
			.
			.
>understand), and way easier to learn. Python is (supposedly :) slower than
>Java in execution, but especially in networked environments, where you spend
>most of your time waiting for the network anyway, you aren't going to notice
>that.
			.
			.
			.
Python is frequently faster in execution than Java.
I know Thomas knows this; I just don't want neophytes
to have an opportunity to misunderstand his jest.
-- 

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


From Barrett at STScI.Edu  Thu Dec  7 08:24:14 2000
From: Barrett at STScI.Edu (Paul Barrett)
Date: Thu, 07 Dec 2000 08:24:14 -0500
Subject: wither PyFITS ?
References: <3A2EB29B.EBB3CCB5@noaa.gov>
Message-ID: <3A2F8F7E.25BE3578@STScI.Edu>

j vickroy wrote:
> 
> claims there is a FITS module (PyFITS) available, but I have not been
> able to find it.
> 
> Any help would be appreciated.

You can find version 0.3.2 of PyFITS as http://ra.stsci.edu/pyfits. It
is not complete, but handles header access well.  The major impendement
to completing it is being able to efficiently hand the binary
extensions.

There is a group of us at STScI working to improve it.  The major
problem with the current module is lack of a good and efficient
interface to binary tables.  In an effort to remedy this situation, we
are also re-designing and re-implementing Numeric 2 which will be able
to handle arrays of records (C-structures) as well has homogeneous
arrays. 

Let me know if you have any further questions.

 -- Paul


From glenfant at equod.com.nospam  Mon Dec 18 10:11:48 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Mon, 18 Dec 2000 16:11:48 +0100
Subject: cron with sched module
References: <91dgfa$q54$1@reader1.imaginet.fr> <mailman.977141221.13933.python-list@python.org>
Message-ID: <91l974$s2$1@reader1.imaginet.fr>

Thanks Andreas,

It's what I've finally done. It's easy and meets what I need.
The "sched" module is definitely stupid: the "run" method is blocking and
there's no way to add new events to a running sched object.
Or did I miss something.
"Andreas Jung" <andreas at andreas-jung.com> a ?crit dans le message news:
mailman.977141221.13933.python-list at python.org...
> On Fri, Dec 15, 2000 at 05:26:33PM +0100, Gilles Lenfant wrote:
> > I'm trying hardly to make a python "cron" like utility that may run
> > functions/methods periodically (say every x seconds) and predetermined
> > hours.
> >
> > I can't use Window "AT" or Unix "cron" cause some functions/methods must
be
> > executed evry 15 seconds. Launching Python executable periodically such
a
> > manner will eat the disk I/O and
> >
> > So, I'm trying with the "sched" module to do this but I can't add
"enter"
> > to a scheduler object already running.
> > So, what's the workaround ?
> > Use directly lower level "threading" module in lace of sched ?
>
> I suggest to use the thread or threading module. The threading
> provides a higher level interface (it's build on the top
> of the thread module). The way how to implement a solution for
> your needs depends on your needs. E.g. you can create a worker
> threads that runs in a loop. To periodic execution you could
> use time.sleep() to delay execution for a period of time.
>
> Andreas
>



From rbsharp at gmx.de  Mon Dec 11 11:33:59 2000
From: rbsharp at gmx.de (Richard Sharp)
Date: Mon, 11 Dec 2000 17:33:59 +0100
Subject: Python Problem - Could not find platform independent
References: <9105de$20l$1@venus.telepac.pt>
Message-ID: <912vql$llc$03$1@news.t-online.com>

Its difficult to know without the exact Linux version. It reminds me of a
problem with Mandrake which is just a packaging of Redhat. Redhat already
uses Python for its configuration utilities. Perhaps their version of Python
is higher up in the search path than your installed version, since they put
theirs in /usr/python1.5 and Python itself installs in /usr/local/lib. I'm
not really sure since it is sometime ago,, but you could have a quick look.

Linux - Sinesoft <linux at sinesoft.com> schrieb in im Newsbeitrag:
9105de$20l$1 at venus.telepac.pt...
> hi!
>
> I had installed python in my linux box, but i had the following error:
>
> bash-2.03# python
> Could not find platform independent libraries <prefix>
> Could not find platform dependent libraries <exec_prefix>
> Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>]
> 'import exceptions' failed; use -v for traceback
> Warning!  Falling back to string-based exceptions
> 'import site' failed; use -v for traceback
> Python 1.5.2 (#1, Apr 28 2000, 06:45:56)  [GCC 2.95.2 19991024 (release)]
on
> linux2
> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>>
>
>
> Can anyone please help me corrected this error message ?  Please answer me
> by email.
>
>
>




From djc at object-craft.com.au  Fri Dec 22 00:38:38 2000
From: djc at object-craft.com.au (Dave Cole)
Date: 22 Dec 2000 16:38:38 +1100
Subject: Sybase module 0.10 released
Message-ID: <m3ofy52ext.fsf@vole.object-craft.com.au>

What is it:

The Sybase module provides a Python interface to the Sybase relational
database system. The Sybase package supports almost all of the Python
Database API, version 2.0 with extensions.

The module works with Python versions 1.5.2 and later and Sybase
versions 11.0.3 and later.  It is based on the Sybase Client Library
(ct_* API), and the Bulk-Library Client (blk_* API) interfaces.

Changes:

- The numeric object has been reimplemented using the
  cs_convert/cs_calc/cs_cmp API.  This has improved the coerce
  behaviour over the previous release.

>>> import Sybase
>>> n = Sybase.numeric(100200300400500L)
>>> n / 12.3456
8116276276608.6702954898911353032659409020
>>> n / Sybase.numeric('12.3456')
8116276276608.67029548989113530326
>>> n.precision, n.scale
(15, 0)
>>> n + 12.3456
100200300400512.345600000000
>>> n + Sybase.numeric('12.3456')
100200300400512.3456
>>> n.precision, n.scale
(15, 0)

  You can also specify a precision and scale in the constructor.

>>> m = Sybase.numeric(n, 30, 2)
>>> m
100200300400500.00
>>> m.precision, m.scale
(30, 2)

Where can you get it:

        http://www.object-craft.com.au/projects/sybase/

- Dave

-- 
http://www.object-craft.com.au


From tegbert at abl.med.utah.edu  Thu Dec  7 16:30:53 2000
From: tegbert at abl.med.utah.edu (Tim Egbert)
Date: Thu, 7 Dec 2000 14:30:53 -0700
Subject: dl module
Message-ID: <Pine.LNX.4.10.10012071428160.28174-100000@abl.med.utah.edu>

How can I get the dl module?  I've read about it in the documentation for
Python 2.0, but can't find it in the distribution or anywhere else.
Supposedly it's a standard module.  What gives?

Tim



From amk at mira.erols.com  Wed Dec 13 20:09:16 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 14 Dec 2000 01:09:16 GMT
Subject: Moving from Perl to Python
References: <3a3a9552.76673185@news.tcn.net> <slrn93fcdt.a8m.scarblac@pino.selwerd.nl>
Message-ID: <slrn93g86m.o0.amk@207-172-112-249.s249.tnt4.ann.va.dialup.rcn.com>

On 13 Dec 2000 17:30:56 GMT, Remco Gerlich <scarblac at pino.selwerd.nl> wrote:
>Personally I don't like Programming Python much. It's becoming old (there
>are rumors of a new version soon) and not very well structured. But then,

At least one online bookseller reports a new edition of _Programming
Python_ in February 2001.  Considering that's only two months away,
and the book isn't listed on O'Reilly's list of upcoming books, that
date is doubtless quite wrong.  

I think experienced programmers who need little handholding will do
best with the Beazley book: it's quite compendious.

--amk


From skip at mojam.com  Sun Dec 31 14:13:10 2000
From: skip at mojam.com (Skip Montanaro)
Date: Sun, 31 Dec 2000 13:13:10 -0600 (CST)
Subject: why no "do : until"?
In-Reply-To: <lgwFRKAp8nT6Ew$P@jessikat.fsnet.co.uk>
References: <92jj3j$ff7$1@nnrp1.deja.com>
	<05C6FUhLDNUU-pn2-tuTnzOCokxem@bill>
	<lgwFRKAp8nT6Ew$P@jessikat.fsnet.co.uk>
Message-ID: <14927.34118.92255.985761@beluga.mojam.com>

    Sonna> Its a design flaw in Python.
    Becker> in this case I have to agree with the pascalers there's no
    Becker> reason not to ...

If my experience programming C for ten years or so is any indication,
do/until or repeat/until would be used rarely enough to not warrant chewing
up two new keywords.  As others have pointed out, the while 1 construct,
while not ideal, is easily recognized as the common workaround.

    Becker> Python is actually a bit broken as continue inside try blocks is
    Becker> not allowed.

That is an admitted limitation of the implementation (because of the
complexity of generating correct byte code), not a design constraint of the
language.  This is being worked on.  In fact, it may be working by now in
the latest CVS version.

    Robin> The malevolence/stupidity of the data set used by Guido will
    Robin> forbid any such minor modifications of the language as his
    Robin> original tutees presumably chose just those constructs which, in
    Robin> his benevolence, he now allows.

You're free to invent your own language.

-- 
Skip Montanaro (skip at mojam.com)
Support the Mojam.com Affiliates Program: http://www.mojam.com/affl/
(847)971-7098



From aahz at panix.com  Wed Dec  6 22:48:48 2000
From: aahz at panix.com (Aahz Maruch)
Date: 6 Dec 2000 19:48:48 -0800
Subject: OT: _Fifth Business_ (was Re: Python Origins)
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu> <slrn92tu7t.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com>
Message-ID: <90n1b0$4jr$1@panix3.panix.com>

In article <slrn92tu7t.lk.amk at 207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com>,
A.M. Kuchling <akuchlin at mems-exchange.org> wrote:
>
>Generally I don't bother to flame do-my-homework-for-me posters and
>simply ignore them, because I don't want c.l.py to approach
>comp.lang.perl.misc's reputation for unfriendliness.  Private mail to
>me with such questions also gets unceremoniously deleted, though
>usually in private mail people are asking for essays on _Fifth
>Business_, not Python help.  (_Fifth Business_ is commonly studied in
>Canadian high schools, so I often get that sort of request.  On one
>occasion I was flabbergasted to get e-mail from a *parent* asking if I
>had an essay on FB their child could use.)

What's _Fifth Business_?  (I did a quick search on Google and Amazon and
didn't find anything related to you.)

And, yes, the irony of this post in this thread does appeal to me.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

What if there were no rhetorical questions?  --Aahz


From tim.one at home.com  Wed Dec 27 19:14:38 2000
From: tim.one at home.com (Tim Peters)
Date: Wed, 27 Dec 2000 19:14:38 -0500
Subject: A Suggestion for Python Colon Syntax
In-Reply-To: <92cu71$9m4$1@nnrp1.deja.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEKEIFAA.tim.one@home.com>

[billtj at my-deja.com]
> Tim, people opinion can change over time,

You apparently haven't met Guido yet <wink>.

> and hopefully Python does not have the principle that once a
> feature is already decided, it can never be changed.  I think
> to survive, a language has to be flexible to follow (rapidly)
> changing technology and therefore modify some of its own
> characteristics.

Bill, I spent too much time arguing about this already.  You may not realize
it, but you have too <wink>.  You should be able to tell from the
overwhelming lack of other response on the newsgroup that there's no
groundswell of desire (for making trailing colons optional) being suppressed
here!  It's not a new suggestion, and it's never been a popular one.  If you
want to put it to a vote, ya, it would lose.  Write a PEP if you like, but
I'm not putting more time into this unless and until there is a PEP (a prime
purpose of the PEP process is to stop endlessly repetitive newsgroup
debates).

Just one gloss:

> ...
> Guido does, but shouldn't we take also the opinions of the
> community?  I don't like to make this analogy, but in the
> history, there were/are communities ruled by kings, and there
> communities ruled by democracy.  Which one works better for
> the welfare of the community?

Guido is often called "the BDFL" here.  That stands for Benevolent Dictator
For Life.  Python design is in fact a dictatorship, and that's widely
perceived as one of Python's greatest strengths:  it gives the language an
internal coherence that committee designs rarely achieve (and, sorry, but
you haven't used Python long enough to judge this yet).  It also gives us
some of Guido's charming idiosyncrasies, but the trick is to view those *as*
charming <wink>.

Note that Guido isn't telling you where to live or how many children you can
have or what ratio of genders they must be.  He's just saying "here's my
language -- feel free to use it!".  Most of us are delighted to.

while-the-rest-plot-against-his-life<wink>-ly y'rs  - tim




From aleaxit at yahoo.com  Fri Dec 22 05:52:10 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 22 Dec 2000 11:52:10 +0100
Subject: mail processing module?
References: <91rcig$rgu$1@nnrp1.deja.com> <m266kd45m0.fsf@freak.kaiserty.com>
Message-ID: <91vbpd027a0@news2.newsguy.com>

"Michael Haggerty" <mhagger at alum.mit.edu> wrote in message
news:m266kd45m0.fsf at freak.kaiserty.com...
    [snip]
> In fact maybe a better breakdown would be one module that represents
> an email message and allows data to be MIMEd into and out of it, one
> module for sending such messages via SMTP, one module for retrieving
> such messages via POP, and one for retrieving via IMAP.  The

An *excellent* architecture.

> interfaces for the two retrieving modules should be as close as
> possible, and ideally the interface to the mail-sending routine should
> be generic enough that one could write a module with the same
> interface but that sends emails by writing them to, e.g., sendmail or
> qmail.

Definitely a great idea.  This would enable separate modules
using, for example, MAPI, or some other proprietary mail-sending
and/or mail-receiving architectures, which may be the only
effective way for mail in/out in some settings.  If we focus
on a central module handling an easy-MIME-I/O-message, and
sending and receiving modules respecting some abstract
interface, writing cross-platform mail handling applications
is made much easier -- the only issue becoming one of ensuring
the 'right' implementation-modules are being used for the
specific platform.


Alex





From pehr at pehr.net  Wed Dec  6 00:48:01 2000
From: pehr at pehr.net (pehr anderson)
Date: Wed, 06 Dec 2000 05:48:01 GMT
Subject: Beginers question
References: <_3NW5.44729$nh5.3218436@newsread1.prod.itd.earthlink.net>
Message-ID: <3A2DD311.25AF9EE4@pehr.net>

I think this isn't a big problem.
You should probably get somebody to 
walk you through how to use the
command line. 

It is hard to explain all the 
concepts involved in navigating 
files and directories without
being there with you an introducing
the commands. 

If you can, find a reference on the 
DOS/Windows command line commands.
These haven't changed much in 15 years.

	-pehr

robert george wrote:
> 
> I started the beginers tutorial. My operating system is Winme.
> The tutorial says I should get a" C:\WINDOWS> " prompt. But I get
> a"C:\WINDOWS\DESKTOP>" prompt.
> Why is that? Will Python work with Winme?
> Thanks for your help
> An extreme novice.
> Rob


From jkndeja at my-deja.com  Tue Dec 19 04:33:32 2000
From: jkndeja at my-deja.com (jkndeja at my-deja.com)
Date: Tue, 19 Dec 2000 09:33:32 GMT
Subject: Solution: Direct access to Printer I/O lines
References: <917k9b$9i9$1@nnrp1.deja.com> <slrn93hrjn.2vp.ssthapa@localhost.localdomain> <91cnp3$g59$1@nnrp1.deja.com> <91d23f0jk7@news1.newsguy.com> <91l4co$jua$1@nnrp1.deja.com> <91lhgi0qt@news1.newsguy.com>
Message-ID: <91na1c$egj$1@nnrp1.deja.com>

Hi Alex


[...'how about this']
>
> Seems nice, though I'd call the method which acquires the resource
> 'acquire' rather than 'claim' -- but that's quibbling.

I can live with acquire(), that's fine ;-)

I was also thinking of having a couple of methods of specifying the
actual LPT 'port number'. Although we have LPT1:, LPT2, LPT3: etc, they
don't actually necessarily correspond to the same port addresses across
PCs! Usually, for sure, but I have encountered exceptions, and Ralph
Brown's interrupt list mentions alternative sets of ports for each of
these. so I may have some constants so you can instantiate in a couple
of ways:

L = directlpt.directlpt(1)
# or
L = directlpt.directlpt(0x278)    # would mean the same thing.

This would allow use on a 'non-standard' implementation.

[...]
>
> The only issue (which depends on the device attached to the
> port, alas...) is IF reading is actually 'non-invasive' so it
> can be allowed without requiring 'acquisition'.  I guess in
> most cases it will be, but...
>

It should be fine, because an 'inb()' is only reading the voltages at
the LPT: pins - it's not strobing anything. Any device connected will
not be able to tell that an inb() has occurred.

Thanks for your suggestions on all of this!

    Jon N


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


From pbjorn at uswest.net  Wed Dec 20 23:12:25 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Wed, 20 Dec 2000 21:12:25 -0700
Subject: Overriding sys.stdout
References: <3A417FA8.29EBC549@home.com>
Message-ID: <3A418329.D1871B68@uswest.net>

Something like this should work...

-- bjorn

import StringIO
import sys

class Logger:
 def __init__(self):
  self.buffer = StringIO.StringIO()
  self.save_stdout = sys.stdout
 def close(self):
  sys.stdout = self.save_stdout
 def getvalue(self):
  return self.buffer.getvalue()
 def write(self,txt):
  self.buffer.write(txt)
  self.save_stdout.write(txt)

>>> import log
>>> import sys
>>> l = log.Logger()
>>> sys.stdout = l
>>> print 5,6,7
5 6 7
>>> print "hello %s" % "world"
hello world
>>> l.getvalue()
'5 6 7\012hello world\012'
>>> l.close()



"John Q. Public" wrote:

> I would like to be able to take a snapshot of stdout while a script is
> running, but still allow stdout to be written to the terminal.  My
> current, working, solution is to hack IDLE but I would like a generic
> solution so we would not be bound to a specific IDE.  I would like to
> use PythonWin on NT, if I can, or do the journaling in batch mode.
>
> My plan is to implement a class/module/whatever that I can used to
> overwrite sys.stdout.  This class/module/whatever would write data to a
> file and send it to the old stdout.  My mail questions are these:
>
>     Does such a class/module/whatever exist?
>     If not, does anyone have an idea where I can start for info?  As a
> last resort, I will start wading through IDLE and PythonWin, but I would
> like some ideas first.
>
> Thanks!
>
> Jonathan Polley
> jwpolley(a)collins(d)rockwell(d)com



From SBrunning at trisystems.co.uk  Thu Dec 21 08:00:20 2000
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Thu, 21 Dec 2000 13:00:20 -0000
Subject: Is there any good documentation on MAPI using Python
Message-ID: <31575A892FF6D1118F5800600846864D5B1658@intrepid>

> From:	Henrik M?rtensson [SMTP:henrik.martensson at frontconsulting.se]
> I would like to use the MAPI with the MS office suite programming in
> Python,
> but I'm having a little bit of a problem with finding good documentation.
> 
> Does anybody have any leads?
 
Python Programming on Win32, by Mark Hammond and Andy Robinson, ISBN
1-56592-621-8.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk
It is easier to optimize correct code than to correct optimized code. - Bill
Harlan




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From jkndeja at my-deja.com  Wed Dec 13 05:50:20 2000
From: jkndeja at my-deja.com (jkndeja at my-deja.com)
Date: Wed, 13 Dec 2000 10:50:20 GMT
Subject: Solution: Direct access to Printer I/O lines
Message-ID: <917k9b$9i9$1@nnrp1.deja.com>

Hello
    People have been asking about getting _direct_ access to a printer
port over the last few weeks, and I've already made some comments on
the matter myself.

By 'direct' in this case, I'm assuming that people want to be able to
individually twiddle the bits of the printer port, so that they can
light LEDs, read switches, control a nuclear power station ;-), etc.
etc.

As several people have pointed out, this is very OS-dependant.
Nevertheless, it's prompted me to write a simple python extension DLL.
This should allow access to the ports of LPT1, LPT2, and LPT3 from
Python, under windows 95 at least (see below for caveats)

You can download a .ZIP file containing the .C source, and the DLL:

http://www.nicorp.f9.co.uk/download/python-ioport.zip

CAVEATS:
    - This is untested!! in the sense that I have not yet got my I/O
hardware connected up, and therefor can't verify that it actually does
what I intend it to. It does behave as expected in other ways though. I
would be very keen to learn of experiences in using it.
    - It's intended to work under Windows 95 only. It won't work under
NT (you need a kernel-mode driver to access I/O ports under NT, and I
haven't written that yet ;-). I don't know about W98. It won't work
under linux either ;-). I'm happy to write a similar extension for
linux if people are interested - it's good practice for me...
    - to be safe, I have restricted the range over which I/O port
access is allowed. This is currently limited to the address spaces of
the LPT ports:

    0x278 to 0x27a (LPT1:)
    0x378 to 0x37a (LPT2:)
    0x3bc to 0x3bf (LPT3:)

    You'll get a ValueError exception
        "ValueError: Access to I/O ports outside LPT[1-3]: not allowed"

    if you try to go outside this range. If you want to remove this
lmitation, you'll have to modify and recompile the code, in which case
I assume you know what you're doing...

    - Oh, I built it with Python v1.5.2 (ie. this is the import Library
I used). Does this mean it will need re-building for V2.0? not sure...

Installation and use:
    - download the .ZIP file
    - unzip, and put the .DLL file in the appropriate directory. On my
W95 setup, it's in <Python Base Dir>\DLLs

Use like this:

import ioport

print ioport.inb(0x278)    # byte read
ioport.outb(0x279, 0x00)   # byte write: (port, val)

# you also have word I/O
# print ioport.inw(0x278)      # word read
# ioport.outw(0x279, 0x0f0f)   # word write: (port, val)

I haven't even checked yet which of the LPT: port addresses are Read-
only, which are Write only, and which are Read/Write - that's how
tested this is! Track down Ralph Browns Interrupt list for the gen. on
this.

Despite all these caveats, I do actually expect that this will work!
I'm happy to receive reports: mailto:jkn at _removethis_.nicorp.f9.co.uk

    Cheers
    jon Nicoll






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


From eliz at is.elta.co.il  Sat Dec 23 10:04:53 2000
From: eliz at is.elta.co.il (Eli Zaretskii)
Date: Sat, 23 Dec 2000 17:04:53 +0200
Subject: Problem with python mode
References: <3a44bcf1.73260781@nntp.interaccess.com>
Message-ID: <3A44BF14.F7721A06@is.elta.co.il>

"Thaddeus L. Olczyk" wrote:
> 
> When i try loading pdb.el in NTemacs I keep getting an error--
> Symbol's function defiition void: toolbar-make-button-list.
> Any idea what the problem is?

This is an XEmacs function, it doesn't exist in NTEmacs.

Are you sure pdb.el works with Emacs?


From fredrik at effbot.org  Tue Dec 19 19:48:19 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 20 Dec 2000 00:48:19 GMT
Subject: xml-rpc module question
References: <dmC%5.75166$yR4.2151642@news1.rdc1.tx.home.com> <91ooea$9qb$1@bunyip.cc.uq.edu.au> <DeT%5.76385$yR4.2212474@news1.rdc1.tx.home.com>
Message-ID: <nlT%5.653$Qb7.44417@newsb.telia.net>

Scott Hathaway wrote:
> Thanks for the response.  When I try your suggested code, here is the error
> I get:
>
> C:\Projects\core>python core_client.py
> Traceback (most recent call last):
>   File "core_client.py", line 5, in ?
>     print testing.CubeIt(3)
>   File "xmlrpclib.py", line 547, in __call__
>     return self.__send(self.__name, args)
>   File "xmlrpclib.py", line 627, in __request
>     response = self.__transport.request(
> AttributeError: 'int' object has no attribute 'request'

you're passing in a bogus second argument to the xmlrpclib
Server constructor.  don't do that.

</F>




From thomas at xs4all.net  Thu Dec 28 11:32:00 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Thu, 28 Dec 2000 17:32:00 +0100
Subject: Interesting behaviour of the assignment
In-Reply-To: <92f4gk$qk6$1@news.nuri.net>; from junaftnoon@nospamplzyahoo.com on Thu, Dec 28, 2000 at 07:23:54PM +0900
References: <92f4gk$qk6$1@news.nuri.net>
Message-ID: <20001228173200.B1811@xs4all.nl>

On Thu, Dec 28, 2000 at 07:23:54PM +0900, June Kim wrote:
> >>> a=100
> >>> b=100
> >>> a is b
> 0
> >>> a=99
> >>> b=99
> >>> a is b
> 1

> Is this an intended/expected result? What is the lying structure beneath
> this?

Yes, this is intended. The 'is' operator is the 'identity' operator, it
returns true if and only if the left hand side of the expression refers to
the same object (not similar, not the same type, not equal, but the *same*)
as the right hand side. Numbers below 100 are always 'unique', they are
created only once, cached and reused whenever you need them. Numbers above
and including 100 are not cached like that, but because of the way the
python constants work, all the constants with the same value end up as the
same object, regardless. Unfortunately, the interactive session behaves
slightly different than a script, so you don't see that behaviour there.

Here is an example:

locust:~> cat spam.py
a = 42
b = 42
c = 41+1  
d = 43-1  

e = 666   
f = 666
g = 665+1
h = 667-1

if (a, b, c, d) == (42,)*4:
    print "a, b, c, d are all 42"

print "id(a) =", id(a)
print "id(b) =", id(b)
if a is b:
    print "a is b"
else:
    print "a is not b"

print "id(c) =", id(c)
print "id(d) =", id(d)
if c is d:
    print "c is d"
else:
    print "c is not d"

if (e, f, g, h) == (666,)*4:
    print "e, f, g, h are all 666"

print "id(e) =", id(e)
print "id(f) =", id(f)
if e is f:
    print "e is f"
else:
    print "e is not f"

print "id(g) =", id(g)
print "id(h) =", id(h)
if g is h:
    print "g is h"
else:
    print "g is not h"

locust:~> python spam.py
a, b, c, d are all 42
id(a) = 134528380
id(b) = 134528380
a is b
id(c) = 134528380
id(d) = 134528380
c is d
e, f, g, h are all 666
id(e) = 134657012
id(f) = 134657012
e is f
id(g) = 134656988
id(h) = 134656976
g is not h

(If you cut&paste the above code into an interactive session, you'll see
that it'll claim that 'e is not f'. That's because constants are handled
slightly differently in the interactive read-eval-print-loop.)

End result ? Don't use 'is' unless you really, really need an identity
check, and not a value check. It's a damned fast check (a single pointer
comparison) but it isn't enough if you really mean '==', not 'is'.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From jweirich at my-deja.com  Tue Dec  5 19:10:58 2000
From: jweirich at my-deja.com (jweirich at my-deja.com)
Date: Wed, 06 Dec 2000 00:10:58 GMT
Subject: "inherited" keyword in Python?
References: <3A2B9EFA.3A64C13B@schlund.de> <90gevs02uu1@news2.newsguy.com>
Message-ID: <90k06d$25n$1@nnrp1.deja.com>

In article <90gevs02uu1 at news2.newsguy.com>,
  "Alex Martelli" <aleaxit at yahoo.com> wrote:
> "Carsten Gaebler" <cg at schlund.de> wrote in
message
> news:3A2B9EFA.3A64C13B at schlund.de...
> > Hi there!
> >
> > Does Python know something like the
"inherited" keyword from Borland
> > Pascal? In Pascal it is used to call a method
of a class's superclass,
> > i.e. instead of saying "superclass.method()"
you say "inherited
> > method()". Can be quite helpful if you change
the name of the superclass
> > or just forgot it. :-)
>
> Python has multiple inheritance, so there is (in
general) no single
> 'superclass' that you could use that way.

I've heard this argument before, but am not
convinced by it.  The resolution that makes sense
to me is to define "super.method()" to invoke the
method that would have otherwise been invoked if
"self.method()" were not defined in the current
class.


Sent via Deja.com http://www.deja.com/
Before you buy.


From sorular at netscape.net  Thu Dec 14 01:23:24 2000
From: sorular at netscape.net (sorular)
Date: Thu, 14 Dec 2000 06:23:24 GMT
Subject: PgSQL and double quotes...
References: <914i7a$q0a$1@nnrp1.deja.com> <5OTZ5.774$4y4.27551@e420r-atl1.usenetserver.com>
Message-ID: <3a3863c2.288751172@news-server.bigpond.net.au>

On Wed, 13 Dec 2000 18:38:02 -0500, "Steve Holden"
<sholden at holdenweb.com> wrote:

>SQL data containing single quotes can be problematic, although you should
>not experience any difficulty with double quotes.
>
>> I wrote a simple class before the execute sintax
>> as :
>>   string.join(string.split(text, "'"), "''")
>> but that still doesnt get rid of this "parsing
>> error" from PgSQL... anyone has any ideas?
>>
>This seems to be a fairly convoluted way to write
>
>    string.replace(text, "'", "''")
>
>but appears to give the same result, at least in common cases (I didn't do
>exhaustive testing).
>
>This works for me using mxODBC to drive SQL Server 6.5 and 7.0, as well as
>Access.  So I can only conclude that either it's the fact you're working
>with bonary data or some implementation deficiency of PgSQL (not, I must
>say, a system I am familiar with).
>
>A small code example might allow the group to exercise its collective mind a
>little more creatively on solving your problem, if you can isolate the
>problem to a ten- or twenty-line test case.
>
>regards
> Steve


thanks for the reply. Just last night, after a careful re-read of psql
docs and a great help from "Bill.Allie at mug.org", I've solved the
problem by replacing the line :

self.cursor.execute("INSERT INTO log (id,type, command,ip_addr) \
values (%d,'%s','%s','%s')" % (conn_id, write_type, data, ip_addr))

with

self.cursor.execute("INSERT INTO log (conn_id, write_type,command,\
ip_addr)  VALUES (%s,%s,%s,%s)" ,conn_id, write_type,data, ip_addr)
		     ^^		^^

which sends following command to postgreSQL:

insert into test values(id,type,data, ip_addr)

execute statement fixes the double/single quotes etc...

ciao

Aus

 )


Note that % and () are excluded and all the data is %s with no %d...


From jbauer at rubic.com  Fri Dec 29 09:58:40 2000
From: jbauer at rubic.com (Jeff Bauer)
Date: Fri, 29 Dec 2000 08:58:40 -0600
Subject: Python with Apache on WinNT newbie question
References: <3a4b6391.4061375@localhost>
Message-ID: <3A4CA6A0.2629456A@rubic.com>

Victor,

Here's a trick I once used to maintain compatibility between
Windows/Unix CGI scripts. Create a C:\usr\local\bin directory
and copy the C:\Python\python.exe file into it. 

Then use the following header line on your portable CGI scripts:

            #!/usr/local/bin/python

I haven't tried this kludge recently, but I think it should
still work.

Jeff Bauer
Rubicon


Victor Muslin wrote:
> 
> To get Apache server to run a Python CGI script, the script needs to
> have on its first line a directive telling Apache the file name of the
> Python interpreter, like so:
> 
>         #!d:/apps/python/python.exe
>         #
>         # The rest of my cgi script follows
>         # ....
> 
> The problem with this is deploying this script on different machines
> that may have Python interpreter installed in different directories.
> Surely others had run into the same issue. What is a typical solution?
> Unfortunately, if this directive can take environmental variables I
> could not figure out how to do it.



From aleaxit at yahoo.com  Fri Dec 29 18:23:34 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 30 Dec 2000 00:23:34 +0100
Subject: 2001 Enchancement Wishlist
References: <3A4CEACB.220998A0@javanet.com>
Message-ID: <92j6i901g7@news1.newsguy.com>

"Raymond Hettinger" <othello at javanet.com> wrote in message
news:3A4CEACB.220998A0 at javanet.com...
> If these ideas meet with favor, I'll write-up a PEP.
>
> 1.  List.append(elem) and List.pop() are great.  Let's add two more
>      methods, List.prepend(elem) and List.leftpop().  The goal is to

Pass (apart from the unhappy names); I think .insert, and slice
assignment in general, are by far enough, but it's a min.r.

> 2. There is already a PEP to extend the eval() function which is currently
>     limited to dictionaries.  The goal is to accept any object defining
>     __getitem__() so that any mapping object can be used to specify the
>    globals and locals.   I recommend that we extend the same courtesy to
>    the % string formatting operator.  For example:

Sure, the more polymorphism, the merrier.  That one I'd like to see.

> 3.  __init__ is the only special method that doesn't allow a return value.

I think it's better off without one, just as it is now.  Use factory
functions if you want to generate with more freedom, and let
class objects as callable make-and-initialize new instances.

> 4.  Allow dictionaries to respond to sequence operations as if .keys()
were
>      in the list object.  Examples:
>
>     knights = { 'lancalot':'good', 'gallahad':'brave', 'robin':'coward' }
    [snip]
>     print knights[0]              # may be any one depending on the hash
table

Unfeasible.  0 (or any other int) might be a key (and it may happen to
fall anywhere in the hash table order, too).  So what does knights[x]
mean -- the value corresponding to key x is present, as now; or, the
value corresponding to key x is present, or else if x is (or, can be
coerced to?) int, the x-th key in hashtable order?  The latter is far
too complex -- and all the complexity for very little gain.

This one would be a real blooper if Python adopted it, IMHO.


Alex





From jepler at inetnebr.com  Fri Dec 15 08:58:47 2000
From: jepler at inetnebr.com (Jeff Epler)
Date: Fri, 15 Dec 2000 13:58:47 GMT
Subject: xreadlines (was Re: while true: !!!)
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <slrn93g37o.9tj.jepler@potty.housenet> <f3%Z5.146409$U46.4768601@news1.sttls1.wa.home.com> <91a6it02flt@news2.newsguy.com>
Message-ID: <slrn93k8sn.lb.jepler@potty.housenet>

On Thu, 14 Dec 2000 11:14:38 +0100, Alex Martelli
 <aleaxit at yahoo.com> wrote:
>...if it was able to iterate over already-opened file-like objects, rather
>than having to open actual files itself, it WOULD conceptually solve it
>all.  I wonder how hard it would be to generalize fileinput in this way --
>probably not very (the in-place-rewriting option would be incompatible
>with having already-opened file-like objects in the list, of course).

xreadlines takes a file object, not a filename, as its argument.

>If [using readline] was a performance problem, it could of course
>also be fixed in a future fileinput version without changing code that
>uses it (again, in-place-rewriting would probably have to inhibit the
>optimization, although that isn't entirely clear).

xreadlines internally uses fp.readlines(sizehint), so 
	for line in xreadlines.xreadlines(fp): pass
is just about as fast as
	for line in fp.readlines(): pass
except that it will not require storage for the whole file.

Jeff


From vasvir at iit.demokritos.gr  Sun Dec 31 15:54:18 2000
From: vasvir at iit.demokritos.gr (Vassilis Virvilis)
Date: Sun, 31 Dec 2000 22:54:18 +0200
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <92m9pm$d19$1@nnrp1.deja.com> <B3z36.122$9s1.18804@e420r-atl2.usenetserver.com> <3A4EC5B1.8912E571@engcorp.com> <x8G36.156$9s1.31209@e420r-atl2.usenetserver.com> <3A4F4B4D.ECD6B172@engcorp.com>
Message-ID: <3A4F9CFA.89C35032@iit.demokritos.gr>

Peter Hansen wrote:

[snip]
> 
> But I can't argue with another view on this.  After all, in C I've been
> using "brace indentation style 3" for the longest time, in apparent
> violation of 98% of coding conventions in the world.  That is, my braces
> are vertically aligned but indented with respect to the "if ()" or "for
> ()" which gets me into the block:
> 
> while (1)
>    {
>    T = f();
>    if (T < 25)
>        break;
>    }
> 
> Use of that convention is probably what has made me quite happy with
> Python's indentation style, which essentially matches this pattern.

I am writing C like this also. Should we form a club or something? :)

	.Bill


From hamish_lawson at yahoo.co.uk  Thu Dec  7 06:05:29 2000
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: Thu, 07 Dec 2000 11:05:29 GMT
Subject: PIL binary for Python 2.0 for Windows ?
References: <3a2f472f$0$235@hades.is.co.za>
Message-ID: <90nqto$447$1@nnrp1.deja.com>

Leon Booyens wrote:

> Where can I find a binary distribution for Windows of the Python Image
> Library 1.1.1.

The bulk of PIL is Python source, which you'll still need to get and
install to where you want. For those core parts that are written in C,
you can get compiled DLLs from

ftp://ftp.reportlab.com/win32-dlls/2.0/

and put them into your installed PIL directory.


Hamish Lawson


Sent via Deja.com http://www.deja.com/
Before you buy.


From blosh at tsbrass.com  Mon Dec 11 15:54:23 2000
From: blosh at tsbrass.com (Brad)
Date: Mon, 11 Dec 2000 20:54:23 GMT
Subject: HPUX install
Message-ID: <913ett$u0b$1@nnrp1.deja.com>

We are running HPUX 10.20 and I was wondering why I cant get python to
install.

./configure seems to work

make does not...it gives the following error

  cd Parser ; make OPT="-O" VERSION="2.0" \  prefix="/usr/local"
exec_prefix="/usr/local" all  cc -Ae -O -I./../Include -I.. -DHAVE_CONFIG_H
-c pgenmain.c (Bundled) cc: warning 480: The -A option is available only with
the C/ANSI C pro duct; ignored. (Bundled) cc: warning 480: The -O option is
available only with the C/ANSI C pro duct; ignored. cpp: "pymem.h", line 7:
error 4036: Can't open include file 'pyport.h'. *** Error exit code 1

Stop.
*** Error exit code 1

Stop.


any help is appreciated


Sent via Deja.com http://www.deja.com/
Before you buy.


From arcege at shore.net  Thu Dec  7 10:58:59 2000
From: arcege at shore.net (arcege at shore.net)
Date: Thu, 07 Dec 2000 15:58:59 GMT
Subject: Accessing "variables" in an extension module
References: <3A2E4B01.4609EE7E@integration-x.com>
Message-ID: <7tOX5.146$JG.14399@news.shore.net>

Christian Wermuth <christian at integration-x.com> wrote:
: I would like to access "variables" (chars and ints) in a C module
: extension for Python from my Python code. Is that possible?
: If I wanted to improve the example from "Extending and Embedding the
: Python Interpreter" like shown below, what would I need to add to the
: spam.c code?

: import spam
: spam.varstring = "test"
: spam.varint = 5
: spam.system('dir c:')

I'm assuming that you want to have predefined variables and constants.
You might want to look at the section 'Adding "constants"' (which, in
Python, are really variables anyway) in the "Python Extension writing"
document.  That might help you.

The URL for the document is
<URL: http://starship.python.net/crew/arcege/extwriting/pyext.html>

Remember that modules, whether a C extension or not, are mutable and
you can always add or remove names there.

  -Arcege



From dnew at san.rr.com  Tue Dec  5 19:44:33 2000
From: dnew at san.rr.com (Darren New)
Date: Wed, 06 Dec 2000 00:44:33 GMT
Subject: an apply question
References: <90jsj0$24o$1@front6.grolier.fr>
Message-ID: <3A2D8BF1.6F655361@san.rr.com>

And in addition to all the other advice, you've defined func1 but you're
trying to call func3. You either need to define func3 or change "if
key=="Key3"" to "if key=="Key1"". :-)

> def func1 ( ):
>     print "Here's func1"
> 
>     if key == "Key3":
>         apply( func, ( ) )

-- 
Darren New / Senior MTS & Free Radical / Invisible Worlds Inc.
San Diego, CA, USA (PST).  Cryptokeys on demand.
There is no "P" on the end of "Winnie the Pooh."


From root at rainerdeyke.com  Mon Dec 18 21:39:38 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Tue, 19 Dec 2000 02:39:38 GMT
Subject: while true: !!!
References: <mailman.977180341.10206.python-list@python.org> <slrn93t7bv.le.grey@teleute.rpglink.com>
Message-ID: <KTz%5.48644$x6.22165257@news2.rdc2.tx.home.com>

"Steve Lamb" <grey at despair.rpglink.com> wrote in message
news:slrn93t7bv.le.grey at teleute.rpglink.com...
> On Mon, 18 Dec 2000 23:32:23 +0100 (MET), Carsten Geckeler
<nospam at no.spam>
> wrote:
> >`while true' would be more readable and clear than `while 1'.  By the
way,
> >explain a newbie that -1 is true. ;)
>
>     0 is true, all else is false.  And?  :P

0 is false, None is false, [] is false, () is false, {} is false, and x is
false where x is a class instance and x.__nonzero__() returns false.
Everything else is true.  I think.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From gregj at pobox.com  Tue Dec  5 02:22:27 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Tue, 05 Dec 2000 07:22:27 GMT
Subject: Script vs Program
References: <y39V5.524$OC.159868@typhoon.aracnet.com> <bHnV5.96503$U46.3094705@news1.sttls1.wa.home.com> <jDFV5.102058$U46.3208014@news1.sttls1.wa.home.com> <3A27BAB5.CF6D060D@san.rr.com>
Message-ID: <TI0X5.116132$U46.3743208@news1.sttls1.wa.home.com>

"Darren New" <dnew at san.rr.com> wrote:

> Java might be considered a scripting language if you think it was designed
> only for being embedded on web pages, like JavaScript seems to have been.

Actually JavaScript used to be called LiveScript, and it was the server-side
scripting language of the Netscape LiveWire web server. Netscape renamed the
language and decided it would be cool to enable it in the web client.
Microsoft was still trailing Netscape at the time so they put JavaScript
into Internet Explorer to stay compatible with Netscape.

According to legend Java (formerly called Oak) predates the web. It was
originally conceived as a cross-platform language for connecting disparate
networked devices. Embedding Java in web pages was apparently a publicity
stunt engineered by Sun, a failed attempt to replace Windows with the write
once, debug everywhere platform.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From joconnor at cybermesa.com  Tue Dec  5 23:27:51 2000
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Tue, 05 Dec 2000 21:27:51 -0700
Subject: Settle Down ( was Re: Python Origins)
In-Reply-To: <90k56f$dhc$1@news.wrc.xerox.com>
References: <90k3pq$50p$1@nnrp1.deja.com>
Message-ID: <3.0.1.32.20001205212751.008eddf0@cybermesa.com>

At 01:36 AM 12/06/2000 GMT, Mark Jackson wrote:
>Nigel <nigeljennison at cs.com> writes:
>>  I am doing some research and I would be grateful if anybody could give
>> me some history on the origins of Python.
>
>But then you wouldn't really be doing any research, would you?


Nigel didn't say *what* he was researching.

He could be researching

a) Web languages and their development
b) History of OO langauges and what languages have branched from which
other languages
c) Scripting languages
d) Dynamically typed versus Statically typed languages

There's a lot of things he could be researching of which knowing the
history of Python would be but a part of a larger picture for which getting
some pointers to sources could be useful without doing his workf for him

It'd be really bad, for example, if Nigel was writing a a magazine article
about Python ascending in prominence on the web and when trying to get some
history information, the best help he got from the Python community was
snotty or sarcastic answers


Take care,
Jay


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

"God Himself plays the bass strings first, when He tunes the soul"



From amk at mira.erols.com  Sun Dec 24 10:54:24 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 24 Dec 2000 15:54:24 GMT
Subject: Python Database Interfaces - Any Standard?
References: <RUo16.204045$RP.37382354@news1.rdc1.md.home.com>
Message-ID: <slrn94c7qm.m4.amk@207-172-57-128.s128.tnt2.ann.va.dialup.rcn.com>

On Sun, 24 Dec 2000 15:48:01 GMT, Brian Knox <brianknox at home.com> wrote:
>One huge question I have about Python is: is there any sort of standard for
>database access with Python? I've been coding Perl DBI apps as long as I've

See the DB-API at http://www.python.org/topics/database/DatabaseAPI-2.0.html.
There's no standard module, though mxODBC is probably close, since it 
supports the largest number of databases.

--amk



From gbreed at cix.compulink.co.uk  Thu Dec 28 11:11:42 2000
From: gbreed at cix.compulink.co.uk (gbreed at cix.compulink.co.uk)
Date: 28 Dec 2000 16:11:42 GMT
Subject: Basic MAPI
References: <91vfsb02cn6@news2.newsguy.com>
Message-ID: <92fonu$auu$1@plutonium.compulink.co.uk>

In article <91vfsb02cn6 at news2.newsguy.com>, aleaxit at yahoo.com (Alex 
Martelli) wrote:

> "Henrik M?rtensson" <henrik.martensson at frontconsulting.se> wrote in 
> > message
> news:nEG06.1008$Qb7.95625 at newsb.telia.net...
> > Hi
> >
> > I'm trying to run a basic MAPI initialisation, but I get stuck with 
> > the
> > dispatcher.
>     ...
> > >>> s=Dispatch("Mapi.Session")
> 
> There is no creatable coclass (that I ever heard of) with progId
> 'Mapi.Session'.

There most certainly is.  It comes with Outlook, among other things, and 
I think lives in cdo.dll ("MAPI" is an abbreviation for "CDO" or 
something).  That Dispatch works fine on my machine, so I don't know 
what's going wrong.  Check that Outlook's installed.


                      Graham


From donn at u.washington.edu  Thu Dec 14 12:45:16 2000
From: donn at u.washington.edu (Donn Cave)
Date: 14 Dec 2000 17:45:16 GMT
Subject: split weirds
References: <QI2l1FAXBMO6EwD6@jessikat.demon.co.uk> <GZXwhJASUNO6EwxV@jessikat.demon.co.uk>
Message-ID: <91b0vc$ccc6$1@nntp6.u.washington.edu>

Quoth Robin Becker <robin at jessikat.fsnet.co.uk>:
| In article <QI2l1FAXBMO6EwD6 at jessikat.demon.co.uk>, Robin Becker
| <robin at jessikat.fsnet.co.uk> writes
| Can somebody explain why strip behaves differently with and without an
| argument?
| 
| >>> from string import split
| >>> split(' ')
| []
| >>> split(' ',' ')
| ['', '']
| >>> split('  ')
| []
| >>> split('  ',' ')
| ['', '', '']
| >>> 
| 
| with the default whitespace arg seems as though a run of whitespace is
| being treated as a single character.

| several people have pointed out that arbitrary whitespace in the docs
| means that so the no arg case really is special. I still find it a bit
| strange that
|
| split(' word ') ==> ['word'], but then the docs use the term separated
| for the no arg case which can't be done for absent words. 

Maybe it would comforting to think of the default separator as
"not non-white space."

In ' word ', 'w' is non-white, but ' ' is not non-white -- and
so is "end of data" not non-white.  So the split classifies this
string as three parts, separator,separated,separator.

Given a specific separator, say ':', and the equivalent string
':word:', null end of data does not qualify as separator, so the
same analysis yields 5 parts.

However you justify it, it's mighty convenient in practice.

	Donn Cave, donn at u.washington.edu


From urner at alumni.princeton.edu  Tue Dec 19 13:18:34 2000
From: urner at alumni.princeton.edu (Kirby Urner)
Date: Tue, 19 Dec 2000 10:18:34 -0800
Subject: Python in "math through programming" curriculum
References: <2g2t3t06r2agrrjmfnr41eu4lc2u39a8qi@4ax.com> <3A3EFF47.67DE69CF@gte.net> <513v3toslkjkggh9oumlf20qc9bi1gdtif@4ax.com> <13M%5.465$AH6.37131@newsc.telia.net>
Message-ID: <ab9v3tsc3el3r3psh8r6c32p1u86tp7uf3@4ax.com>

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

>Kirby Urner wrote:
>> Maybe the Mac people are working on alternatives/enchancements
>> to IDLE?  Does anyone have some info on this?
>
>    http://starship.python.net/crew/just/
>
>    "Yes, Guido is my brother"
>
>    "I am the author of the IDE for MacPython. This place
>    used to be the download spot for it, but since Python
>    1.5.1 it is included in the standard MacPython distribution.
>    If you can't find it: it lives in :Mac:Contrib:IDE. Starting
>    with MacPython 1.5.2b1 /.../ the IDE is included in the
>    distribution with full source code and a Python Style
>    License (PSL)."
>
></F>

Yes, when I installed Python 1.x for the iMac at my daughter's
school, it came with IDLE and I used IDLE after installing it.

But the key words didn't seem to be color coded, i.e. when 
I typed 'import x', the word 'import' did not automatically 
turn orange (or any other color).  This detracted from the 
kid-friendliness of the interface.  Especially when coding 
modules, you want color-coded key words, as per what any 
high-end programming editor provides.

My question is:  did I miss something, e.g. some setting, 
in the IDLE interface? and/or does Macworld offer alternative
GUI IDEs for Python, e.g. in the Windows world there are 
several?

><!-- (the eff-bot guide to) the standard python library:
>http://www.pythonware.com/people/fredrik/librarybook.htm
>-->

Have it, use it.

Kirby



From slhath at home.com  Thu Dec 28 17:30:31 2000
From: slhath at home.com (Scott Hathaway)
Date: Thu, 28 Dec 2000 22:30:31 GMT
Subject: automating excel with python ... getting excel to go away
Message-ID: <baP26.99236$yR4.2642456@news1.rdc1.tx.home.com>

Hello,

I am trying to automate excel with python.  Here is my code:

 import win32com.client, os
 myExcel = win32com.client.Dispatch('Excel.Application')
 myExcel.Workbooks.Open('c:\\test.html')
 myExcel.Visible = 0
try:
  os.remove(newReport)
 except:
  pass
 myExcel.ActiveWorkbook.SaveAs('c:\\test.xls)
 myExcel.Quit
 myExcel = None

This code works fine, in that the HTML file is converted to xls and saved,
but the problem is that excel is still running in the background when the
script is through.  How can I get excel to go away?

Thanks,
Scott




From tim.one at home.com  Fri Dec 22 21:37:22 2000
From: tim.one at home.com (Tim Peters)
Date: Fri, 22 Dec 2000 21:37:22 -0500
Subject: Good introduction to functional programming with Python?
In-Reply-To: <slrn947ebd.qfb.qrczak@qrnik.zagroda>
Message-ID: <LNBBLJKPBEHFEDALKOLCAECPIFAA.tim.one@home.com>

[Alex Martelli]
> There must indeed be some Pythonic 'aura' to Haskell, I guess,
> given the somewhat surprising correlation between liking Haskell
> and liking Python among people who've tried both.

[Marcin 'Qrczak'  Kowalczyk]
> Indeed. I am very surprised how often Haskell is mentioned on the
> Python's list, given quite different paradigms of these languages.
>
> Elegance of design of a language? Significant layout? :-)

My vote:  both languages place extreme emphasis on being *readable*.  Listen
to a Scheme True Believer ragging on a Haskeller for all the "unnecessary"
syntactic sugar in Haskell, then go back in the c.l.py archives and watch
the same folks ragging on Python for the same thing. Languages' detractors
often stumble into key similarities years before the supporters figure it
out <wink>.

prophecy-is-remembering-the-past-ly y'rs  - tim




From smnordby at yahoo.com  Wed Dec 13 08:07:18 2000
From: smnordby at yahoo.com (Steve)
Date: Wed, 13 Dec 2000 05:07:18 -0800
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <VrtZ5.3567$Kd1.415273@newsb.telia.net>
Message-ID: <3A377486.C986273@yahoo.com>

Yes, I should have prefaced my post with: "The files I read are less
than 3000 lines and less than 150kB, so I can get away with using a for
loop". 

-Steve-

Fredrik Lundh wrote:
> 
> "Steve" wrote:
> > I dislike infinite while loops and use:
> >
> > for line in file.readlines()
> >     ...do stuff...
> 
> what makes you so sure that's not an infinite loop?
> 
> (well, in the current implementation, it isn't: if your file
> has more than 2147483647 lines, your code won't read
> all of them... ;-)
> 
> </F>


From glenfant at equod.com.nospam  Thu Dec 14 08:22:09 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Thu, 14 Dec 2000 14:22:09 +0100
Subject: comp.lang.python.announce
Message-ID: <91aha7$791$1@reader1.imaginet.fr>

Hi,

My ISP's news server does not host "comp.lang.python.announce".
It is not hosted at deja.com too.
Is there a public Web interface or read only news server to this newsgroup
somewhere ?

TIA



From entropiamax at jazzfree.com  Mon Dec 18 03:20:35 2000
From: entropiamax at jazzfree.com (entropia)
Date: Mon, 18 Dec 2000 09:20:35 +0100
Subject: Error evaluating numbers starting in zero
References: <3A3D07BD.85EFEB3D@wag.caltech.edu> <87k88ysv93.fsf@lisboa.ifm.uni-kiel.de>
Message-ID: <3A3DC8D2.9E6207EA@jazzfree.com>

Python, like most languages interpret a number starting with 0 as a octal
(and 0x as a hexadecimal). 08 doesn't exists in octal (8=010 in octal).
Because this you should use int("08") and not eval("") OR
 eval("010")-->8
 int("08")-->8


Janko Hauser escribi?:

> "Richard P. Muller" <rpm at wag.caltech.edu> writes:
>
> > "01/08/99", python runs into an error when it tries to eval("08").
> >
> The use of int("08") may help
>
> HTH,
> __Janko
>
> --
>   Institut fuer Meereskunde             phone: 49-431-597 3989
>   Dept. Theoretical Oceanography        fax  : 49-431-565876
>   Duesternbrooker Weg 20                email: jhauser at ifm.uni-kiel.de
>   24105 Kiel, Germany
> --
> http://www.python.org/mailman/listinfo/python-list




From kragen at dnaco.net  Sat Dec 30 18:00:53 2000
From: kragen at dnaco.net (Kragen Sitaker)
Date: Sat, 30 Dec 2000 23:00:53 GMT
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <slrn94shbe.150.grey@teleute.rpglink.com> <X8s36.2731$Qb7.305339@newsb.telia.net> <slrn94smds.150.grey@teleute.rpglink.com>
Message-ID: <FOt36.17$vd1.2224@e420r-atl2.usenetserver.com>

In article <slrn94smds.150.grey at teleute.rpglink.com>,
Steve Lamb <morpheus at here.not.there> wrote:
>On Sat, 30 Dec 2000 21:08:07 GMT, Fredrik Lundh <fredrik at effbot.org> wrote:
>Sorry, *plonk*.  You want to argue for
>argument's sake, fine.  But I don't think I need any of your poor advice.

Hahaha!  You just plonked Fredrik Lundh in comp.lang.python.  He was
being helpful and polite, and you plonked him because you disagreed
with him.  I can guess how much help you'll get next time you have a
Python question.  :)

I think I'm going to save this entire thread; it's just such a
delightful example of how to make a fool of yourself on Usenet.
-- 
<kragen at pobox.com>       Kragen Sitaker     <http://www.pobox.com/~kragen/>
Perilous to all of us are the devices of an art deeper than we possess
ourselves.
       -- Gandalf the White [J.R.R. Tolkien, "The Two Towers", Bk 3, Ch. XI]




From dale at out-think.NOSPAMco.uk  Mon Dec  4 09:17:14 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Mon, 04 Dec 2000 14:17:14 +0000
Subject: Printable time
Message-ID: <7m9n2tcp5ff10o97sr1d57qokakjqsa2m9@4ax.com>

Is there a more succinct way of getting the current local time as text than:

time.strftime('<format>', time.localtime(time.time()))

?
Thanks
--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants




From victor at prodigy.net  Thu Dec 28 16:57:22 2000
From: victor at prodigy.net (Victor Muslin)
Date: Thu, 28 Dec 2000 21:57:22 GMT
Subject: Python with Apache on WinNT newbie question
References: <3a4b6391.4061375@localhost> <slrn94n584.2cg.gerhard.nospam@lilith.hqd-internal.de>
Message-ID: <3a4bb647.25235453@localhost>

Thanks. Though my subject mentions NT, I would also like to know how
this is typically addressed on Unix/Linux? Symbolic links?

On Thu, 28 Dec 2000 20:21:34 +0100, gerhard.nospam at bigfoot.de (Gerhard
=?iso-8859-1?Q?H=E4ring?=) wrote:

>I recommend to add the following line to httpd.conf under WinNT:
>
>ScriptInterpreterSource Registry
>
>and give your Python scripts the extension ".py". A Python CGI script can
>then look like:
>
>#!/usr/bin/env python
>print 'Content-type: text/plain\n\nHello!'
>
>and work perfectly under WinNT and Linux!
>
>Gerhard H?ring
>-- 
>mail:   gerhard <at> bigfoot <dot> de
>web:    http://highqualdev.com
>
>On Thu, 28 Dec 2000 16:07:07 GMT, Victor Muslin <victor at prodigy.net> wrote:
>>To get Apache server to run a Python CGI script, the script needs to
>>have on its first line a directive telling Apache the file name of the
>>Python interpreter, like so:
>>
>>	#!d:/apps/python/python.exe
>>	#
>>	# The rest of my cgi script follows
>>	# ....
>>
>>The problem with this is deploying this script on different machines
>>that may have Python interpreter installed in different directories.
>>Surely others had run into the same issue. What is a typical solution?
>>Unfortunately, if this directive can take environmental variables I
>>could not figure out how to do it.



From lists at TheWebsons.com  Sat Dec 23 08:30:24 2000
From: lists at TheWebsons.com (Ben Ocean)
Date: Sat, 23 Dec 2000 05:30:24 -0800
Subject: Setting Environment Variables
Message-ID: <LOBBJMHOIFOLFIPFANHJAEFGCFAA.lists@TheWebsons.com>

Thanks. I did as you suggested, adding the path to my *webmaster* account,
but it's not part of the sys.path, which (I presume) is what I want and
need:

thor:/home/webmaster# python
Python 2.0 (#1, Oct 16 2000, 18:10:03)
[GCC 2.95.2 19991024 (release)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> import sys
>>> sys.path
['', '/usr/local/lib/python2.0', '/usr/local/lib/python2.0/plat-linux2',
'/usr/local/lib/python2.0/lib-tk', '/usr/local/lib/python2.0/lib-dynload',
'/usr/local/lib/python2.0/site-packages']

Please advise.
TIA,
BenO

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of David Porter
Sent: Saturday, December 23, 2000 4:18 AM
To: Python List
Subject: Re: Setting Environment Variables


* Ben Ocean <beno at TheWebsons.com>:

> I'm trying to set environment variables so I can define my own path for my
> modules. My problem is I don't know where to set the PYTHONPATH. My book
> says do it in the .cshrc file: where the heck's that? My Linux book says
it
> should be in the /home directory, but it isn't. I own this server and it's
a
> standard RedHat 6.2 install. Please advise.

What is your shell? It is probably bash. if so, add something like this to
/home/username/.bashrc:

export PYTHONPATH=/path/to/modules

If adding more than one path put a colon between them.


David

--
http://www.python.org/mailman/listinfo/python-list




From donn at oz.net  Fri Dec 22 00:47:41 2000
From: donn at oz.net (Donn Cave)
Date: 22 Dec 2000 05:47:41 GMT
Subject: 2 question from a python newbie!
References: <91tl43$rko$1@neptunium.btinternet.com> <slrn944q0t.919.scarblac@pino.selwerd.nl> <91tq58$cq6$1@uranium.btinternet.com> <91tvmk07qs@news2.newsguy.com>
Message-ID: <91uptt$5ru$0@216.39.151.169>

Quoth "Alex Martelli" <aleaxit at yahoo.com>:
| "A scripting language and nothing more" suggests a language
| that only lives for limited customization/automation purposes "on
| board" of a 'real' application; although this may be an outdated
| view of higher-level (aka 'scripting':-) languages, the 'nothing
| more' strongly suggests that this is the speaker's view.

At least, it strongly suggests that the speaker understimates
the ambiguity of this unfortunate usage.

	Donn Cave, donn at oz.net


From aleaxit at yahoo.com  Sat Dec  9 08:09:29 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 9 Dec 2000 14:09:29 +0100
Subject: using execfile() in class/instance methods
References: <slrn932qna.rtk.wcuddy@lserv.ja10629.home>
Message-ID: <90tb1302s27@news2.newsguy.com>

"Wayne Cuddy" <wcuddy at lserv.ja10629.home> wrote in message
news:slrn932qna.rtk.wcuddy at lserv.ja10629.home...
> I am attempting to use the execfile() function to "source" a file that
> has valid python code to create a diction object:
>
> bn { 'key' : 'value1',
>         'key2' : 'value2' }

This is not valid Python code -- it's probably missing an equal-sign
between the identified bn and the open-brace.


> I am issueing this command from a class constructor method and would
> like bn to be bound the class instance.  So in the future I can say:
>         print self.bn['key']

Piece of cake.

-- file: ba.txt
bn = { 'key' : 'value1',
     'key2' : 'value2' }
-- end of file: ba.txt

-- file: ba.py
class Ba:
    execfile('ba.txt')
    def bu(self):
        print self.bn['key']

ba = Ba()

ba.bu()
-- end of file: ba.py

D:\PySym>python ba.py
value1

D:\PySym>


> Is this possible?  I assume I am screwing up the global() and locals()
> arguments to the function which I can say that I don't fully

I think it was probably just the missing '=' sign, as above.  The default
for globals and locals seem to be OK in this case.

> understand.  Also is there another better/worse way to do this???just
> so I can see what else is out there..

execfile is very risky if the textfile you're executing may contain
undesired
stuff, of course; then, you might want to build up the self.bn dictionary
in a safer manner, which might then be highly preferable.

But if the textfile is fully under your control, this might be quite OK.


Alex





From billtj at z.glue.umd.edu  Thu Dec 21 16:09:47 2000
From: billtj at z.glue.umd.edu (William Djaja Tjokroaminata)
Date: 21 Dec 2000 21:09:47 GMT
Subject: A Suggestion for Python Colon Syntax
References: <91tc8h$2pp$1@hecate.umd.edu> <W2r06.1076$AH6.86249@newsc.telia.net> <91tfg1$282$1@hecate.umd.edu> <EEr06.842$Qb7.73987@newsb.telia.net>
Message-ID: <91trir$52l$1@hecate.umd.edu>

I saw many replies that give the reasons why the colons are
*needed*.  My suggestion is to make it *optional*, like the semicolon at
the end of statements.

Yes, for people who cares about making it easier to read for newbies, they
can always put colons.  Yes, for people who use emacs to automatically
indent the next line, they can put colons too.

However, as Guido wrote that he wanted an ABC that appeals to Unix/C
users, one of the unix designs is its minimalist style in terms of command
or syntax.  To me, I like Python because when I want to create a function
I just write def as the first keyword and the rest follows, without much
punctuation.  However, from time to time, the colons always make me come
back to fix them.  For me personally, deleting the colons do not change
the readability of the code at all.  Probably the study should be
conducted not only on the newbies, but also on people who have programmed
for a long time.  Besides, fewer keystrokes are always better, in my
opinion.  Unix C shell does not require colons for the clauses.

Therefore, my suggestion is really not to remove the colons, but to make
them optional.  I want opinion of you who have programmed for a long time
whether there will be some harms if the colons are made optional.  If not,
probably someone can tell me how I can forward this suggestion to the core
Python development team?

Regards,

Bill


From billtj at y.glue.umd.edu  Fri Dec 22 15:38:08 2000
From: billtj at y.glue.umd.edu (William Djaja Tjokroaminata)
Date: 22 Dec 2000 20:38:08 GMT
Subject: A Suggestion for Python Colon Syntax
References: <mailman.977446446.16022.python-list@python.org> <9207a3$av6$1@hecate.umd.edu> <mailman.977512869.31157.python-list@python.org>
Message-ID: <920e3g$bt9$1@hecate.umd.edu>

D-Man <dsh8290 at rit.edu> wrote:
: On Fri, Dec 22, 2000 at 06:42:11PM +0000, William Djaja Tjokroaminata wrote:

.......

:> I am not too familiar with the parsing stuff.  However, in my simplistic
:> opinion, Python is not a free-form language like C or Perl.  Therefore,
:> probably it is reasonable for any parser to breaks a Python code first
:> into lines, even in backward parsing, instead of parsing it token by
:> token first.  In parsing backwards, can then it just detect first that the
:> line is at different indentation level rather than try to detect the
:> colon?

: If the parser (I believe you are referring to someone's emacs mode
: thing) used the indentation level to infer the structure, it would
: totally defeat the purpose.  The purpose of that parser is with the
: following input:

: if 1 :<enter>print "hello"

: To give the following result in the editor:

: if 1:
: 	print "hello"


: I use vim myself and rather like this feature in C/C++/Java modes.  It
: also handles the unindent when a } is typed.


: I for one rather like having the colon and I have more experience
: using C++, C, and Java than Python.

.......

Well, I think there is not much support in making the colon optional.  I
use emacs, but it does not have the python mode (yet).  But the beauty of
Python is, in C or C++ I don't think I want to program without a
color-coding editor, but with Python, even in simple text mode, I can
program easily.  I guess I just have to install the python mode in emacs,
and based on the automatic indentation caused by the colon, probably the
colon will help me in coding (by automating the indent) rather than keep
giving me syntax errors from time to time.

I will start another thread about another suggestion (regarding the
dictionary), and I guess it indicates how useful Python to me is.

Regards,

Bill


From marsh1963 at my-deja.com  Fri Dec 15 20:26:23 2000
From: marsh1963 at my-deja.com (marsh1963 at my-deja.com)
Date: Sat, 16 Dec 2000 01:26:23 GMT
Subject: Python newbie can't launch freeze; Winnt/python20
Message-ID: <91egbq$v6g$1@nnrp1.deja.com>

Hi Gurus,

I'm trying to create a simple wrapper in Python for some Expect
scripts. Don't want to have to install Python on everyones PC....
Sorry if this has already been covered, I didn't find a reference to it
in my searches. Hope someone can tell me where I took the wrong
exit.... (besides having Windows... not my choice) :)

I have Windows NT 4.0, Python 2.0, and the Installer_b_03f.

At 1st it looked like it couldn't find python15.dll (I have
python20.dll on my system) so I changed the references of python15.dll
in Builder.py and in \projects\distr_b3 to python20.dll . I still have
problems.

It looks to me like it's trying to find win32api.pyd, which I don't
have.

When I try to execute:
C:\Python20>python c:\python20\installer\freeze.py cliScript

I get a popup: pyhton.exe - Unable to locate DLL
The dynamic link library python15.dll could not be found in the
specified path
c:\python20;.;c:\winnt\system32;c:\winnt\system;c:\winnt;c:\winnt\system
32;c:\winnt (not sure where it got the dup entries)

The MSDOS window echos:
File 'cliScript.cfg' already exists. Overwrite? [yn]y
> "C:\Python20\python.exe" c:\python20\installer\Builder.py
cliScript.cfg
Cannot determine your Windows or System directories
Please add them to your PATH if .dlls are not found
The win32api module is not available - modules listed
in the registry will not be found.
Traceback (most recent call last):
  File "c:\python20\installer\Builder.py", line 9, in ?
    from MEInc.Dist import ltoc
  File "c:\Python20\installer\MEInc\Dist\ltoc.py", line 2, in ?
    from MEInc.Dist import finder, tocfilter, resource
  File "c:\Python20\installer\MEInc\Dist\resource.py", line 4, in ?
    import carchive
  File "c:\Python20\installer\MEInc\Dist\carchive.py", line 9, in ?
    import zlib
ImportError: DLL load failed: The specified module could not be found.

C:\>echo %PATH%
C:\WINNT\system32;C:\WINNT;"C:\Program
Files\Hummingbird\Connectivity\7.00\Accessories\"

Any help greatly appreciated,
Marshall



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


From jbranthoover at my-deja.com  Mon Dec 11 11:01:52 2000
From: jbranthoover at my-deja.com (jbranthoover at my-deja.com)
Date: Mon, 11 Dec 2000 16:01:52 GMT
Subject: A Newbie Question: How To Call Other Python Programs From A Main Python Program.
Message-ID: <912tpb$efl$1@nnrp1.deja.com>

Hello All,
	I have written several small Python programs.  These programs
are not just functions but complete programs with their own menuing
systems.

        My question is,  how do I (and is it a good idea) call these
independent Python programs from one main program that selects which
program to run from a main menu?  I tried to use the import command.
This seems to work OK except that the programs only run once.  If the
same program is selected from the menu,  nothing happens.  It seems to
return immediately without running the program.

        Any information that you can give me will be greatly
appreciated.  Thank you for your time.


Sent via Deja.com http://www.deja.com/
Before you buy.


From matt at virtualspectator.com  Mon Dec 18 13:42:02 2000
From: matt at virtualspectator.com (matt)
Date: Tue, 19 Dec 2000 07:42:02 +1300
Subject: ? kalman filters
In-Reply-To: <3A3E4C55.9C785306@san.rr.com>
References: <B43D149A9AB2D411971300B0D03D7E8B0FBCBB@natasha.auslabs.avaya.com> <91lclt02qr3@news1.newsguy.com> <3A3E4C55.9C785306@san.rr.com>
Message-ID: <00121907445100.00900@localhost.localdomain>

Has anyone written any wrappers for any kalman filter libraries?

.... else I'll go do it.

thanks
Matt  



From jochen at pc1.uni-duesseldorf.de  Thu Dec  7 05:02:45 2000
From: jochen at pc1.uni-duesseldorf.de (Jochen =?iso-8859-1?q?K=FCpper?=)
Date: 07 Dec 2000 11:02:45 +0100
Subject: Programming Habits in Python
References: <cspU5.82459$U46.2617014@news1.sttls1.wa.home.com> <8vvb6p$ct2$1@news.wrc.xerox.com> <XiMU5.87123$U46.2803855@news1.sttls1.wa.home.com> <900cpl02ss1@news2.newsguy.com> <Ue3V5.91231$U46.2956112@news1.sttls1.wa.home.com> <slrn92tv0u.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com>
Message-ID: <udn1e8k0qi.fsf@box.home.de>

>>>>> A.M. Kuchling wrote on 7 Dec 2000 02:42:19 GMT:

AMK> On Wed, 29 Nov 2000 08:37:08 GMT, Greg Jorgensen <gregj at pobox.com> wrote:
>> I'm not contending that all junior programmers are dummies, or even that
>> everyone should read Knuth, though it's hard for me to imagine how anyone
>> claiming to be a programmer can sleep without Fundamental Algorithms under
>> their pillow ;-). 

AMK> I sleep quite well, thank you. :) I've started reading Knuth's TAoCP
AMK> several times but never gotten very far, mostly because vol. 1 quickly
AMK> embroils you in MIX,

Me too...

AMK> making it very hard to distill anything of practical value from
AMK> the book.

... but I go back once in a while when I need some basic
understanding. Pick the right section and go through -- then you'll
understand where the other books come from:)

AMK> I liken TAoCP to Newton's _Principia_: monumental, quite
AMK> important, but not too useful except for specialists.  If a
AMK> bright but unstudied person asked me how to learn physics, 

Not to destroy the metaphor, but you cannot learn modern physics from
Principia! Classical Mechanics at most. But you can learn the
physicists way of thought... [1]

AMK> I'd certainly point them at a text such as Halliday & Resnick,
AMK> not the _Principia_!

Feynman's lectures do a pretty good job here, actually.


And this points at a big difference between Principia and TAOCP:
Physics has gone a really long way since Newton. On the other hand
Knuth, to my understanding, is still giving a elaborate description of
*todays* algorithms.

There are more physics classics then Principia, and comparing it to
TAOCP 'cause both are the first in their topic isn't the right
thing. I won't say you have to compare TAOCP to Feynman-lectures, at
least they are from the same time, though...


AMK> The algorithm textbook parallel to Halliday/Resnick, for me at
AMK> least, would be the Cormen/Leiserson/Rivest book; clearly
AMK> written, covers lots of topics, fairly rigorous, but not so
AMK> detailed as to be impossible to read.

The last sentence actually is a tradeoff everybody has to decide on by
himself; probably one has to decide on with every new problem:) I have
Sedgewick, great for a very quick peek on this stuff, Knuth for the
basics, but actually most of the time I look into Golub, van Loan:-))

Greetings,
Jochen

[1] quantum physics is celebrating it's hundredth anniversary this
month, relativity is getting close, and Heisenberg's/Schr?dinger's
quantum theories are at ~75:-)) ...
-- 
Einigkeit und Recht und Freiheit                http://www.Jochen-Kuepper.de
    Libert?, ?galit?, Fraternit?                GnuPG key: 44BCCD8E
        Sex, drugs and rock-n-roll


From gregj at pobox.com  Sat Dec 23 19:34:03 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Sun, 24 Dec 2000 00:34:03 GMT
Subject: Some Python 2.1 ideas
References: <mailman.977590932.19335.python-list@python.org> <Cw616.23054$qN6.3308205@typhoon.we.rr.com> <3A452881.7686E717@uswest.net>
Message-ID: <%vb16.187233$U46.5801214@news1.sttls1.wa.home.com>

"Bjorn Pettersen" <pbjorn at uswest.net> wrote in message
news:3A452881.7686E717 at uswest.net...
> string.chomp() (a better name would be nice) needs to handle all of:
>
>  - "foo\n"
>  - "foo \n"
>  - "foo \r\n"
>  - "foo \n\r"
>  - "foo"
>
> etc. correctly, which means it should delete between zero and two bytes
> (characters) from the end of the string. This is very different from
> removing leading or trailing whitespace (whitespace can be significant
> you know <wink>).

How should these be handled?

  "foo\n\n\n"
  "foo\n\r\n\r"

etc.

In other words, does the chomp() method always remove 0-2 characters, or all
trailing /n/r characters? Is it a good idea to add a string method that
makes assumptions about what the file.readline and file.readlines methods
return? Wouldn't it be cleaner to enhance the file object with additional
methods:

  f. readline_chomp()
  f.readlines_chomp()

Or perhaps add an attribute to the file object to indicate if newlines
should be stripped for text files?

  f.chomp = 1  # 1=strip trailing /n and /r characters, 0=leave them
(default)

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From jrwren at xmtp.net  Tue Dec 12 10:36:15 2000
From: jrwren at xmtp.net (Jay R. Wren)
Date: Tue, 12 Dec 2000 15:36:15 GMT
Subject: Python support for MIDL RPC
References: <3A354BEE.7D5C9895@ct.marconimed.com>
Message-ID: <PBrZ5.142665$hD4.36246501@news1.rdc1.mi.home.com>

Depending on what you are trying to accomplish, you may want to consider onc
rpc on windows.  There are a number of 3rd party onc rpc implementation for
win32 and some are even free.

I find it the best quick and dirty method of getting NT too communicate with
Unix and others at an rpc level.

To answer your question directly.  I know of no DCE RPC implementations for
Python.  Specifically I know of no DCE RPC implementations short of NT
itself or a UNIX with DCE running.

Further, NT's DCE RPC's are implenented below the COM API.  It is doubtful
you could even use win32com under python to call NT DCE RPC's directly.  If
you are only trying to get NT peers to communicate, you could use DCOM and
write a DCOM server in python with DCOM python clients communicating.  Use
win32com for this.

-Jay R. Wren

"tanya" <tanya at ct.marconimed.com> wrote in message
news:3A354BEE.7D5C9895 at ct.marconimed.com...
> I need to implement DCE RPC Server/Client in Python using TCP/IP to
> communicate with NT peers.
> What Python has to support it? Do I have to write the
> interface/marshaling stuff myself? (Hope not...???)
>
> Thans.
>




From grante at visi.com  Wed Dec 20 17:37:44 2000
From: grante at visi.com (Grant Edwards)
Date: Wed, 20 Dec 2000 22:37:44 GMT
Subject: capturing stdout from xwindows program
References: <mailman.977218382.5431.python-list@python.org>
Message-ID: <Ywa06.11261$TU6.1080052@ptah.visi.com>

In article <mailman.977218382.5431.python-list at python.org>, Roch'e Compaan wrote:

>How does one capture the result of a program that needs to
>attach to a running X Server?

It appears that you _did_ capture the program's output.

>If I execute a python script from a terminal while working in X
>it works fine, but in this case the system will execute the
>script and I need to tell it to connect to a X Server.

Are you asking how to tell the program what X server to use?
That is done by passing a DISPLAY env variable to the program.

>I tried popen2 but it returns 'program: cannot connect to X server'

Was the DISPLAY env varible set correctly?

-- 
Grant Edwards                   grante             Yow!  Is this BOISE??
                                  at               
                               visi.com            


From fredrik at effbot.org  Thu Dec  7 12:47:01 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Thu, 07 Dec 2000 17:47:01 GMT
Subject: operator precedence
References: <90ogt3$mkr$1@nnrp1.deja.com>
Message-ID: <p2QX5.2021$Kd1.296353@newsb.telia.net>

b_wiedermann at my-deja.com wrote:
> I am looking for a list of operator precedence that has been updated to
> include information for Python 2.0. Also, I am looking for association
> information. Should we assume that associativity is always from left to
> right, except for comparison operators?

as far as I can tell,

    http://www.python.org/doc/current/ref/summary.html

is up to date.

> What about assignment operators; they do not appear in the language
> reference at

there are no assignment operators in Python.  assignment is a
statement.

what looks like C's assignment operators are known as "augmented
assignment statements", and are documented here:

    http://www.python.org/doc/current/ref/augassign.html

</F>




From gregor at mediasupervision.de  Tue Dec  5 08:32:09 2000
From: gregor at mediasupervision.de (Gregor Hoffleit)
Date: Tue, 5 Dec 2000 14:32:09 +0100
Subject: controlling crashing Win9X apps ?
Message-ID: <20001205143209.B22835@mediasupervision.de>

Hi,

luckily enough, I don't have profound experiences with Windows, but now I'm
at the point where I have a Python script that walks down a directory tree
and for each file calls a Win32 command line executable via os.system().

That all works well, but for a few files in the tree, the executable will
crash and opens open up a modal panel ('close' or 'details') that blocks the
script until I close a window (to be exact, the Win32 exe subsequently calls
a few other exe's, and one of those crashes and blocks).

Does Python on Win32 allow for some kind of process control that allows me
to catch those errors and work around them ?

    Gregor
    



From moshez at zadka.site.co.il  Sun Dec 31 10:39:21 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Sun, 31 Dec 2000 17:39:21 +0200 (IST)
Subject: how do i check if a file exists with python?
In-Reply-To: <3a4c95dc.88792787@nntp.sprynet.com>
References: <3a4c95dc.88792787@nntp.sprynet.com>, <fBN26.99045$yR4.2640173@news1.rdc1.tx.home.com>, <RlM26.98766$yR4.2639444@news1.rdc1.tx.home.com> <92g5v8$r5k$1@nnrp1.deja.com> <mailman.978078157.26485.python-list@python.org>
Message-ID: <20001231153921.D0794A84F@darjeeling.zadka.site.co.il>

On Fri, 29 Dec 2000 14:02:25 GMT, ullrich at math.okstate.edu (David C. Ullrich) wrote:
> On Fri, 29 Dec 2000 16:54:56 +0200 (IST), Moshe Zadka
> >Now, it may be that I'm wrong, in which
> >case just go ahead an use os.path.isfile
   
> It's certainly true the the second foo is better than the first,
> in fact yes the first is evil. But there are other reasons a person 
> might want to know whether a file exists (although probably
> things like foo are the most likely explanation.)

You seemed to miss the above part of my posting. While not having
used isfile, I've frequently used isdir to my advatage before.

> I sometimes make cumulative backups of files containing
> important stuff. I sort of need to determine that file.001,
> file.002 and file.003 already exist and file.004 does not
> before deciding what to name the current backup file.

That's *really* evil. What if there is another version of your
process running???  Damn!, just lost that backup file. In that
case, I'd go for a Maildir-like algorithm, which promises no colisions.
Actually, in that case, usually what you want is CVS, but that's
a matter for another day.

> (When I tell people they're asking the wrong question I
> get in trouble for it...)

You're safe with us, buddy.

as-long-as-you-pay-your-protection-fee-in-time-ly y'rs, Z.
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From glenfant at equod.com.nospam  Mon Dec 18 12:12:02 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Mon, 18 Dec 2000 18:12:02 +0100
Subject: UN/EDIFACT related packages
Message-ID: <91lg8h$2mo$1@reader1.imaginet.fr>

Hi,

I'm looking for packages related to UN/EDIFACT EDI messages.
Yes we got business partners who did not make the XML jump.
I'm 100% newbie in the EDIFACT field, so if some of you have useful related
packages/tooks, thei're welcome.
I did not find throug usual Python related search engines :(

Thanks in advance

Gilles Lenfant




From eaj at ricochet.net  Wed Dec 27 00:46:56 2000
From: eaj at ricochet.net (Eric Jacobs)
Date: Wed, 27 Dec 2000 01:46:56 EDT
Subject: Homogeneous Sequences in Python?
In-Reply-To: <92bc2i$72n$1@nnrp1.deja.com>
References: <92bc2i$72n$1@nnrp1.deja.com>
Message-ID: <t4j3uamp6dve8c@corp.supernews.com>

cbarker at jps.net:

:
> The classic answer to why Python cannot be easily made faster, even if 
> it were compiled, is that because of Python's dynamic nature, an 
> expression like "x = a + b" cannot be compiled, because it is not known 
> until runtime what types a and b are, and what it means to __add__ them. 
> I am a BIG fan of Python's dynamic nature, so I would not like to
> see any of that flexibility go away.

The "holy grail" here would be to have automagic static type
inference, I think. This has been done for Scheme, for example,
without losing any dynamic flexibility.

> I have seen proposals for "optional static typing" that would address 
> this issue (I believe Guido is talking about it for Py3k). That would 
> certainly help, but it could be somewhat ugly as well. What occurs to me
> are two observations:

I hope that these proposals are not intended to solve the speed
problem!

> 1) the occasional staticly typed integer is not going to make much 
> difference. What would make a difference is a large collection of
> integers, for example

It could. Consider an integer that is incremented inside nested
loops. If it gets above 100, Python will end up allocating and
initializing new integer objects to increment it.

> 2) More often than not, when I build a large sequence, it is 
> homogeneous, that is a large list (or tuple, whatever) of the same type:
> maybe a list of strings ( file.readlines() ), a list of integers
> (range(n) ), or even a list of lists.
> 
> So, what we need to speed things up is to have a way for Python to 
> process large sequences of the same type without having to type check
> each item in the sequence.
>
> The simple way to implement this is to have an optionally static type 
> declaration for sequence: list.is_list_of_integers, or something like 
> that. Then when this list is processed, the interpreter would know what 
> it was dealing with.

But the interpreter already knows what it's dealing with. All it has
to do is to check its ob_type object! If the ob_type of a variable
is known to be constant at _compile_ time (say, if you do a for loop
over a statically typed list of integers), then it's possible to
move the vtable-style lookup and possibly some sanity checks out of
the loop (of course, no bytecode instructions exist to facilitate
this)...

> It then struck me that it could certainly be done 
> automatically, on-the-fly: Every time an item was added to a sequence, 
> it's type would be checked, and some kind of "homogeneous
> flag" would be set:

..but if you don't know anything about the type at compile time,
you have to generate code for the general case -- which uses the
ordinary bytecodes which do the method lookup and type checks as
usual. Of course, if you can't infer it absolutely (what if
someone goes to the interpreter prompt and makes a list of class
Foo), you can compile two versions of the function -- a generic
list, and a list-of-ints, say -- assuming you know what types it
will likely be called with.

What I think I hear you proposing is to have objects be able to
track their own type beyond the single layer that is indicated in
the ob_type field. Then if we know how to make optimized functions
for certain cases, we can dispatch based on this "rich type":

> a = [] # type unknown

__getitem__: raise IndexError


> a.append(4) # homogeneous list of integers

__getitem__: returns object of IntegerType or raise IndexError


> a.append(56) # still a homogeneous list of integers

__getitem__: returns object of IntegerType or raise IndexError


> a.append(4.5) # Not homogeneous

__getitem__: returns object of NumericProtocolType or 
raise IndexError

(I'm assuming the Py3K-ish type/class consolidation here -- none
of this is "real".)

If we have a compiled version of a function that is optimized
for the case where its argument's __getitem__ method returns a
specific type, we can dynamically dispatch to that version
instead of the generic one. From that point on, we can avoid all
dynamic dispatching.

> I have not idea how much this would slow append operations, but I 
> suspect much less than the potential speed up of know that a list is 
> homogeneous when it is. A plus of doing this automatically, is that it 
> would take NO changes to existing code to take advantage of the new
> feature.

I agree that we should look at optimizations that will work without
changes to existing code. Actually, here we may be able to work the
best of both worlds. Think of the extended type tracking as kind of
a "profiler" which collects statistical type information based on
what types are encountered in the real world. Then, this data can
be fed back into the compiler which optimizes for those static types.
Once the optimized versions are entered, the extended type codes
(actually, any type codes) are no longer needed in the wonderful
world of static types, which continues from that point onwards
(call-graphically speaking.)

The profiler could even run transparently as the program runs,
generating the fast VM code (or maybe even machine code) in real
time and keeping tracking of itself in a pickle for successive runs.
It could try to find that magic threshold where Pythoners would
naturally split up their program into Python (dynamic dispatch based
on rich type) and C (statically typed) modules. Now that would be
crazy.

The statistical, "real life" approach could possibly produce better
optimizations than a perfect static type inferencer could (like I
mentioned before, the existence of the interactive interpreter
nullifies any possible proof of what kind of arguments a function
must receive.)

> Would this really speed up Python code much, even for large homogeneous 
> sequences? I have no idea. I am in no way qualified to make that
> judgment, I'd like to hear what folks think.

It depends a lot on how you divide up the run-time and compile-time
layers, and how they're going to interact.

> I suspect that it won't speed up straight, run-through-the-interpreter 
> code enough to make it worth it, but it could have some real benefits in
> three places:
> 
> 1) some of the sequence oriented operations that are being introduced:
> list comprehensions, and the possible future Elementwise/Objectwise
> Operators (see PEP # 225)
> 
> 2) Py2c, the Python to C translator. I think a really important element 
> of Python's future may be Py2C (and perhaps eventually a Python compiler 
> or JIT compiler). It seems that has been the direction of number of 
> interpreted languages. With optional static typing and compilation, LISP 
> can be as fast (or faster) that C++ for numerics. MATLAB now has a 
> translator to C++ that can create faster code and fully compiled stand 
> alone applications. That has almost eliminated the need to write 
> external functions in C. Py2C can only be so fast without some kind of 
> static typing.

Or type inference. It surprises me to hear you say this, because you
seemed to be making a case for type inference with the list-of-integers
argument. Or maybe I don't understand what you're getting at?

> I think homogeneous sequences could do it. Right now Py2C 
> can convert "for i in range(1000)" into a fast C loop, but can't do 
> anything with "for i in list_of_integers", If it had a way of knowing 
> that it was a list of integers, it could be highly optimized.

Right, but Py2C would need to know it, not just Python. That's
the challenge ;-)

> 3) Extensions that you have to pass a lot of data to. An example is 
> wxPython. In order to draw a polyline with a couple of thousand points, 
> for example, you have to pass a list of 2-tuples of integers of the 
> coordinates of the points. The extension has to do a type check on every 
> one of the numbers in every one of the tuples in the list. Frankly, this 
> is still pretty fast, but not as fast as it could be.

External functions can, of course, provide known type signatures which
can greatly assist any type inference engine. Then again, perhaps it
is best not to rely too much on C function's demands: I often do a
PyNumber_Int, followed by a PyInteger_AsLong to accept any kind of
number object as input. It's kind of nice how Python does automatic
conversions like that.


> My goal is to be able to write applications that contain some 
> numerically intensive code (or other repetitious simple processing) and 
> not to have to use anything but Python. I'm going to have to keep 
> writing C for a large amount of the core code if something isn't done a 
> little differently.

This is, I think, the "Python approach" for problems like this:
use a C module for the time-intensive stuff, use Python for the
"high-level" control. Could this be done automatically? I think
so.

> What about Numeric?
> 
> NumPy provides the Multiarray object, which does provide what I am 
> talking about, a homogeneous sequence ( at least of numbers, which is 
> mostly what I need). As a result, it can implement very fast array 
> oriented operations that eliminate the need for a large amount of hand 
> coded C extensions. The problem is that core Python has no idea that it 
> is a homogeneous sequence, so when you have to do some looping, things
> slow WAY down.

The critical point here is, I think, who's getting that info: core Python
is only the first part of the battle. The code that Python is running
has to know too.

-- 



From torppa at polykoira.megabaud.fi  Tue Dec 19 21:19:38 2000
From: torppa at polykoira.megabaud.fi (Jarkko Torppa)
Date: 20 Dec 2000 02:19:38 GMT
Subject: How to get the os.system data into a variable
References: <91o5ih$2kv$1@news.wrc.xerox.com>
Message-ID: <91p4vq$94b$1@news.kolumbus.fi>

In article <91o5ih$2kv$1 at news.wrc.xerox.com>, jicman wrote:
>I know that
>
>t = os.system('ls -l')
>
>return code and assigns it to t.  I would like to get the data, not the code, 
>that is resulted from the command into a variable.  (i.e.

if 'ls -l' really is the command that you want to get the data from
then this does pretty much the same thing.
# 1.5 way
l=map(lambda x:(x,os.stat(x)),os.listdir('.'))
# 2.0 way
l=[ (e,os.stat(e)) for e in os.listdir('.') ]

the list is totally unformatted and as such more
suited for further processing. Look at stat module
for hints of what to do with the values.

-- 
 Jarkko Torppa                torppa at staff.megabaud.fi
  Megabaud Internet-palvelut


From aleaxit at yahoo.com  Sun Dec 17 04:43:54 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 17 Dec 2000 10:43:54 +0100
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <r0pzoi11lhz.fsf@black.cs.washington.edu> <5jvZ5.3588$Kd1.418770@newsb.telia.net> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com> <kuu28971kp.fsf@lasipalatsi.fi> <3A380E4E.8537A0FB@san.rr.com> <91a1vr02b02@news2.newsguy.com> <Co9_5.25465$x6.13756579@news2.rdc2.tx.home.com> <91bbq102mtk@news1.newsguy.com> <rWd_5.26820$x6.14119694@news2.rdc2.tx.home.com> <91d5el0me7@news1.newsguy.com> <rOr_5.31120$x6.15241856@news2.rdc2.tx.home.com> <slrn93o7j4.tu0.quinn@dinar.ugcs.caltech.edu> <JnW_5.37708$x6.18119061@news2.rdc2.tx.home.com>
Message-ID: <91i2d405qq@news2.newsguy.com>

"Rainer Deyke" <root at rainerdeyke.com> wrote in message
news:JnW_5.37708$x6.18119061 at news2.rdc2.tx.home.com...
    [snip]
> > >class UserInt:
    [snip]
> > >  def __iadd__(self, other):
    [snip]
> > Just as a matter of style, would it be superior to add a __coerce__
method
> > that knows how to coerce ints, longs, and floats into UserInts, or make
an
    [snip]
> __coerce__ would probably have been superior, since it eliminates
duplicate
> code and more closely resembles the way real integers work in Python.

__coerce__ is what I had used in my generic Mutable class, previously
posted, together with a __getattr__ and a few others, to try and mimic
a variety of objects it could be holding by delegating stuff to whatever
it's holding in a way reasonably economical of code.

But it doesn't really work for __iadd__ and friends, does it?  Unless I'm
missing something, those have to be implemented one by one.  Maybe
__getattr__ tricks could help -- if name.startswith('__i'), one could use
name.replace('__i','__',1) to look a function up in library module
'operator', etc.  But that doesn't exactly feel very 'clean' to me.  (The
performance issues could be ameliorated by having Mutable instances
cache a directory of such methods once looked-up and built, but that
doesn't make the overall architecture any cleaner, it seems to me).


Alex





From grante at visi.com  Mon Dec  4 12:24:33 2000
From: grante at visi.com (Grant Edwards)
Date: 4 Dec 2000 11:24:33 -0600
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Dec  4)
Message-ID: <8C419A8040EAFDCB.1774103991763D9E.2F034C54E78FF82C@lp.airnews.net>

    Alex Martelli discusses the issues involved in using Python
    within Windows Script Control.
        http://deja.com/=dnc/getdoc.xp?AN=698664345
    Alex Martelli shows a clever way to overload a method based on
    number of arguments.
        http://deja.com/=dnc/getdoc.xp?AN=975947338
    
    Dale Strickland-Clark and Alex Martelli on using Python
    programs as drop targets under Win32.
        http://deja.com/=dnc/getdoc.xp?AN=700537932
        http://deja.com/=dnc/getdoc.xp?AN=700557153
    
    Andrew H. Chatham announces Python bindings for Ogg/Vorbis (MP3
    patent/licensing free replacement for MP3) -- and requests help
    with dlopen().
        http://deja.com/=dnc/getdoc.xp?AN=975946985
    
    Nick Mathewson shows how to implement a general overloading
    module for python.
        http://deja.com/=dnc/getdoc.xp?AN=700076223
    
    Gordon McMillan releases fixed Installer for Python 2.0
        http://deja.com/=dnc/getdoc.xp?AN=700179180
    
    Raymond Hettinger shows a module to provide dynamic inheritance
    and convenient composition.
        http://deja.com/=dnc/getdoc.xp?AN=700743048
    
    Johannes Nix releases Python wrappers for ALSA (Advanced Linux
    Sound Architecture) that enable real-time audio processing.
        http://deja.com/=dnc/getdoc.xp?AN=975947772


========================================================================

Everything you want is probably one or two clicks away in these pages:

      Python.org's Python Language Website is the traditional
      center of Pythonia
          http://www.python.org
      although PythonLabs.com bills itself as "The Python Source",
      which is becoming increasingly true in 2000
          http://www.pythonlabs.com/

      PythonWare complements the digest you're reading with the
      daily python url
           http://www.pythonware.com/daily  

      comp.lang.python.announce announces new Python software.  Be
      sure to scan this newly-revitalized newsgroup at least weekly.
          http://deja.com/group/comp.lang.python.announce

      Andrew Kuchling writes marvelous summaries twice a month of
      the action on the python-dev mailing list, where the future
      of Python is truly determined
	  http://www.amk.ca/python/dev

      The Vaults of Parnassus ambitiously collects Python resources
          http://www.vex.net/~x/parnassus/   

      Much of Python's real work takes place on Special-Interest Group
      mailing lists
          http://www.python.org/sigs/

      Python Consortium emerges as an independent nexus of activity
          http://www.python.org/consortium

      Cetus does much of the same
          http://www.cetus-links.de/oo_python.html

      Python FAQTS
          http://python.faqts.com/

      Python To-Do List anticipates some of Python's future direction
          http://www.python.org/cgi-bin/todo.py

      Python Journal is at work on its second issue
          http://www.pythonjournal.com

      Links2Go is a new semi-automated link collection; it's impressive
      what AI can generate
          http://www.links2go.com/search?search=python

      Archive probing trick of the trade:
          http://www.dejanews.com/dnquery.xp?QRY=&DBS=2&ST=PS&defaultOp=AND&LNG=ALL&format=threaded&showsort=date&maxhits=100&groups=comp.lang.python


Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://purl.org/thecliff/python/url.html
or
  http://www.dejanews.com/dnquery.xp?QRY=~g%20comp.lang.python%20Python-URL%21


Suggestions/corrections for next week's posting are always welcome.
[http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
e-mail us ideas directly.]

To receive a new issue of this posting in e-mail each Monday morning,
ask <claird at neosoft.com> to subscribe.  Mention "Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.
-- 

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


From dvoitenko at qode.com  Fri Dec 29 10:26:31 2000
From: dvoitenko at qode.com (Voitenko, Denis)
Date: Fri, 29 Dec 2000 10:26:31 -0500
Subject: HTML Parser
Message-ID: <0FF17853157CD31197E100902784B34B7A1D06@FILESERVER1>

Hello,

I am trying to write an HTML parser. I am starting off with a simple one
like so:

# html_parser.py
import re
import string

newline=re.compile('\n')
HTMLtags=re.compile('<.*>')

file=open('C:\Documents and Settings\dvoitenko\My
Documents\Python\index.jsp', 'r')
input_file = file.read()
file.close()

jsp_content = newline.split(input_file)


# loop thru lines ...
for line in jsp_content[:]: 
	result=HTMLtags.search(line)
	tag_content = line[result.start()+1:result.end()-1]
	print '<'+string.upper(tag_content)+'>'

Which simply uppercases all html tags. Well, it actually uppercases
everything else which I do not like. What seems to be wrong here? Say if I
have a link <a href=hello.jsp>Hello</a> it will result into <A
HREF=HELLO.JSP>HELLO</A>... which is not right.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20001229/6aeaf02e/attachment.html>

From phil at geog.ubc.ca  Tue Dec 26 14:59:57 2000
From: phil at geog.ubc.ca (Phil Austin)
Date: 26 Dec 2000 11:59:57 -0800
Subject: Season's greetings (was Re: Some Python 2.1 ideas)
References: <mailman.977765348.13376.python-list@python.org> <928otj02l39@news1.newsguy.com>
Message-ID: <m34rzrx8aq.fsf@curlew.geog.ubc.ca>

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

> 
> One thing I wasn't able to easily answer -- what's the
> *easiest* way for a Python beginner to do function-plots,
> with nicely labeled graphs, minor/major ticks, color,
> auto-scaled, &c &c?  On-screen display the major need,
> though nice printouts wouldn't hurt either.  Speed no
> issue, as his new PC is twice as fast as mine!-)
> 
> I tried downloading DISLIN but it appears to embed its
> own Python, need its own setting of PYTHONPATH; and
> (from a superficial look at it) it doesn't seem to do auto
> scaling.  Any suggestions?  I'd like to keep amazing him
> with Python power and ease, as what dad wouldn't like
> for his late-teenage son...:-).
> 
> 
> Alex

I've just finished using dislin in a remote sensing/radiative
transfer course I taught to about 25 third  year students with
little computing experience.  I'd take another  look at it;
it's a standard python extension (dislin.dll), and although
the help documents seem to require installation in c:/dislin,
I think the dll can go anywhere in your pythonpath.  The source
of the extension module is available with  the linux installation,
if you want to doublecheck.

For some examples  of  dislin in action, you can check out
some of my problem sets -- in particular see mplotII.py
at:

(http://www.geog.ubc.ca/~phil/courses/a301/ps4.html)

Not ready for release yet, but it might give you a general
idea about dislin's capabilities.  There's also 
testimage.py, if he'd like to look at 7 channel Landsat
Terrestrial Mapper imagery of Morro Bay, California (see
the accompanying png files, produced with imageplot.py).

Regards, Phil

-- 
Voice: (604) 822-2175
Fax:   (604) 822-6150
email: paustin at eos.ubc.ca

Associate Professor and Chair
Atmospheric Sciences Programme
Department of Earth and Ocean Sciences
The University of British Columbia
6339 Stores Road
Vancouver, BC  V6T 1Z4


From manyong at gnu.org  Sun Dec 10 19:51:19 2000
From: manyong at gnu.org (Man-Yong Lee)
Date: Mon, 11 Dec 2000 00:51:19 GMT
Subject: pattern matching
References: <DZQY5.7555$KY1.20484@news1.rivrw1.nsw.optushome.com.au>
Message-ID: <3A342524.B3856100@gnu.org>

How about glob or re module?

Especially ``re'' module is worth learning, I think.

deadmeat wrote:
> Is there another standard library I can use for pattern
> matching, other than
> fnmatch? I need to be able to check for metacharacters, 
> fnmatch doesn't allow me to escape them...

-- 
Happy Python!
             Man-Yong (Bryan) Lee  <yong at linuxkorea.co.kr>


From jblaine at kickflop.ne.mediaone.net  Mon Dec 18 20:27:44 2000
From: jblaine at kickflop.ne.mediaone.net (Jeff Blaine)
Date: Tue, 19 Dec 2000 01:27:44 GMT
Subject: Tool name (posted here, can't recall), documentation generation
References: <slrn93r69b.ggs.jblaine@kickflop.ne.mediaone.net> <91ks40$e30$1@nnrp1.deja.com>
Message-ID: <slrn93teju.gjd.jblaine@kickflop.ne.mediaone.net>

>DocBook?

No, the tool was written in Python.

Thanks for trying, though.

Anyone else?


From paul.magwene at yale.edu  Thu Dec 14 10:43:30 2000
From: paul.magwene at yale.edu (Paul Magwene)
Date: Thu, 14 Dec 2000 10:43:30 -0500
Subject: Dijkstra's Shortest Path algorithm
References: <Pine.GSO.4.21.0012090042370.15110-100000@y.glue.umd.edu> <dvrd3t40u9l7atgdro59m5ttg7dsfv7c4i@4ax.com> <Pine.GSO.4.21.0012122249480.19682-100000@y.glue.umd.edu> <3A37055C.D831CACF@yale.edu> <Pine.GSO.4.21.0012131307060.761-100000@y.glue.umd.edu>
Message-ID: <3A38EAA2.10C9EB04@yale.edu>

Roy Katz wrote:
> 
> On Wed, 13 Dec 2000, Paul Magwene wrote:
> > If you're interested in graphs and programming, how about a clique
> > finding algorithm for undirected graphs?
> 
> Alright, I checked out some background material on cliques, and I'll take
> a look at Andrew Dalke's code (I looked at his site).  If I understand
> correctly, all that is needed is to generate a clique-finding extension in
> C?

That would work, and would be relatively straightforward to do given
that Andrew provides some C code.  More interesting, from my
perspective, would be to see a python solution using Numeric.  My guess
is that it would be nearly efficient as the C solution...


> This raises another question for me:  Suppose my exension (1) relies on
> another extension (2); is there a more efficient way for (1) to access
> (2) without going through the Python layer?
> 
> Roey


If both extensions are shared libraries, written in C, I imagine that
with the correct headers it should be easy to do so.

-- 

Paul Magwene
---------------------
paul.magwene at yale.edu
Postdoctoral Associate
Department of Ecology & Evolutionary Biology
Yale Univeristy


From aleaxit at yahoo.com  Sun Dec 17 12:46:55 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 17 Dec 2000 18:46:55 +0100
Subject: Singleton vs Proxies DP (was Re: Solution: Direct access to Printer  I/O lines)
References: <917k9b$9i9$1@nnrp1.deja.com> <slrn93hrjn.2vp.ssthapa@localhost.localdomain> <91cnp3$g59$1@nnrp1.deja.com> <91d23f0jk7@news1.newsguy.com> <91gfnp$7r4$1@animus.fel.iae.nl> <91gkh801e49@news2.newsguy.com> <3A3CB83D.45C8775E@darwin.in-berlin.de>
Message-ID: <91iu95013ta@news2.newsguy.com>

"Dinu C. Gherman" <gherman at darwin.in-berlin.de> wrote in message
news:3A3CB83D.45C8775E at darwin.in-berlin.de...
    [snip]
> The GOF book defines a Singleton rather clearly as a
> *unique instance* of a class (while it does not exclude
> that a limited number > 1 of instances might also be
> useful, sometimes). IMHO, a unique identity is part of
> the pattern as you will often have the case where the

I agree that unique id is part of the Singleton Pattern:
that is what makes it problematic.  One does not really
_care_ about that identity, only about the state and
behavior it represents.

> *usage* of a Singleton, i.e. the time windows in which
> it is used by multiple clients do overlap. In fact, I

Yes, that often happens.

> think this is the regular case for this pattern, while
> you seem to argue in favour of use-cases where clients
> immediately drop their handle to that one or *same*
> instance and there is no overlap.

No!  I'm not saying two clients can't hold their
'handles' (lw proxies) at the same time.  It IS
pretty typical.  But it's not at all typical for client
code to be interested in ensuring _identity_ --
state and behavior are typical.


> Of course, you're free to implement the pattern in what-
> ever way you like, but if, as you seem to do, define your
> Singleton pattern as multiple instances sharing behaviour
> *and* state, i.e. while allowing for different identities
> as well (in the general case) I'm tempted to suggest, may-

Right, it gets the gist, except that I don't see where
you get the non-overlap idea you mentioned above.

> be, a different name, like Clone, Replicant or something
> similar. A Proxy, at least to me, always means a place-
> holder for some *other* object, like "intelligent poin-
> ters" or the like. Well, if you add a Decorator to your
> pool of Clones you might get a Singleton again, but then,
> why would I do that? Doesn't a large pool of Clones mean
> a waste of memory?

I know 'Proxy' is the name of other patterns -0- it's just
that I dunno where I placed the C++ Report issue where
the pattern I prefer was first presented, to check up how
the author named it.

A small object takes a tiny amount of memory.  In C++,
a typical 'lightweightclone' would take 8 bytes -- 4 for
the virtual-table pointer, 4 for the data-pointer.  I think
in Python it would be 16 or so.  How many of those *at
the same time* will you have active?  A few hundreds
on a typical PC of today, at most?  I wish my worries
were all as bad as this one!-)


Alex


>
> Anyway, it probably all depends on the the problem you
> want to solve and the forces being at work, as well as
> your personal preferences...
>
> Glad-to-see-more-pattern-threads-again'ly,
>
> Dinu
>
> --
> Dinu C. Gherman
> ................................................................
> "The only possible values [for quality] are 'excellent' and 'in-
> sanely excellent', depending on whether lives are at stake or
> not. Otherwise you don't enjoy your work, you don't work well,
> and the project goes down the drain."
>                     (Kent Beck, "Extreme Programming Explained")




From aleaxit at yahoo.com  Tue Dec 19 05:10:30 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 19 Dec 2000 11:10:30 +0100
Subject: while true: !!!
References: <mailman.977180341.10206.python-list@python.org> <slrn93t7bv.le.grey@teleute.rpglink.com> <KTz%5.48644$x6.22165257@news2.rdc2.tx.home.com>
Message-ID: <91nc7h0s0p@news2.newsguy.com>

"Rainer Deyke" <root at rainerdeyke.com> wrote in message
news:KTz%5.48644$x6.22165257 at news2.rdc2.tx.home.com...
> "Steve Lamb" <grey at despair.rpglink.com> wrote in message
> news:slrn93t7bv.le.grey at teleute.rpglink.com...
> > On Mon, 18 Dec 2000 23:32:23 +0100 (MET), Carsten Geckeler
> <nospam at no.spam>
> > wrote:
> > >`while true' would be more readable and clear than `while 1'.  By the
> way,
> > >explain a newbie that -1 is true. ;)
> >
> >     0 is true, all else is false.  And?  :P
>
> 0 is false, None is false, [] is false, () is false, {} is false, and x is
> false where x is a class instance and x.__nonzero__() returns false.
> Everything else is true.  I think.

Another couple of 'false's: a class instance which doesn't define
__nonzero__ but does define __length__ when the latter returns 0;
and, C-implemented user objects, similarly to class instances.

I don't _think_ there are any more...


Alex





From paul.moore at uk.origin-it.com  Thu Dec 14 08:36:20 2000
From: paul.moore at uk.origin-it.com (Paul Moore)
Date: Thu, 14 Dec 2000 14:36:20 +0100
Subject: How to use mxTextTools
Message-ID: <2bc4OuYTMVp87e49nlcz01wv0ybq@4ax.com>

Hi,
I'm looking at mxTextTools to see if it would be suitable for some
types of text parsing work I am interested in (nothing concrete yet,
so I can't give specifics...)

The example in the documentation of tagging HTML looks fine - I
understand what's going on there, and as I understand it, this will
give me back a taglist, which is (effectively) the text stream with
portions tagged as I ask.

What I dont't see (yet), and I can't find any good examples for, is
what to do with the resulting taglist. There seem to be no functions
for working with taglists, and the lists themselves seem like
relatively complex data structures, so is it right that I should be
manipulating them "by hand"?

More information, or better still, some complete examples, would be
very helpful. (All the examples in the distribution just use
print_tags() to display the tags, and don't do anything with them...)

Paul



From sdm7g at virginia.edu  Tue Dec 19 12:44:46 2000
From: sdm7g at virginia.edu (Steven D. Majewski)
Date: Tue, 19 Dec 2000 12:44:46 -0500 (EST)
Subject: while true: !!!
In-Reply-To: <Pine.LNX.4.21.0012191723380.30860-100000@justitia.jura.uni-tuebingen.de>
Message-ID: <Pine.NXT.4.21.0012191242010.229-100000@localhost.virginia.edu>



On Tue, 19 Dec 2000, Carsten Geckeler wrote:

> > > 0 is false, None is false, [] is false, () is false, {} is false, and x is
> > > false where x is a class instance and x.__nonzero__() returns false.
> > > Everything else is true.  I think.
> > 
> > Another couple of 'false's: a class instance which doesn't define
> > __nonzero__ but does define __length__ when the latter returns 0;
> > and, C-implemented user objects, similarly to class instances.
> > 
> > I don't _think_ there are any more...
> 
> Nice; It takes three tries of -- I assume -- experienced Python
> programmers to find all cases of all statements evaluating as false. ;)
> 

No: There's an infinite number of statements that evaluate to false in
Python (and in all programming languages for that matter!) so don't
wait for someone to come along and iterate them all for you. I don't
think they were trying to find them all -- they were trying to classify
them all. 

-- Steve Majewski <sdm7g at Virginia.EDU> 





From aahz at panix.com  Tue Dec 12 19:20:59 2000
From: aahz at panix.com (Aahz Maruch)
Date: 12 Dec 2000 16:20:59 -0800
Subject: Spinning off time intensive procedures to speed CGI
References: <916bpn$a6i$1@nnrp1.deja.com>
Message-ID: <916fdb$gc$1@panix2.panix.com>

In article <916bpn$a6i$1 at nnrp1.deja.com>,  <leejor at my-deja.com> wrote:
>
>I have a variety of time consuming proceedures that slow the response
>time of my CGI scripts. These are proceedures that do not send
>any information back to the web browser. I want to spin off these
>proceedures to allow the original program to imediately continue.
>Examples of these proceedures include: saving to Log files, sending
>emails, or caching database queries.

You might try using threads.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

What if there were no rhetorical questions?  --Aahz


From nospam at no.spam  Mon Dec 11 10:05:26 2000
From: nospam at no.spam (Carsten Geckeler)
Date: Mon, 11 Dec 2000 16:05:26 +0100 (MET)
Subject: global __dict__
Message-ID: <Pine.LNX.4.10.10012111601001.487-100000@erde.kosmos.all>

Hi.

I was wondering if there is a __dict__ dictionary for the global namespace
as for classes and modules.  I was trying the following

>>> var = 1
>>> __dict__

and expected that __dict__ would contain something like {..., 'var': 1},
but there is no global __dict__.  Where can i find the global dictionary?

Cheers, Carsten
-- 
Carsten Geckeler:  carsten dot geckeler at gmx dot de
To get proper email-address replace `dot' and `at' by the corresponding symbols.




From kacmaz at my-deja.com  Tue Dec  5 23:38:43 2000
From: kacmaz at my-deja.com (kacmaz at my-deja.com)
Date: Wed, 06 Dec 2000 04:38:43 GMT
Subject: Using Dictionary look-up
Message-ID: <90kfsj$edl$1@nnrp1.deja.com>

This is kind of a follow-up to an earlier question I posted. Thankyou
for all who answered it....

I have written a simple dictionary look up where necessary values are
received from a single-item dictionary (later on I will open this up to
a larger number of items..!) (I saw a similar one in Mr Guido's site I
think!).....

The dictionary would be {'UserXXXX':'123123'}

----Configuration.py (start)----
...
UserName=UserXXXX
UserNumb="123123"
...
----Configuration.py (ends)----

----Main.py (start)----
...
from configuration import *
class Authenticator:
    def __init__(self):
        self.authenticate=Dict_lookup(UserName=function(UserNumb))

    def lookup_key (self,*args):
        return apply(self.authenticate.lookup_key,args)

class Dict_lookup:
    def __init__(self,**kw):
        self.dict=kw

    def lookup_key (self,sn):
        return self.dict[sn]
...
----Main.py (ends)----

But python gives me error when I run it like this as :
           Exceptions.KeyError : UserName

but still if I change the above line with :
    def __init__(self):
        self.authenticate=Dict_lookup(UserXXXX=function(UserNumb))

then it works fine...!!!

The general structure of the dict_lookup suits what I want to do but
why can't I call up UserName in this fashion?

thanks

HK


Sent via Deja.com http://www.deja.com/
Before you buy.


From timd at macquarie.com.au  Thu Dec 28 18:56:47 2000
From: timd at macquarie.com.au (Timothy Docker)
Date: 29 Dec 2000 10:56:47 +1100
Subject: unix pipes + python
Message-ID: <m4iu27odrr4.fsf@macquarie.com.au>


I must be doing something silly here...

| import os
| 
| def child( infd  ) :
|     "Run a child process"
|     pid = os.fork()
|     if pid == 0:
|         if infd != None: os.dup2( infd, 0 )
|         os.execv( '/usr/bin/cat', ['cat']  )
|     else:
| 	return pid
| 
| toPipe,fromPipe = os.pipe()
| child( fromPipe )
| os.write( toPipe, "hello\nthere\n" )
| os.close( toPipe )


This script behavious almost as expected. I end up with the text
echoed back, but the 'cat' subprocess is left running - I was
expecting it to exit when the write side of the pipe was closed in the 
parent, but it is still reading from fd 0.

Any suggestions on what's wrong?

Thanks,

Tim


From mmiller3 at iupui.edu  Mon Dec 11 21:37:37 2000
From: mmiller3 at iupui.edu (Michael A. Miller)
Date: 11 Dec 2000 21:37:37 -0500
Subject: Python-like compiled language
References: <913jr3$dmv$1@sylvester.vcn.bc.ca>
Message-ID: <87y9xm1i1a.fsf@lumen.med.iupui.edu>

>>>>> "Alexander" == Alexander Lawhead <alawhead at vcn.bc.ca> writes:

    > I'm trying to build some command-line driven programs to do
    > some image processing and am trying to find a Python-like
    > compiled language that would be faster than Python+NumPy.

Have you looked into the python imaging lib (PIL)?  It's no more
(or less) compiled than NumPy, but it will allow you to do
reasonably quick image processing.

Mike


From SBrunning at trisystems.co.uk  Mon Dec 18 09:12:41 2000
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Mon, 18 Dec 2000 14:12:41 -0000
Subject: search for pattern in StringList
Message-ID: <31575A892FF6D1118F5800600846864D5B1614@intrepid>

> From:	Daniel.Kinnaer at AdValvas.be [SMTP:Daniel.Kinnaer at AdValvas.be]
> Sent:	Monday, December 18, 2000 2:07 PM
> To:	python-list at python.org
> Subject:	search for pattern in StringList
> #try to show every item in MyList which contains an 'a'
> #that is 'an' and 'call'.   In Pascal :
> #for i:=0 to MyList.Items.Count -1 do
> #    if pos('a',MyList.Items[i])>0 then writeln(MyList.Items[i]) 
> #in Python : ???
 
myList = ['an', 'been', 'call', 'door']
for i in range(0, len(myList)):
    if myList[1].find('a') <> -1:
        print i, myList[i]





-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From nospam.newton at gmx.li  Fri Dec  8 03:46:36 2000
From: nospam.newton at gmx.li (Philip 'Yes, that's my address' Newton)
Date: Fri, 08 Dec 2000 09:46:36 +0100
Subject: Programming Habits in Python
References: <cspU5.82459$U46.2617014@news1.sttls1.wa.home.com> <8vvb6p$ct2$1@news.wrc.xerox.com> <XiMU5.87123$U46.2803855@news1.sttls1.wa.home.com> <900cpl02ss1@news2.newsguy.com> <Ue3V5.91231$U46.2956112@news1.sttls1.wa.home.com> <slrn92tv0u.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com> <FSHX5.120474$U46.4048998@news1.sttls1.wa.home.com>
Message-ID: <ct413tkg6mh85o7f9tscdap15kr715p6qp@4ax.com>

On Thu, 07 Dec 2000 08:28:21 GMT, "Greg Jorgensen" <gregj at pobox.com> wrote:

> Knuth has since expressed some regrets about using
> such a low-level language, but at the time it seemed a good choice
> (programmers pretty much had to know how to read and write assembly back
> then), and the high-level languages then in vogue (PL/I, Fortran) have so
> many shortcomings that the example implementations would have suffered.
> 
> Regardless of the language Knuth had chosen to use for examples, the
> examples would be antiquated today.

He makes that point on his web pages somewhere, probably in the FAQ. What
language should he choose? Regardless of his choice, it'll be antiquated in a
few years' time. If had chosen PL/I or Fortran, it'd be old now. If he had
chosen C, some people might say "it should be C++ or Java". If he chooses Java
-- well, will that still be current in 10 years' time? Will the basic principles
and algorithms of computer programming still be around? Probably no to the first
question, yes to the second.

And he does have a new assembly-type programming language called MMIX now, which
is based on a RISC-type machine, if I remember correctly. You can download the
spec as a PostScript paper. AIUI, he'll change the examples in the first few
volumes from MIX to MMIX if he produces a new edition, and upcoming volumes will
use MMIX from the start.

Cheers,
Philip
-- 
Philip Newton <nospam.newton at gmx.li>
If you're not part of the solution, you're part of the precipitate.


From ten.llun at ttam  Sun Dec 17 18:51:45 2000
From: ten.llun at ttam (Matthew Schinckel)
Date: Mon, 18 Dec 2000 10:21:45 +1030
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <915jcf0c6o@news2.newsguy.com> <9162p3$7t2$1@animus.fel.iae.nl> <MBWZ5.334$h67.37748@newsread1.prod.itd.earthlink.net> <3A397C74.6C8906E1@cosc.canterbury.ac.nz> <91d5520m6j@news1.newsguy.com> <wvw_5.10652$h67.670217@newsread1.prod.itd.earthlink.net> <slrn93mh41.1nk.scarblac@pino.selwerd.nl>
Message-ID: <3a401df0$1_1@news.chariot.net.au>

In message <slrn93mh41.1nk.scarblac at pino.selwerd.nl>, Remco Gerlich wrote:
> Oldayz <sill at localhost.kitenet.net> wrote in comp.lang.python:
> > Hey, how about if you think while 1: looks ugly
> > you do true =3N 1 somewhere on top and use
> > while true: ? I think it's much more obvious
> > and clean and you don't need to change syntax.
> 
> Yes! And you don't need the ugly "break" statement, you can simply do
> "true =3N 0" and the loop will end! (Remember to set it back to 1 later)
> 
:-)

(I'm the guy who forgets to set it back :-)

In my freshman Pascal course, all those years ago (although, until last year
they still taught the same course), I was really pleased that I came up with
(may be bad syntax, haven't used Pascal since 93 :)

HELL_FREEZES_OVER =3N 0

repeat:
 # do stuff
until HELL_FREEZES_OVER


-- 
Matthew Schinckel
matt at null dot net



From MarkH at ActiveState.com  Tue Dec  5 00:16:28 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Tue, 05 Dec 2000 05:16:28 GMT
Subject: ActiveState: PythonWin not using PythonPath for import?
References: <906ki2$s16$1@nnrp1.deja.com> <3A283CDB.8020800@ActiveState.com> <90gjah$7bh$1@nnrp1.deja.com>
Message-ID: <3A2C78DF.6000603@ActiveState.com>

Bill Seitz wrote:

> Thanks, that worked (once I quit/relaunched PythonWin). Is this (a)
> recognized outside PythonWin, and (b) documented?

Yes, and somewhere, but almost certainly not as prominently as it should be.

Mark.



From bradh at mediaone.net  Fri Dec  8 09:50:11 2000
From: bradh at mediaone.net (Brad Howes)
Date: Fri, 08 Dec 2000 14:50:11 GMT
Subject: DCOracle Memory Leak?
References: <m3g0k1qih5.fsf@h0050047c4794.ne.mediaone.net> <t2v9fq7v9ns565@corp.supernews.com>
Message-ID: <m3aea7dl24.fsf@h0050047c4794.ne.mediaone.net>

"Jose Martin" <joseamartin at venmex.com> writes:

> Hi, sorry i cant help you, but , can you help me ?

Yeah. You missed some of the install instructions in the README file under the
src directory. After you build the *.so files there, you need to move them into
../DCOracle/ so they may be loaded by the Python files there.

BTW, I found that if I went to the latest version of DCOracle, my leaks went
away. All is well now.

Brad

-- 
    "Were people this stupid before TV?" -- _White Noise_ by Don DeLillo


From bernhard at abnoba.intevation.de  Thu Dec 28 04:36:13 2000
From: bernhard at abnoba.intevation.de (Bernhard Reiter)
Date: 28 Dec 2000 09:36:13 GMT
Subject: Source for Linux Binaries?
References: <92at88$rig$1@nnrp1.deja.com>
Message-ID: <92f1id$re4$3@newsserver.rrzn.uni-hannover.de>

In article <92at88$rig$1 at nnrp1.deja.com>,
	chris_barker at my-deja.com writes:

> I'm looking for a source for Linux binaries for Python 2.0, and a whole
> bunch of modules to go along with it (PIL, Numeric, wxPython, etc.) 

> Has anyone set up a comprehensive source for Linux binaries? My foirst
> choice is RedHat 6.1 compatable rpms, 

At least I don't know any.
You probably have tried ActiveState and BeOpen I guess...

Somebody would assume that the debian people do have everything
ready, though.
	Bernhard
-- 
Professional Service around Free Software                (intevation.net)  
The FreeGIS Project                                         (freegis.org)
Association for a Free Informational Infrastructure            (ffii.org)
FSF Europe                                            (www.fsfeurope.org)


From thomas at xs4all.net  Sat Dec 23 12:00:40 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Sat, 23 Dec 2000 18:00:40 +0100
Subject: Building Python with Linux/ccc on the Alpha Platform
In-Reply-To: <91vt3l$l97@news.Informatik.Uni-Oldenburg.DE>; from jnix@medi.physik.uni-oldenburg.de on Fri, Dec 22, 2000 at 04:48:04PM +0100
References: <91vt3l$l97@news.Informatik.Uni-Oldenburg.DE>
Message-ID: <20001223180040.C5515@xs4all.nl>

On Fri, Dec 22, 2000 at 04:48:04PM +0100, jnix at medi.physik.uni-oldenburg.de wrote:

> I am interested in building Python with Numeric extension using the
> Compaq CCC compiler on a Alpha21264 machine under Linux. I have a good
> reason for this: ccc binaries usually run three to five times faster
> than the corresponding gcc binaries, especially if they are
> floating-point intensive - at this point in time, gcc just not
> optimizes well on the Alpha.

> Anybody did this before?

> What do I have to observe?

Probably nothing, as long as you make *everything* use 'ccc' -- that means
'configure' as well. And don't forget to run 'make clean' or (pref'rably)
'make distclean' (but save your Setup file if you don't want to re-edit it)
before you run configure. Then call configure with the 'CC' environment
variable set to 'ccc' (or the full path, if you need it.)

(use 'env CC=ccc ./configure <--opts>' in a c-style shell, or
'CC=ccc ./configure <--opts>' in a sh-style shell.)

If you want to compile some parts using ccc, and others using gcc, it might
get a bit tricky, but that really depends on how different gcc and ccc are,
and I have no clue on that, sorry.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From alexander.loehr at zuken.de  Mon Dec  4 10:50:09 2000
From: alexander.loehr at zuken.de (alexander.loehr at zuken.de)
Date: Mon, 04 Dec 2000 15:50:09 GMT
Subject: Tkinter.CreateFileHandler Unix/Windows
Message-ID: <90gefd$310$1@nnrp1.deja.com>

Hello!

I've got a question concering the function _tkinter.createfilehandler.
In the Module _tkinter.c are different places, where it is said, that
the FileHandler-functions doesn't work:

/* Tcl_CreateFileHandler() changed several times; these macros deal
   with the  messiness.  In Tcl 8.0 and later, it is not available on
   Windows (and on Unix, only because Jack added it back); when
   available on Windows, it only applies to sockets. */

I've got an application which uses threads with one thread running a
python-interpreter with a tk-mainloop. I have to send a notification to
the python-thread and want to use a socket for this. It seems okay on
Unix, but on Windows? I need a clear answer.

Pleae reply by email.

Thank's to everyone very much.
Alex.


Sent via Deja.com http://www.deja.com/
Before you buy.


From aleaxit at yahoo.com  Sun Dec 10 17:11:25 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 10 Dec 2000 23:11:25 +0100
Subject: _oleobj_ and dir
References: <910lpp$p2g$1@nnrp1.deja.com>
Message-ID: <910v3j01os2@news2.newsguy.com>

<yelled at yahoo.com> wrote in message news:910lpp$p2g$1 at nnrp1.deja.com...
> Hi,
>
>  I am trying to get available functions and attributes of COM obj:

Two steps: first, you must ensure that the Python wrappers for
the typelibrary you're using are on -- the simplest way may be
to use:
adoConn = win32com.client.gencache.EnsureDispatch('ADODb.connection')
rather than:
adoConn = win32com.client.Dispatch('ADODb.connection')

Once that is done, dir(adoConn.__class__) will give you, among
other things, the COM methods of the class:

>>> print dir(adoConn.__class__)
['BeginTrans', 'CLSID', 'Cancel', 'Close', 'CommitTrans', 'Execute', 'Open',
'Op
enSchema', 'RollbackTrans', '__call__', '__doc__', '__int__', '__module__',
'__s
tr__', '_prop_map_get_', '_prop_map_put_']

CLSID is the CLSID, and the names starting with '__' are Python-special
ones.  But _prop_map_get_ and _prop_map_put_ are of direct
interest to your purpose -- they're dictionaries (for all properties,
and R/W properties only, respectively) whose keys are the various
property names (never mind the associated values).


Alex






From mats at laplaza.org  Tue Dec 19 08:52:27 2000
From: mats at laplaza.org (Mats Wichmann)
Date: Tue, 19 Dec 2000 13:52:27 GMT
Subject: Source Code Management
References: <fb4f3tsgffnmvme64aol1cn397te7ln349@4ax.com> <91aq96$tba$1@nnrp1.deja.com> <698_5.3930$Kd1.455411@newsb.telia.net>
Message-ID: <3a3d1470.5243319@news.laplaza.org>

On Thu, 14 Dec 2000 18:17:38 GMT, "Fredrik Lundh" <fredrik at effbot.org>
wrote:

>gtalvola at NameConnector.com wrote:
>> If you don't mind paying money, I believe Perforce fits all your
>> requirements.  http://www.perforce.com/  We use it and have nothing
>> but good things to say about it.
>
>Couldn't agree more.
>
>(PythonWorks 1.1 can talk to Perforce, of course...)


Perforce, in fact, provided my incentive to learn Python.  I picked
Perforce for a particular client project because it was the easiest to
deploy and the small number of developers involved made it
cost-effective.  When we needed some scripts to extend functionality,
several of them were in Python... 

Perforce had, may still have, a deal that the product is free if you
declare that what you're developing is open-source.  You do have to
sign a paper to this effect, but it seems like a pretty good deal
nonetheless.


Mats Wichmann



From slhath at home.com  Tue Dec 26 09:36:11 2000
From: slhath at home.com (Scott Hathaway)
Date: Tue, 26 Dec 2000 14:36:11 GMT
Subject: xml-rpc server in python returning python lists...
References: <hih16.85504$yR4.2492277@news1.rdc1.tx.home.com> <925o9s$pqk$1@bunyip.cc.uq.edu.au> <HZx16.87713$yR4.2521840@news1.rdc1.tx.home.com> <928evf$u2$1@bunyip.cc.uq.edu.au>
Message-ID: <v1226.92343$yR4.2546834@news1.rdc1.tx.home.com>

Thanks, I will try the code snippet below.

Scott


"Chui Tey" <teyc at bigfoot.com> wrote in message
news:928evf$u2$1 at bunyip.cc.uq.edu.au...
> As far as I know, the python xmlrpc server conforms to the xmlrpc
standard.
>
> > cannot seem to get working.  Which one did you use with VB?
>
> I'm not sure. I downloaded both. I'm probably using the other one...
>
>     Set m_Client = New XMLRPCC.Client
>     Private m_DOM       As DOMDocument
>     Set m_DOM = m_Client.Execute(URL, "getServerName")
>     m_Client.ResponseToVariant m_DOM, sName
>
> I think mine needs VB6 + IE5 at least to work.
>
> theArray(0) and theArray(1) should return something..., try stepping
through
> the vb code
>
> (Sorry Pythonistas, probably should move over to a vb forum soon)
>
>
>




From ian.hobson at ntlworld.com  Tue Dec 12 15:28:24 2000
From: ian.hobson at ntlworld.com (Ian Hobson)
Date: Tue, 12 Dec 2000 20:28:24 +0000
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com>
 <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu>
 <976153091.257544@clam-ext> <3A30A181.4302F102@caltech.edu>
 <p6qr93ik7ks.fsf@informatik.hu-berlin.de> <90u5mo$kr@gap.cco.caltech.edu>
Message-ID: <K$zuiOAopoN6Ewbc@ntlworld.com>

In article <90u5mo$kr at gap.cco.caltech.edu>, Nathaniel Gray
<nospam at caltech.edu> writes
>
>It seems to me that flaming newbies only serves to make this a less 
>friendly newsgroup.

Come on. 

Its not fair to blame the flaming newbies.

<grin> 

Ian Hobson 

Every time we teach a child something, we prevent him from inventing 
it himself. - Jean Piaget


From amk at mira.erols.com  Sat Dec 30 11:14:30 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 30 Dec 2000 16:14:30 GMT
Subject: how to sort a dictionary by the values ?
References: <Pine.GSO.4.10.10012292313350.21787-100000@ux11.cso.uiuc.edu> <92kjor$ll7$1@news.nuri.net>
Message-ID: <slrn94s38i.gd.amk@207-172-111-106.s106.tnt1.ann.va.dialup.rcn.com>

On Sat, 30 Dec 2000 21:16:13 +0900, 
	June Kim <junaftnoon at nospamplzyahoo.com> wrote:
> >>> dic={ ... }
> >>> z=dic.items()
> >>> z.sort(lambda x,y:cmp(x[1],y[1]))

Note that this won't compare the keys at all, then; the resulting list
would be sorted by the value, and tuples with the same value would be
ordered randomly.  lambda x,y: cmp( (x[1], x[0]), (y[1], y[0]) ) would
sort by value and then key.  But look at all the tuple construction
and destruction going on; it would be even slower.  It might be
preferable, depending on the size of the dictionary and what you're
doing with the list 'z', to reverse all the tuples first and then sort
them.

>>> z = map(lambda x: (x[1], x[0]), z)
>>> z.sort()

--amk


From DanielK at aracnet.com  Mon Dec 25 13:45:47 2000
From: DanielK at aracnet.com (Daniel Klein)
Date: Mon, 25 Dec 2000 10:45:47 -0800
Subject: __init__ question
Message-ID: <EvM16.270$LU6.118037@typhoon.aracnet.com>

I'm trying to decipher some Python code and came across the line:

    def __init__(self, *args):

What is the meaning of the '*args' expression?

Thanks,
Dan:




From per at sbc.su.se  Fri Dec 15 10:36:13 2000
From: per at sbc.su.se (Per Kraulis)
Date: Fri, 15 Dec 2000 16:36:13 +0100
Subject: __repr__ policy problem: should module be prefixed to output?
Message-ID: <3A3A3A6D.2C3B1A91@sbc.su.se>

Having just released a Python extension module crng
(http://www.sbc.su.se/~per/crng/), I have received some constructive
criticism, one issue being save/restore functionality. My plan is to
define a __repr__ for each type in the module.

However, I have realized that there is a policy problem. Should __repr__
add the module name as prefix to the output string, or not? Or is there
some other solution?

Alternative 1:

>>> from crng import *
>>> r=ParkMiller()
>>> repr(r)
crng.ParkMiller(seed=314159265)

Alternative 2:

>>> from crng import *
>>> r=ParkMiller()
>>> repr(r)
ParkMiller(seed=314159265)

The problem is that Alt 1 is appropriate in the context of "import
crng", while Alt 2 is better when having done "from crng import *".

A suggestion is that there should be a settable module variable that
controls this. Feels like a kludge to me...

Any ideas which way to go?
-- 
Per J. Kraulis, Ph.D.                        per at sbc.su.se
Stockholm Bioinformatics Center (SBC)        http://www.sbc.su.se/~per
Dept. Biochemistry, Stockholm University     phone +46 (0)8 - 674 78 17
SE-106 91 Stockholm, SWEDEN                  fax   +46 (0)8 -  15 80 57


From jhg at galdon.com  Fri Dec 22 05:56:42 2000
From: jhg at galdon.com (Juan Huertas)
Date: Fri, 22 Dec 2000 11:56:42 +0100
Subject: About WebBrowser
Message-ID: <91vbr9$eo1q2@SGI3651ef0.iddeo.es>

It's possible to print the content of a Html page using the webbrowser
module into a python module?

For example:
Into Windows  (IE)  the webbrowser object has a method  (ExecWB) for
printing, it's possible to call this method inside a python module importing
Webbrowser?.

Using pythonwin this is possible previous register of IE.

Thank You in advance.

Juan H. Galdon

jhg at galdon.com





From skip at mojam.com  Wed Dec 13 22:12:21 2000
From: skip at mojam.com (Skip Montanaro)
Date: Wed, 13 Dec 2000 21:12:21 -0600 (CST)
Subject: null bytes in re pattern - difference between 1.5.2 and 2.0?
Message-ID: <14904.14997.494451.840804@beluga.mojam.com>

I want to delete control characters from some strings.  Accordingly, I
tried:

    name = re.sub("[\000-\037\177]", "", name)

This works in Python 2.0 but not in 1.5.2.  In 1.5.2 I find I need to use
raw strings:

    name = re.sub(r"[\000-\037\177]", "", name)

Accordingly, using raw strings in 2.0 fails.  Is there some form that will
work both in 1.5.2 and 2.0?

Is this a change I should have expected?  I assume it has something to do
with Unicode support in 2.0.

-- 
Skip Montanaro (skip at mojam.com)
Support the Mojam.com Affiliates Program: http://www.mojam.com/affl/
(847)971-7098



From roche at ybm.co.za  Tue Dec 19 04:35:25 2000
From: roche at ybm.co.za (Roch'e Compaan)
Date: Tue, 19 Dec 2000 11:35:25 +0200
Subject: capturing stdout from xwindows program
Message-ID: <002101c0699f$03db15a0$0102a8c0@roche.up-front.co.za>

How does one capture the result of a program that needs to attach to a
running X Server?

If I execute a python script from a terminal while working in X it works
fine, but in this case the system will execute the script and I need to tell
it to connect to a X Server.

I tried popen2 but it returns 'program: cannot connect to X server'

Roch?




From aleaxit at yahoo.com  Thu Dec 14 04:15:23 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 14 Dec 2000 10:15:23 +0100
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <3a36c85e.4409136@News.CIS.DFN.DE> <owen_nospam-1312001131590001@cs204-90.fsmodem.washington.edu> <3A37DD31.C9A5F18C@pacbell.net> <918rll02olg@news1.newsguy.com> <3A3800A3.D8E936D8@acm.org>
Message-ID: <91a33o02cce@news2.newsguy.com>

"Michael Khaw" <michael.khaw at acm.org> wrote in message
news:3A3800A3.D8E936D8 at acm.org...
    [snip]
> It's pretty hit-and-miss though. I see mom-and-pop stores in small towns
> all over Italy which prominently feature Disney characters in their
> logos/store names (Topolino=Mickey, Papagallo=Donald), yet Disney hasn't
> gotten around to sic-ing its lawyers on them.

'Topolino' (which Disney Italia uses as Mickey Mouse's name) just
means 'little mouse' in Italian, as 'Paperino' (Donald Duck) means
'little duck', 'Pippo' (Goofy) is a common abbreviation of Filippo
(just like 'Phil' would be in English), etc.  These choices hark
back to the '30s, by the way.  Hard to get intellectual-law
protection for such generic terms.

('Pappagallo' means 'parrot' and is not the Italian name of any
Disney character).

It's true that sometimes small shops &c do use somewhat-recognizable
_drawings_ of the copyrighted characters to go with the names -- THAT
could certainly induce Disney to sue if they learned about it.  But
even more incredible to me, I just saw the other day a new fast-food
shop downtown (falafel and kebabs, mostly) -- it's called 'McKing',
and the logo for the 'M' uses _perfectly_ recognizable 'golden arches'.

With McDonald's well-known proneness to aggressive defense of their
trademarks, I'm astonished anybody could provoke them in such wise,
particularly a small direct competitor (maybe the strategy behind it
is that the lawsuit will produce a lot of publicity, and the David vs
Goliath effect, plus the likely antipathy against McDonald's on the
part of typical local consumers of Arab-cuisine fast-food, will in turn
increase the small shop's customer base by more than enough to pay
for the lawsuit's costs...?).


Alex





From tmf002 at email.mot.com  Fri Dec 22 12:21:09 2000
From: tmf002 at email.mot.com (Mark Wyatt)
Date: Fri, 22 Dec 2000 17:21:09 +0000
Subject: Serial I/O (Windows) Documentation
Message-ID: <3A438D85.2C52324A@email.mot.com>

Newbie Question:

Have found several references to WIN 32 serial I/O. No one seems to
mention documentation however. What is available? What I would really
like is to find a tutorial on serial I/O programming on the WIN 32/DOS
platforms (UNIX is way down my list of things to support for this little
project). Failing that, any kind of documentation would be a help.
	
	Thanks
	
			Mark


From slhath at home.com  Mon Dec 11 21:56:48 2000
From: slhath at home.com (Scott Hathaway)
Date: Tue, 12 Dec 2000 02:56:48 GMT
Subject: winreg
References: <sh9Z5.6204$bw.481106@news.flash.net> <913ifd018tv@news2.newsguy.com>
Message-ID: <QtgZ5.63981$yR4.1670115@news1.rdc1.tx.home.com>

Thanks!

Scott

"Alex Martelli" <aleaxit at yahoo.com> wrote in message
news:913ifd018tv at news2.newsguy.com...
> "Scott Hathaway" <slhath at home.com.nospam> wrote in message
> news:sh9Z5.6204$bw.481106 at news.flash.net...
> > Hello Everyone,
> >
> > I was looking at the online documentation on the winreg module that
allows
> > editing of the windows registry.  I am too new to python to understand
how
> > to code it.  Can someone get me started in the right direction?
> >
> > I need to add a key and then add 5 values to that key.
>
> With Python 2, for example:
>
> import _winreg as w
>
> parentKey = w.OpenKey(w.HKEY_CLASSES_ROOT,r'file\shell',0,w.KEY_WRITE)
> newKey = w.CreateKey(parentKey,'SubkeyName')
> w.SetValueEx(newKey,'AValueName1',0,w.REG_SZ,'Something here')
> w.SetValueEx(newKey,'AValueName2',0,w.REG_SZ,'Something else')
>
> will create a key 'file\shell\SubkeyName' (assuming 'file' and
'file\shell'
> already exist!) under HKEY_CLASSES_ROOT, with no default (unnamed)
> value, and two named values under it.
>
>
> Alex
>
>
>




From wware at world.std.com  Wed Dec 20 12:09:41 2000
From: wware at world.std.com (Will Ware)
Date: Wed, 20 Dec 2000 17:09:41 GMT
Subject: Music synthesizer in Python/C
References: <G588B3.6A5@world.std.com>
Message-ID: <G5vMC5.HE5@world.std.com>

Will Ware (wware at world.std.com) wrote:
> I have been tinkering with a few iterations of a software emulation
> of a music synthesizer...

I have done some further work on this. Performance improvements,
code cleanup, more tunes, more voices, etc. Bon appetit.
http://www.deja.com/getdoc.xp?AN=706876401&fmt=text

-- 
# - - - - - - - - - - - - - - - - - - - - - - - -
# Resistance is futile. Capacitance is efficacious.
# Will Ware	email:    wware @ world.std.com


From katz at Glue.umd.edu  Tue Dec 12 16:19:09 2000
From: katz at Glue.umd.edu (Roy Katz)
Date: Tue, 12 Dec 2000 16:19:09 -0500
Subject: Thank you, Fredrik. (was Re: Label-Value)
In-Reply-To: <WntZ5.3566$Kd1.414873@newsb.telia.net>
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <WntZ5.3566$Kd1.414873@newsb.telia.net>
Message-ID: <Pine.GSO.4.21.0012121550530.19682-100000@y.glue.umd.edu>

On Tue, 12 Dec 2000, Fredrik Lundh wrote:


> Sigh.  The problem is that you keep using terms that mean
> something for you, whether or not they apply to the current
> situation.  [snip]
> If you want to learn something new, you have to stop doing
> that.  [snip]
> Here's how it's done in Python:
> [wonderful explanation snipped]


Thank you very much, I understand now more or less how this works. 
Furthermore I appreciate your thoughtful, measured, and flame-free tone.  


Roey


--------

to everyone else:

Thanks for your responses.  
just to clarify something about '@identifier':  I'm playing devil's
advocate. So don't see this as an attack.  Judging
from the tone of voice in some responses, it seems they took my original
post (which was just a what-if) as some affront to the 
language.  Ordinarily, So I'm fine either way the discussion swings, as
long as it is constructive.  


I do, however, take to heart when people take potshots at me. 
If I say I've been programming Python for N years, take it
at face value.  It seems to make people feel better
when they come down on someone else's experience. "What, you say you've
been programming for *THREE* years and you *STILL* can't get it??" is pure
ill will.  I'm not trying to come off as some hotshot, I was only asking a
question.  If you must know why I have had difficulty grasping Python's
labelling system, perhaps it is because I have had 10+ years' C/C++
experience.  So that is what I know. 

Sometimes you've got to give someone the benefit of the doubt.  If s/he
mis-assumes, that's no reason to start letting loose with these kinds of
comments.  No doubt someone will append here some line like 'well you
should have read the python docs if you wanted to know
xyz..'.  Whatever. If I had found and understood this exact
documentation in the Python docs, I would not have posted the question.


Anyhow, Like I said, thank you all very
much for your responses, I appreciate them. To anyone who still feels 
a need to nitpick, go ahead and fix my spelling mistakes.  Honestly don't
care anymore; I got what I needed and now I'm done.








From scarblac at pino.selwerd.nl  Tue Dec 12 06:09:38 2000
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 12 Dec 2000 11:09:38 GMT
Subject: compiled or interpreted?
References: <w8jZ5.74702$IP1.2487256@news1.giganews.com>
Message-ID: <slrn93c1n3.76k.scarblac@pino.selwerd.nl>

Matt Frey <freydaddy at kricket.net> wrote in comp.lang.python:
> Is Python a compiled language that can be run on a platform like c++, or is
> it interpreted like java, or is it in a class of its own?

It is like Java; it is first compiled to bytecode, then the bytecode
is interpreted. A lot of the libraries are compiled C.

(But the Python interpreter runs the compiler by itself, you don't have
to call the compiler by hand).

-- 
Remco Gerlich


From gerson.kurz at t-online.de  Sat Dec 30 13:34:17 2000
From: gerson.kurz at t-online.de (Gerson Kurz)
Date: Sat, 30 Dec 2000 18:34:17 GMT
Subject: Embedding Python in Windows Applications
Message-ID: <3a4e28bf.16852652@news.t-online.de>

If I embedd Python in a Win32 console application (using
Demo\embed.c), everything works fine. If I take the very same piece of
code and put it in a Win32 Windows application (not MFC, just a plain
WinMain()) I see no output (and more importantly so, no errors),
because the application does not have a stdout/stderr set up. 

I have made the following "hack" to make it work (in VisualStudio)

    // stdout, stderr as defined in the MSVCRT
    _iob[1] = *fopen("D:\\stdout.log","w");
    _iob[2] = _iob[1];

but that surely cannot be a longterm solution. I don't want the output
to be written to a temp file, I want to see it in a callback. Is that
possible, and if so, how ? 



From amk at mira.erols.com  Thu Dec 21 19:52:12 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 22 Dec 2000 00:52:12 GMT
Subject: re different between US and German versions of NT
References: <mailman.977362144.17768.python-list@python.org> <slrn942rjn.l1.amk@207-172-111-180.s180.tnt1.ann.va.dialup.rcn.com> <r1i06.769$Qb7.67123@newsb.telia.net>
Message-ID: <slrn945a70.g7.amk@207-172-112-244.s244.tnt4.ann.va.dialup.rcn.com>

On Thu, 21 Dec 2000 07:10:15 GMT, Fredrik Lundh <fredrik at effbot.org> wrote:
>SRE uses isalnum only if you're using the LOCALE flag when
>you compile the pattern.

So it does.  I have to stop making quick glances at _sre.c to answer
questions; every time I do it leads to error.

--amk


From rzilavec at tcn.net  Wed Dec 13 10:40:04 2000
From: rzilavec at tcn.net (Richard Zilavec)
Date: Wed, 13 Dec 2000 15:40:04 GMT
Subject: Moving from Perl to Python
Message-ID: <3a3a9552.76673185@news.tcn.net>

I just got back from two book stores, nothing on Python, I was
shocked!  So I went online and found a pile but I hate ordering
online, I like to browse through them first.  Anyways, I really like
O'Reilly books, is the Programming Python really advanced or is more
along the lines of the Programming Perl where it covers the basics to
advanced?

man python on my machine doesn't work, I'm assuming something is wrong
or maybe its a perldoc equivalent?  Where can I find the FAQs?

Thanks in advanced for any help or suggestions.

--
 Richard Zilavec
 rzilavec at tcn.net


From fgeiger at datec.at  Sat Dec 23 04:48:30 2000
From: fgeiger at datec.at (Franz GEIGER)
Date: Sat, 23 Dec 2000 10:48:30 +0100
Subject: A Q concerning map (a comparison to Maple V)
Message-ID: <921sil$moe$1@newsreaderg1.core.theplanet.net>

When I encountered Python a few month ago now, it made me remember Maple V.
Maple has map, select, zip etc. like Python. But Maple handles map e.g. more
generously in terms of argument passing. So you can apply
f:=(x,y)->x^2+y
to a list and an ADDITIONAL argument like so:
map(f, L, 5);
If L was [1,2,3] this would yield [6, 9, 14].
If Python could do this too, I e.g could apply replace to a list of strings
like so:
map(replace, ["abc", "cde"], "c", "C");
instead of having to create a lambda function.
Wouldn't it make sense to have that in Python too? Do there come more such
opportunities into your mind?

Best regards
Franz GEIGER





From maxm at normik.dk  Wed Dec 13 08:15:58 2000
From: maxm at normik.dk (maxm)
Date: Wed, 13 Dec 2000 14:15:58 +0100
Subject: Getting PIL fonts into my win32 python PIL
Message-ID: <917sqb$d2f$1@news.inet.tele.dk>

I am using PIL 1.0 on windows, and it's tugging along nicely thank you.

But I have to use Helvetica Condensed to add some text to a gif, for a
(long) series of buttons. Does anybody know how to convert a truetype font
to a PIL font ???

My other alternative is to cut out each letter in Phototshop to save them as
images and then use these as a pseudo font. I would rather avoid this.

Regards Max M




From tismer at tismer.com  Mon Dec  4 11:53:30 2000
From: tismer at tismer.com (Christian Tismer)
Date: Mon, 04 Dec 2000 18:53:30 +0200
Subject: Python mailing list/usenet bridge.
References: <8vt9t4$r0f$1@nnrp1.deja.com> <900365$fp6$1@nnrp1.deja.com> <905t5s$6ef$1@nnrp1.deja.com> <200012041005.EAA23178@x56.deja.com> <20001204180354.C4396@xs4all.nl>
Message-ID: <3A2BCC0A.1ABDE864@tismer.com>


Thomas Wouters wrote:
> 
> On Mon, Dec 04, 2000 at 10:05:53AM +0000, Simon B. wrote:
> 
> > *Still* nothing. This is getting a bit irritating - deja is horrid.
> > *Please*, anyone know who I can nag to get this sorted?
> 
> Noone. The only people we can nag to have been nagged to already, and they
> can't fix it (except in the long term, which is what Barry *is* fixing.) If
> you want read-access, though, you can use 'tyrannewsaurus.xs4all.nl' as your
> newsserver. It's the publically accessible newsserver my company hosts, as a
> test-bed. We don't charge, we don't expect you to fill in forms, and we
> don't use your ipadress for anything. My first posting that said this
> attracted 500+ concurrent connections, so I'm wondering howmuch more this
> one is going to attract :-)

Thanks a lot! Works fine :-)

Something seems to be in fact very messy. I noticed that no messages
from me made it to the newsgroupsince Nov. 11, since then I wrote
at least ten.
What's the best advice for posting:
To the news, and maybe it doesn't get to the mailing list,
or vice versa?

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From mwh21 at cam.ac.uk  Thu Dec  7 10:25:00 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 07 Dec 2000 15:25:00 +0000
Subject: Printing modules contents
References: <3A2F84CC.3056C755@hotmail.com> <m3itow4cmi.fsf@atrus.jesus.cam.ac.uk> <3A2FA5C4.CC479A19@hotmail.com>
Message-ID: <m3d7f445kj.fsf@atrus.jesus.cam.ac.uk>

joonas <keisari_ at hotmail.com> writes:

> I tried this but it didnt worked.
> 
> <clip>
> import sys
> list = dir(sys)
> for x in list:
>   print x," = ",sys.x

print x,"=",repr(getattr(sys,x))

should work.

The output might be a bit bigger than you want, though...

M.

-- 
  The ability to quote is a serviceable substitute for wit.
                                                -- W. Somerset Maugham


From SBrunning at trisystems.co.uk  Mon Dec 11 10:25:25 2000
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Mon, 11 Dec 2000 15:25:25 -0000
Subject: while true: !!!
Message-ID: <31575A892FF6D1118F5800600846864D5B15AD@intrepid>

> From:	Jaroslav Gergic [SMTP:j_gergic at yahoo.com]
	(snip)
> Do you know better construction to avoid both - duplicate lines
> and horrible 'while 1' construction?
> 
> Something like:
> 
> while( "" != (line = fh.readline()) ):
>   ... do something ...
 
Unless the file is *really* big, I usually use:

for line in fh.readlines():
    do something with line

Simple, fast, and I've not personally had to process a file large enough
that this won't work.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk





-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From brian_zhouNOSPAM at agilent.com  Fri Dec  8 21:08:45 2000
From: brian_zhouNOSPAM at agilent.com (Brian Zhou)
Date: Fri, 8 Dec 2000 18:08:45 -0800
Subject: Jpython not work in Win2000
References: <3A3112D7.20504@hotmail.com>
Message-ID: <976327716.216220@emperor.labs.agilent.com>

http://www.egroups.com/message/jpython/4075
http://www.egroups.com/message/jpython/4071

Regards,

-Brian

"Hongzhen Zhang" <hzzhang at hotmail.com> wrote in message
news:3A3112D7.20504 at hotmail.com...
> Hi,
> I just installed Jpython on Windows 2000. After successful
> installation, under dos prompt, I type:
> jpython
> it complains that 'jpython' is not recognized as an internal
> or external command, operable program or batch file.
> The jpython is a executable jar file, according to the file
> property. How can that happen? But it works in Windows NT.
> Any suggestions and recommendations are highly welcome.
>
> zhang
>
>




From mmueller at dgfz.de  Wed Dec 13 08:30:12 2000
From: mmueller at dgfz.de (Mike Müller)
Date: Wed, 13 Dec 2000 14:30:12 +0100
Subject: unpacking Python tuple in C
References: <917e64$3bp@micky.ibh-dd.de> <m31yvc38hm.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <917t88$fi3@micky.ibh-dd.de>

>>
>> PyObject* floatFromPy0 = PyTuple_GetItem(result, 0);
>>   *double1= PyFloat_AsDouble(floatFromPy0)
>>
>> since the tuple is pretty long and I need all of the tuple items
>> PyTuple_GetItem seems to not the best solution.
>
>Why not?  Why not use a loop and stuff the values into an array of
>doubles?

Thanks for the tip. This works. As you can tell I don't have much experience
with C,
so simple things can cause big trouble.


>
>Supplying the right number of d's to PyArg_ParseTuple, maybe?

Tried "ddd...d" with an without ( ). No success.
But the loop with the array works (see above) and this just what I want.

>Stepping back from your immediate problem, what are you actually
>trying to do?  I can infer it from what you've said here, I'm afraid.

I am writing my program in Python (it's a water quality model). There is a
"main" model written by somebody else in FORTRAN which calls  submodels as
DLLs. My Python model is one of these submodels. Now I try to speeze
everything into a DLL. So C is the "glue" and just has to pass some args
back and force from the main to the submodel. I do NOT program any
functionallity in C (just the argument passing via the DLL). Besides this it
seems I need to improve my knowledge of C.

Thanks for your help.

Mike




From skidrow at nownuri.net  Sun Dec 10 06:06:53 2000
From: skidrow at nownuri.net (¼­Áö¿ø)
Date: Sun, 10 Dec 2000 20:06:53 +0900
Subject: questions 'bout snack module...
Message-ID: <90vo12$lu1$1@news1-2.kornet.net>

I get to know record/play with snack module,

but what I wanna do is get sound data and send it through network in
real-time...

how can i do that?..(should I use Thread or something?..) plz give me brief
example...

p.s) oh, I think I must use UDP instead of TCP/IP

but when I use SOCK_DGRAM instead of SOCK_STREAM, I get "socket error"

when the sock listens...

if possible, plz give me UDP socket program example also... :)




From erno-news at erno.iki.fi  Wed Dec 13 13:06:01 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 13 Dec 2000 20:06:01 +0200
Subject: Python-like compiled language
References: <913jr3$dmv$1@sylvester.vcn.bc.ca> <3A35835B.AAA92005@bioeng.ucsd.edu> <kuzoi178kg.fsf@lasipalatsi.fi> <9163mr02v2t@news1.newsguy.com> <kulmtl6xqx.fsf@lasipalatsi.fi> <917ipt023tk@news2.newsguy.com>
Message-ID: <kulmtk5h86.fsf@lasipalatsi.fi>

In article <917ipt023tk at news2.newsguy.com>, "Alex Martelli"
<aleaxit at yahoo.com> writes:

|| [good scheme suggestion snipped]
|| 
|| it is also rumored that ocaml has a good optimizing compiler.

| It has an _excellent_ compiler, but it has no similarity to
| Python whatsoever.

well, compared to c/c++ at least it is quite high-level and it has
nice syntax. :)

but i agree scheme is probably a better match.

  -- erno


From syver at NOSPAMcyberwatcher.com  Fri Dec 22 10:34:16 2000
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Fri, 22 Dec 2000 16:34:16 +0100
Subject: mail processing module?
References: <91rcig$rgu$1@nnrp1.deja.com> <m266kd45m0.fsf@freak.kaiserty.com>
Message-ID: <91vsej$246$1@troll.powertech.no>

"Michael Haggerty" <mhagger at alum.mit.edu> wrote in message
news:m266kd45m0.fsf at freak.kaiserty.com...

> In fact maybe a better breakdown would be one module that represents
> an email message and allows data to be MIMEd into and out of it, one
> module for sending such messages via SMTP, one module for retrieving
> such messages via POP, and one for retrieving via IMAP.  The
> interfaces for the two retrieving modules should be as close as
> possible, and ideally the interface to the mail-sending routine should
> be generic enough that one could write a module with the same
> interface but that sends emails by writing them to, e.g., sendmail or
> qmail.

Couldn't this be accomplished by extending the mimetools.Message class? I
checking out mail messaging via smtplib yesterday, and saw that the
mimetools.Message class seemed to have rather good support for "decoding" a
received message, but none for building a message. So what I basically felt
I needed was a message class for outgoing messages that could be used with
smtplib.SMTP. What I did to avoid having to think about getting the protocol
right all the time, was to make a small smtpmessage class that allowed you
to set such things as from, to, subject, message. I could then pass in the
SMTP object in the send method.




From moshez at zadka.site.co.il  Fri Dec 29 01:59:47 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Fri, 29 Dec 2000 08:59:47 +0200 (IST)
Subject: One Python 2.1 idea
In-Reply-To: <92f1it01at1@news1.newsguy.com>
References: <92f1it01at1@news1.newsguy.com>, <mailman.977978231.26163.python-list@python.org>
Message-ID: <20001229065947.71C20A84F@darjeeling.zadka.site.co.il>

On Thu, 28 Dec 2000, "Alex Martelli" <aleaxit at yahoo.com> wrote:

> Now, I find out that, for that one instance only, I want to
> override the normal Ogre.foo method with another function
> which I also happen to have around:
> 
> def fun(self):
>     print 'have fun!'
> 
> 
> What are the 'good Pythonic ways' to do this, as opposed to
> 'dirty tricks'?  It seemed to me that, say:
> 
>     x.foo = new.instancemethod(fun, x, Ogre)
> 
> might be simpler and clearer than some alternative such as:
> 
>     x.__class__ = new.classobj('NewOgre', (Ogre,), {'foo':fun})

Personally, I avoid the ``new'' module completely (last I checked,
it wasn't portable to Jython, and it's all around tricky).

I'd simply do:

class _temp(Ogre):
	foo = fun
x.__class__ = _temp

Sometimes even encapsulating it in a function:

def add_method_to_instance(instance, name, function):
	class _temp(instance.__class__):
		pass
	setattr(_temp, name, function)
	instance.__class__ = _temp

One additional benefit of the __class__ method is that it is
honest: you change the class when you changed the methods,
because behaviour *has* changed.

-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From bob at staff.cs.usyd.edu.au  Wed Dec 27 19:26:56 2000
From: bob at staff.cs.usyd.edu.au (Bob Kummerfeld)
Date: 28 Dec 2000 11:26:56 +1100
Subject: Python vs. PHP (& Java?)
References: <mailman.977897470.4734.python-list@python.org>
Message-ID: <92e1cg$5n2@staff.cs.usyd.edu.au>

Checkout PMZ (Poor Man's Zope) at http://www.suxers.de/python/pmz.htm

>From the home page:

"PMZ is a small Python script that allows you to combine HTML and
Python code. It is very similar to Microsoft Active Server Pages or
PHP3 or PHP4. So you can easily integrate Python into your HTML pages "

Bob Kummerfeld

-- 



From aleaxit at yahoo.com  Wed Dec 27 04:45:06 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 27 Dec 2000 10:45:06 +0100
Subject: Python vs PHP (& Java?)
References: <mailman.977897649.4877.python-list@python.org>
Message-ID: <92cdn00hck@news2.newsguy.com>

"Ben Ocean" <zope at TheWebsons.com> wrote in message
news:mailman.977897649.4877.python-list at python.org...
> Hi;
> I?m an intermediate-level programmer that?s new to Python (via Zope) and I
> like it a lot. However, it appears to me that, while Python is a good
> replacement for Perl, it cannot replace much of the functionality of PHP,
> which can be scripted into HTML (whereas Python, it appears, must be

I'm not sure what you mean.  There are many setups which let you
mix Python code right into HTML 'templates', just as in Active
Server Pages (ASP's) or PHP approaches; for example, ASP's themselves
work fine with Python if your server is Microsoft IIS (using
Hammond's "win32all" extensions, or the ActivePython distribution
from ActiveState, you can install Python as an activescripting
language and call it from anywhere that supports activescripting,
which includes many important Microsoft applications & servers).

So, _what_ "much functionality" of PHP couldn't be "replaced" by
Python solutions?

> called). So, I have to wonder about the speed of Python vs. PHP for
> Internet-based applications. Python has the advantage of byte-code and
> compilation, but it?s server-side. PHP might not be as fast on the server,
> but it doesn?t have to make as many trips back and forth between server
and
> client.

Performance issues need good benchmarking practice.  But why would
a Python solution require any more round-trips than any other
server-side solution?!

> Plus, I?m curious as to why Python wasn?t written in such a way that
> it could be written directly into HTML documents like PHP. On a similar

What makes you think it wasn't?  As I said, there are many deployment
approaches that let you do that, if that's what your heart hankers for.

Of course, Python can be deployed in a zillion other ways as well --
it's not JUST for server-side scripting of HTML -- and in any given
case other deployment options might prove faster, more productive, or
whatever.

> subject, someone once informed me that Java was by far the best language
for
> scripting communication programs over the Internet (like chat, IRC, etc.).
> How does Python compare? Your comments are appreciated.

I've done a reasonable amount of Java coding, and, though it was a good
while ago, I don't think the net-oriented parts of Java have changed
by enough to modify my judgment.  My observation is that Java gave me
about 15-20% net programming productivity enhancement over C++ (being
an expert at C++, and relatively new at Java); Python, today, is giving
me about a range of 50% to 200% programming productivity enhancement
over C++.  Thus, although I have not tried out the *same* task in both
Python and Java, I confidently predict that Python would still let me
be MUCH more productive than Java would.

Just about the only plus for Java would be that a client's browser is
likely to have SOME Java version in it -- hard to tell how good or
updated it will be, but _some_ *very* elementary applets should work,
if you can bear to code at VERY, VERY minimal lowest-common-denominator
standards (be prepared to wave bye-bye to ANY productivity if you take
this route: just testing on AT LEAST half a dozen Java/Platform combos,
and empirically finding out just how low that lowest c.d. is going to
be, IS going to impact your productivity-to-deployment by an order of
magnitude -- if you're lucky).  With Python, assuming you want to do
distributed processing rather than just server-side, you cannot rely on
a deployment strategy based on 'everybody already having Python': you
have to get some Python installed on every target machine (Gordon McMillan
has a great program to let you prepare minimal installers for such
purposes).  On the plus side, for that one-time cost you gain much better
control on *what* Python versions are going to be on either side of the
communication channel, so the testing headaches go down A LOT (so much,
in fact, that you should consider a similar deployment strategy even for
Java or whatever else, even though it may not be as handy as for Python).


Do notice one thing -- if you master Python and its accompanying tricks
of the trade (e.g., preparing installers), there is almost no limit to
what you can do with it (excepting only computationally-heavy, and low-
level-system-programming, stuff; for which tasks your Python code does
need to be driving lower-level modules -- but there are a LOT of those,
free and high-quality, so the problem might reduce to finding the right
one:-).  Python gives you enormous leverage in the real world -- it
"plays well with others* in a way that no other language I know does.

Server-side scripting?  Python's great at that.  Need to work in a JVM
driving some existing Java classes?  Jython, the JVM-version of Python,
may solve that for you.  Network-distributed processing?  Python would
be my first choice for that -- take your pick of Corba, DCOM, SOAP,
HTTP, or Python-specialized approach such as pyro.  Need to script and
automate Visio or Powerpoint or Excel or some other MS app?  Python is
insanely great at doing COM Automation, too.  Heavy numerical processing
of big arrays of numbers?  Python, with the 'Numeric' module, is a
prime candidate for the task.  Symbolic and unlimited-precision
numeric computing?  PySymbolic and my own little gmpy at your service --
though they're pre-alpha, I'm told the speed already exceeds other
competing packages, and there may be further tuning soon.  Image
processing?  Why, PIL is the cat's pajamas.  Want to teach programming
to a total newbie?  Python's ideal for that.

It's hard to find a programming task for which Python is not among
the prime candidates (possibly with a leg up from some lower-level
module, see above).  How does it do it?  First and foremost, I think,
its *SIMPLICITY*.  If you follow this newsgroup for a while, you'll
see a zillion threads complaining about the lack of feature X, with
responses explaining why you don't really need X and/or how easy it
is to emulate it with Python's existing machinery.  If Python had
followed even just 10-15% of the suggestions/requests/demands for
such 'improvements' over the years, it would be just as big, complex,
unwieldy, and less-practical, as any other decent language.  But,
fortunately, the genius of Guido van Rossum, Python's inventor and
Benevolent Dictator For Life (BDFL), is in resisting such urges --
a unique mix of prudent conservatism, aesthetic intuition, and sheer,
clever practicality, which has permitted Python to grow into the
truly unique platform it is today.  Even he slips from time to time
(only one example of that, actually: the cursed "print >>" construct!-),
but one slip or so per decade is still a better batting average than
anybody else's!-).  And gradually, the whole 'culture' of Python has
grown up to realize this source for the language's strength and
support it intensely.


Alex





From wcuddy at lserv.ja10629.home  Sat Dec  9 13:07:03 2000
From: wcuddy at lserv.ja10629.home (Wayne Cuddy)
Date: Sat, 09 Dec 2000 18:07:03 GMT
Subject: using execfile() in class/instance methods
References: <slrn932qna.rtk.wcuddy@lserv.ja10629.home> <90tb1302s27@news2.newsguy.com>
Message-ID: <slrn934t6r.s9l.wcuddy@lserv.ja10629.home>

In article <90tb1302s27 at news2.newsguy.com>, Alex Martelli wrote:
>"Wayne Cuddy" <wcuddy at lserv.ja10629.home> wrote in message
>news:slrn932qna.rtk.wcuddy at lserv.ja10629.home...
>> I am attempting to use the execfile() function to "source" a file that
>> has valid python code to create a diction object:
>>
>> bn { 'key' : 'value1',
>>         'key2' : 'value2' }
>
>This is not valid Python code -- it's probably missing an equal-sign
>between the identified bn and the open-brace.
I made a typo, there actually was an '=' after the bn in my file.
>
>
>> I am issueing this command from a class constructor method and would
>> like bn to be bound the class instance.  So in the future I can say:
>>         print self.bn['key']
>
>Piece of cake.
>
>-- file: ba.txt
>bn = { 'key' : 'value1',
>     'key2' : 'value2' }
>-- end of file: ba.txt
>
>-- file: ba.py
>class Ba:
>    execfile('ba.txt')
>    def bu(self):
>        print self.bn['key']
>
>ba = Ba()
>
>ba.bu()
>-- end of file: ba.py
>
>D:\PySym>python ba.py
>value1
>
(I hit something in emacs to make everything go upper case in the
following text)
Your Example Worked But I Want Multiple Instances Of This Class And Be
Able To Pass The Filename As An Arg.  I Also Would Prefer That Bn Be A
Local Variable Of The Function Containing The Execfile() Call.  This
Would Allow Me To Either Return It Or Bind It To Any Variable Name Of
The Instance.  So I Want To Do Something Like This:

Class Bla:
      ...
      ...
      Def F1(Self,Filename):
	  Execfile(Filename)
	  Self.Dictfromdisk = Bn

      Def F2(Self):
	  Print Bn

I Have Been Able To Get Different Behavior By Passing In Globals() And
Locals() But I have Not Been Able To Get Bn To Be A Local Method Variable.
>
>> Is this possible?  I assume I am screwing up the global() and locals()
>> arguments to the function which I can say that I don't fully
>
>I think it was probably just the missing '=' sign, as above.  The default
>for globals and locals seem to be OK in this case.
>
>> understand.  Also is there another better/worse way to do this???just
>> so I can see what else is out there..
>
>execfile is very risky if the textfile you're executing may contain
>undesired
>stuff, of course; then, you might want to build up the self.bn dictionary
>in a safer manner, which might then be highly preferable.
>
>But if the textfile is fully under your control, this might be quite OK.
>
I am in total control..

thanks,
Wayne
>
>Alex
>
>
>


From tdelaney at avaya.com  Tue Dec 19 19:00:52 2000
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 20 Dec 2000 11:00:52 +1100
Subject: while true: !!!
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B0FBCD8@natasha.auslabs.avaya.com>

> "Delaney, Timothy" wrote:
> > maybe_buffered_stream = MaybeBufferedFileInputStream(file_object)
> >
> > while maybe_buffered_stream.has_unread_contents:
> >        line = maybe_buffered_stream.readline()
> 
> so you prefer slow and fragile code over fast and robust
> code.  fine.  now please shut up, and let this thread die.

ROTFL. Or were you serious?

There is nothing here which is any more fragile than any other method. Of
course, I omitted the code that dealt with exceptional circumstances for
clarity ... perhaps this confused you.

Shutting up now.

Tim Delaney
Avaya Australia



From glenfant at equod.com.nospam  Thu Dec  7 07:48:32 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Thu, 7 Dec 2000 13:48:32 +0100
Subject: Oracle 8.1.6 8.1.7
References: <90np2v$7hk$1@proxy.fe.internet.bosch.de>
Message-ID: <90o0n0$9it$1@reader1.imaginet.fr>

This is a way... (look at mxODBC package from this URL)

http://www.lemburg.com/files/python/


"Jens Vogelgesang" <Jens.Vogelgesang at de.bosch.com> a ?crit dans le message
news: 90np2v$7hk$1 at proxy.fe.internet.bosch.de...
> Hi there,
>
> ist there a way to get a connection to an Oracle DB with Version 8.1.6 or
> 8.1.7.
>
> Jens
>
>



From fredrik at effbot.org  Sat Dec  9 10:54:12 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sat, 09 Dec 2000 15:54:12 GMT
Subject: sgmllib nit
References: <eyErVLAEHlM6EwA4@jessikat.fsnet.co.uk>
Message-ID: <EAsY5.2808$O%.435892@newsc.telia.net>

Robin Becker wrote:
> The current version of sgmllib doesn't allow the detection of attributes
> without values.

as stated in the documentation, the current version of sgmllib
only handles enough SGML to be able to parse HTML.

in HTML, attributes without values (minimized form) are treated
as boolean attributes.  boolean attributes can have one of three
formats:

    attribute="attribute" (full form: true)
    attribute (minimized form: true)
    (not present: false)

if you need to tell the difference between "true" and "true", you're
clearly not working with HTML (or anything compatible with HTML).

</F>




From fgeiger at datec.at  Wed Dec 27 12:31:00 2000
From: fgeiger at datec.at (Franz GEIGER)
Date: Wed, 27 Dec 2000 18:31:00 +0100
Subject: Exception Handling and Inheritance
References: <921sdr$mlq$1@newsreaderg1.core.theplanet.net> <806k4tc5t9js74aqv81u214ijiqsni9q2j@4ax.com>
Message-ID: <92d95s$8uq$1@newsreaderg1.core.theplanet.net>

Thank you, Robert,

in the meanwhile I had an other idea:

def Formatted(self):
    try:
        return "Exception '%d' (%s) in '%s'" % (self.iCode, self.sInfo,
self.sSource)
    except:
        return str(self.args)

Exception.Formatted = Formatted    # !!!

class TException(Exception):
    :
    :

class TException_Data(TException):
    :
    :

This would make Exception.Formatted() act like __str__ and polymorphically
call the Formatted() method on derived classes in situations like this:

except Exception, e:
    :
    :

So now there's only one handler at all even if I raise my TException.


> Exception as a catch all since it tends to mask programming errors. Of

Admitted.


Thanks and best regards
Franz GEIGER


"Robert Roy" <rjroy at NO-SPAM.magma.ca> wrote in message
news:806k4tc5t9js74aqv81u214ijiqsni9q2j at 4ax.com...
> On Sat, 23 Dec 2000 10:45:57 +0100, "Franz GEIGER" <fgeiger at datec.at>
> wrote:
>
> >
> >I'd like to specialize Exceptions. So I derived a TException from
Exception
> >and added all the things I think I need to have.
> >
> >But now there is a problem: I have created a second branch in the
> >inheritance tree:
> >Exception
> >    StandardError
> >        IOError
> >        ...
> >    TException
> >        TDataException
> >        ...
> >
> >When I want to catch TExceptions I write
> >try:
> >    obj.meth(anyStuff)
> >except TException, e:
> >    print >> sys.stderr, "Panic: %s!" % e.Formatted()
> >
> >But what about Exceptions? They are not caught by my TException handler!
And
> >Exception does not have my Formatted method.
> >
>
> You could always add __str__ = Formatted to your class definintion
> since the published interface to Exception supports the str()
> function.
>
> Then  you could write
> try:
>     obj.meth(anyStuff)
> except Exception, e:
>     print >> sys.stderr, "Panic:", e
>
> Since TException is derived from Exception this would work.
>
> >Do I really have to catch them in a 2nd handler? Or can I add some
> >functionality somehow to Exception instead of implementing it in a
derived
> >TException?
>
> As Aahz mentionned, it is usually considered a Bad Idea to use
> Exception as a catch all since it tends to mask programming errors. Of
> course you could always Exception then do:
> if e.__class__ == AException:
> handle a
> elif e.__class__ == TException:
> handle t
> etc...
>
> but this does not really improve your lot in life over multiple
> handlers
>
> >
> >Many thanks in advance and best regards
> >Franz GEIGER
> >
> >
>
> Bob




From moshez at zadka.site.co.il  Sun Dec 31 10:31:43 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Sun, 31 Dec 2000 17:31:43 +0200 (IST)
Subject: Interesting behaviour of the assignment
In-Reply-To: <92j83q03gh@news1.newsguy.com>
References: <92j83q03gh@news1.newsguy.com>, <92f4gk$qk6$2@news.nuri.net> <92fhsr01u88@news1.newsguy.com> <mailman.978059501.14500.python-list@python.org> <92j4n1$mhu$0@216.39.162.232>
Message-ID: <20001231153143.86832A84F@darjeeling.zadka.site.co.il>

[Thomas Wouters]
> In other words, 'x == y' first does 'x is y', and if it's true, the
> result
> of the operation is true. Only if 'x is y' is false will it start the
> (possibly very expensive) comparison-by-contents operation.

[Michael Esveldt]
> This is a somewhat ignorant question, but if 'x == y' checks for 'x is
> y' first, why use 'is' at all? If you only get a tiny bit of speed out
> of using 'is' is it worth it to have this in the language at all? (Not
> that it would ever change, just curious.)

[Alex Martelli]
> Moshe gave an example: *defeating* polymorphism attempts on
> the part of object implementers, based on their implementing
> __cmp__.  Why one would actively work to impede polymorphism,
> I don't know, but it sure can work that way.

Here's an example where I'd rather use "is" then "==":

# module foo.py
SPAM = 0
EGGS = 1

def eat(what):
	if what is SPAM:
		eat_spam()
	if what is EGGS:
		eat_eggs()
	raise FoodError()

# user code:
import foo

foo.eat(foo.SPAM)
foo.eat(foo.EGGS)

The semantics are accurate here: foo.eat should not be passed 0.0,
it should be passed only foo.SPAM or foo.EGGS, so a FoodError
will alert the programmer to his bug sooner.
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From neelk at alum.mit.edu  Sat Dec 30 09:32:57 2000
From: neelk at alum.mit.edu (Neelakantan Krishnaswami)
Date: 30 Dec 2000 14:32:57 GMT
Subject: Interesting behaviour of the assignment
References: <92fhsr01u88@news1.newsguy.com>, <92f4gk$qk6$2@news.nuri.net> <mailman.978059551.14512.python-list@python.org>
Message-ID: <slrn94s0c2.mr6.neelk@alum.mit.edu>

On Fri, 29 Dec 2000 09:07:21 +0200 (IST), Moshe Zadka wrote:
>On Thu, 28 Dec 2000, "Alex Martelli" <aleaxit at yahoo.com> wrote:
>
>> (Actually, in the case of None, the semantics of is and
>> == are identical, but 'is' can be a tiny little bit faster).
>
>No.
>
>>>> class FoolAlex:
>...         def __cmp__(self, other):
>...             return cmp(None, other)
>...
>>>> FoolAlex() == None
>1
>>>> FoolAlex() is None
>0
>
> Which is why, in code I write, I always use "x is None" -- faster
> and more reliable.

Oddly enough, this is why I always use '=='. I figure that if the
class's author wanted the object to test the same as None, he or
she had a good reason. :)


Neel



From trotts at llnl.gov  Sun Dec 10 22:03:25 2000
From: trotts at llnl.gov (Issac Trotts)
Date: Sun, 10 Dec 2000 19:03:25 -0800
Subject: Newbee seeking help! Extending Python with C/C++
In-Reply-To: <DLEOLHIBOMHGMMPFOKGMCEPODEAA.pkocmid@atlas.cz>; from pkocmid@atlas.cz on Mon, Dec 11, 2000 at 03:21:05AM +0100
References: <3A33D106.3F755E1@magnet.fsu.edu> <DLEOLHIBOMHGMMPFOKGMCEPODEAA.pkocmid@atlas.cz>
Message-ID: <20001210190325.A16977@llnl.gov>

You can also download swig at 

http://sourceforge.net/projects/swig/

Issac

On Mon, Dec 11, 2000 at 03:21:05AM +0100, Petr Kocmid wrote:
> www.boost.org Hot new stuff is Boost C++ Python Library from David Abrahams.
> 
> Petr Kocmid
> pkocmid at atlas.cz
> 
> > -----Original Message-----
> > From: python-list-admin at python.org 
> > [mailto:python-list-admin at python.org]On Behalf Of Khaled Khairy
> > Sent: Sunday, December 10, 2000 7:53 PM
> > To: python-list at python.org
> > Subject: Newbee seeking help! Extending Python with C/C++
> > 
> > 
> > www.swig.org seems to be down or not responding. The links I got to
> > through the search engines are to documentaion that is fairly old
> > (Python 1.4 and rearlier). I have Active State Python 2.0 installed and
> > need to extend it using a lot of C++ code that is already written. Pease
> > advise.
> > Khaled Khairy
> > 
> > -- 
> > http://www.python.org/mailman/listinfo/python-list
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list



From claird at starbase.neosoft.com  Tue Dec 26 09:37:34 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 26 Dec 2000 08:37:34 -0600
Subject: registry tweaking program
References: <9250ek$ueo$1@nnrp1.deja.com> <9255kv0qf8@news1.newsguy.com>
Message-ID: <7BF73DCDDEBF06D0.F27538BDF52BDECD.535DA9B20E5D15B3@lp.airnews.net>

In article <9255kv0qf8 at news1.newsguy.com>,
Alex Martelli <aleaxit at yahoo.com> wrote:
><kevrjohn at my-deja.com> wrote in message news:9250ek$ueo$1 at nnrp1.deja.com...
>>    I would like to write a program that will automatically modify
>> certain parts of the registry in win98.  I have never programed before
>> (but I am trying to learn) and I thought that this would be a
>> good "real app" for me because I modify the registry a lot, and I would
>> like to see if I could automate the process to save some time.  I
>> really only want it to go and do a few preprogrammed things.  Nothing
>> spectacular like regedit.exe.  So how do you get a python script to
>> access and/or modify the registry?  I have looked around and I haven't
>> found any HOW-TO on the subject.
>
>The _winreg module, 20.2 in Python 2's Library Reference, or e.g.:
>http://www.python.org/doc/current/lib/module--winreg.html
>may be what you're looking for -- somewhat low-level, but OK.
			.
			.
			.
I want to add specific encouragement to the original
poster:  Python is a *wonderful* first computing lan-
guage.  If you already are in the habit of working
with the Registry with distribution Microsoft tools,
you're likely to find programming automation a nearly
religious experience.  Look forward to a lot of fun.
-- 

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


From erno-news at erno.iki.fi  Sat Dec 23 08:27:47 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 23 Dec 2000 15:27:47 +0200
Subject: namespaces & indentation
References: <mailman.977538666.16764.python-list@python.org> <92267v$q3e$1@news.mch.sbs.de>
Message-ID: <kuk88ri7xo.fsf@lasipalatsi.fi>

>>>>> "Chris" == Chris Schaller <Chris.Schaller at web.de> writes:

| I've already taken a look in my despair at that document, but
| couldn't believe that the smallest piece of code with its own
| namespace is a function.

well, keep trying :)

| This implies (at least to me) to keep functions as small as
| possible in order to avoid further complications (at least some
| functional programming philosophy...).

what complications?

it's good to keep your functions reasonably small, but
"as small as possible" is probably too small :)

try reading some existing python code.

|| 1. Strive to test all paths thru each function (good advice
|| regardless of language).

| That's right, but that's not so easy in Python (sorry, I'm a little
| bit fastidious from using FPLs) due to its dynamics.  Otherwise such
| cases would have been analyzed and criticized when parsing the
| function.  The same applies to a function that gets a wrong argument
| type and runs amuck.

static typing is no substitute for actually testing the code!

  -- erno


From grey at despair.rpglink.com  Fri Dec 29 14:01:08 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Fri, 29 Dec 2000 19:01:08 -0000
Subject: Great New Book
References: <t4o1vr8hktk531@corp.supernews.com> <92ij47$n3p$1@bob.news.rcn.net>
Message-ID: <slrn94pnrk.irm.grey@teleute.rpglink.com>

On 29 Dec 2000 17:54:15 GMT, Fred Yankowski <fcy at enteract.com> wrote:
>I actually prefer heavier paper since it allows me to wield my
>highlighter with abandon without worrying about marring the backside
>of the marked page.

    Not to mention 2'5" isn't all that thick.  I have fictional paperbacks
that are that thick.  :)

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From emile at fenx.com  Sat Dec 23 21:21:18 2000
From: emile at fenx.com (Emile van Sebille)
Date: Sat, 23 Dec 2000 18:21:18 -0800
Subject: About WebBrowser
References: <91vbr9$eo1q2@SGI3651ef0.iddeo.es> <9235bj$6e6$1@troll.powertech.no> <9236f3$6rb$1@troll.powertech.no> <3a45260d.52012638@news.skynet.be>
Message-ID: <923mk9$5qhqq$1@ID-11957.news.dfncis.de>

You'll need to download and install Mark Hammond pythonwin
extensions.  The easiest is to download the activestate
python release, (these urls will wrap)
http://www.activestate.com/download/ActivePython/windows/2.0
/ActivePython-2.0.0.202.msi  which includes python 2.0 as
well.  Or just get the extension from
http://www.activestate.com/download/ActivePython/windows/win
32all/win32all-135.exe

HTH,

--

Emile van Sebille
emile at fenx.com
-------------------


"Daniel" <Daniel.Kinnaer at AdValvas.be> wrote in message
news:3a45260d.52012638 at news.skynet.be...
> On Sat, 23 Dec 2000 22:43:36 +0100, "Syver Enstad"
> <syver at NOSPAMcyberwatcher.com> wrote:
>
>
> >Okay so I had to find out, is this what you're looking
for?
> >
> >import win32com.client
> >from win32com.client import constants
> >browser =
win32com.client.Dispatch('InternetExplorer.Application')
> >browser.Visible = -1
> >browser.Navigate('www.python.org')
> >browser.ExecWB(constants.OLECMDID_PRINT,
constants.OLECMDEXECOPT_PROMPTUSER)
> >
> >PS: Remember to use the makepy utility to generate
wrapper classes and
> >constants for you.
> >
> >
>
> Sorry to barge in like that, but as a pyton newbie, I like
to see what
> Python can do for me, therefore I tried the above code but
stumbled
> upon an error I do not understand(yet).
>
> Here it is :
>
> Traceback (innermost last):
>   File "C:/Python20/AProject/test1/browzer.py", line 1, in
?
>     import win32com.client
> ImportError: No module named win32com.client
>
> I suppose I need this  makepy-utility you write about?
What do I have
> to do to make this sourcecode work for me?
>
> Thanks for your time.
>
> Daniel
>
>




From aleaxit at yahoo.com  Wed Dec  6 12:04:07 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 6 Dec 2000 18:04:07 +0100
Subject: Schwartzian Transform (was Re: Sort)
References: <NlnX5.1444$yv1.2700363@nnrp5.proxad.net> <oa8s2tsls5foll62r8fpa6nlcpisb6v1ek@4ax.com> <90l9ir02avk@news2.newsguy.com> <3A2E5F2A.9E650D50@erols.com>
Message-ID: <90lrhl06t@news2.newsguy.com>

"Edward C. Jones" <edcjones at erols.com> wrote in message
news:3A2E5F2A.9E650D50 at erols.com...
> Alex Martelli wrote:
>
> > Faster for sorting a large list is a "Schwartzian Transform" approach:
> >
> >     auxlist = [ (x[2], x) for x in thelist ]
> >     auxlist.sort()
> >     thelist = [ x[1] for x in auxlist ]
>
> What is a "Schwartzian transform"?

Exactly what I exemplify here (and you quote the example): to sort a
sequence X,

    a. generate an auxiliary sequence Y, exactly as
        long as X, each item Y[i] being a sequence of
        the sort fields computed from X[i] followed
        by X[i] itself

        a1. if a stable-sort is needed, be sure to
            insert i itself in Y[i], just before X[i]

    b. Y.sort()

    c. X = [ x[-1] for x in Y ]

where steps b and c could, I guess, also be expressed
in English, but it would be too cumbersome & verbose:-).

There are alternative, similar ideas.  For example, you
could have i at the end of Y[i] NOT followed by X[i],
then step c would become

    c. X = [ X[x[-1]] for x in Y ]


The technique is named after its inventor Randall
Schwartz, who thought it up in the context of Perl
(but it works even better in Python:-).  The key
idea is that, in scripting languages such as Perl
and Python, 'sort' slows down a lot if each of its
O(N log N) comparison involves interpreting script
code, rather than just the fast, built-in "lexical
comparison" that the language provides.

Hence, we'd dearly like to just call
    something.sort()
rather than
    something.sort(lambda a, b: complex stuff
        involving a and b)
but how could an unadorned sort produce the order
we want...?

Answer: if the sortfields that are to be computed
or extracted from the original item X[i] were
*prepended* to it, appropriately.  Whence, pass a
of the Transform (pass c just 'undoes' that to get
the clean list right back).

Note that the possibly - complex transformations
yielding the sortkey are performed O(N) times, not
O(N log N) times.  Whence the speedup.


Example: perform a stable sort, case-independent,
over the first 20 characters of each line of the
838-lines README.txt file that comes with Python
2.0.

Naive approach...:

def compare(a,b):
    keya = a[:20].lower()
    keyb = b[:20].lower()
    return cmp(keya, keyb)

import time

ldata = open("D:/Python20/README.txt").readlines()

lines = ldata[:]

start1 = time.clock()
lines.sort(compare)
stend1 = time.clock()

# oops, this ISN'T stable, is it...?  Pretty
# hard to make Python's sort stable _without_
# a Transform... "here I come to save the day":

tlins = ldata[:]
start2 = time.clock()
aulin = [ ( x[:20].lower(), i)
    for x, i in zip(tlins, range(len(tlins))) ]
aulin.sort()
tlins = [ x[-1] for x in aulin]
stend2 = time.clock()

print stend1-start1
print stend2-start2


D:\PySym>python tesor.py
0.220513718779
0.0353952707969

D:\PySym>


Note how, even on this small dataset, the Transform
version is over 6 times faster (AND satisfies the
specs about the sort being stable, a nice plus:-).
An idiom WELL worth keeping in mind...


Alex





From tim.one at home.com  Thu Dec 21 20:02:23 2000
From: tim.one at home.com (Tim Peters)
Date: Thu, 21 Dec 2000 20:02:23 -0500
Subject: while true: !!!
In-Reply-To: <v0422080bb66846e9440a@[192.168.11.33]>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEAOIFAA.tim.one@home.com>

[Blair Lowe]
> ...
> WHO says that while 1 is a "standard python idiom"? Certainly not in
> my company!

I don't know of any defn. for "std idiom" that doesn't appeal to what
experienced programmers in a language routinely do.  Since Python's std
library contains "while 1" more than 400 times, it's hard to see how that
could *not* be considered a std Python idiom.

Don't know about Fredrik, but I'd be confused at first sight if it were
spelled in any other way.

then-again-i-confuse-easily<wink>-ly y'rs  - tim




From gregj at pobox.com  Thu Dec 21 02:20:49 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Thu, 21 Dec 2000 07:20:49 GMT
Subject: while true: !!!
References: <mailman.977268243.15476.python-list@python.org> <NnS%5.482$Qb7.42792@newsb.telia.net> <4bY%5.172261$U46.5444054@news1.sttls1.wa.home.com> <91q3gr$nbr$1@nnrp1.deja.com>
Message-ID: <lbi06.176440$U46.5556651@news1.sttls1.wa.home.com>

<eff at my-deja.com> wrote in message:

> so?  this newsgroup is cross-linked with a mailing list.
> how about showing the people on that list some respect?

Participating in a popular thread is disrespectful, but randomly flaming
people is OK? Who put you in charge?

> lots of new names in the killfile, you mean?

By all means, please add me to your killfile. And please get your
prescription refilled soon.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From aleaxit at yahoo.com  Wed Dec 27 15:37:30 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 27 Dec 2000 21:37:30 +0100
Subject: newbie : array initialization, anyone ?
References: <3a4a1352.9818548@news.t-online.de> <wgr26.1691$JG.213419@news.shore.net>
Message-ID: <92dpti01vqa@news2.newsguy.com>

<arcege at shore.net> wrote in message
news:wgr26.1691$JG.213419 at news.shore.net...
    [snip]
> This is in the FAQ (#4.50).  You will want to create each inner dimension
> individually:
>
> >>> multidim = [ None ] * 3
> >>> for i in range(3):
> ...   multidim[i] = [ 0 ] * 10

Or, more compactly, in Python 2:

>>> mudi = [ [0]*10 for i in range(3) ]

Not sure if the FAQ shows this handy list-comprehension use...?


Alex






From aleaxit at yahoo.com  Wed Dec 13 17:16:38 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 13 Dec 2000 23:16:38 +0100
Subject: Python-like compiled language
References: <913jr3$dmv$1@sylvester.vcn.bc.ca> <3A35835B.AAA92005@bioeng.ucsd.edu> <kuzoi178kg.fsf@lasipalatsi.fi> <9163mr02v2t@news1.newsguy.com> <kulmtl6xqx.fsf@lasipalatsi.fi> <917ipt023tk@news2.newsguy.com> <kulmtk5h86.fsf@lasipalatsi.fi>
Message-ID: <918sk802pg9@news1.newsguy.com>

"Erno Kuusela" <erno-news at erno.iki.fi> wrote in message
news:kulmtk5h86.fsf at lasipalatsi.fi...
> In article <917ipt023tk at news2.newsguy.com>, "Alex Martelli"
> <aleaxit at yahoo.com> writes:
>
> || [good scheme suggestion snipped]
> ||
> || it is also rumored that ocaml has a good optimizing compiler.
>
> | It has an _excellent_ compiler, but it has no similarity to
> | Python whatsoever.
>
> well, compared to c/c++ at least it is quite high-level and it has
> nice syntax. :)

We'll have to disagree on the latter, I fear.  I'm on record as
loathing C's syntax (and C++'s, because of compatibility with
C's), but OCaml's horrid crowded syntax is a factor that keeps
me away from it.  Do you LIKE writing
    2.1 +. 3.2 /. 4.3
...?  Just about every other infix-syntax language in the world
lets you write
    2.1 + 3.2 / 4.3
but OCaml insists on the dotted-operators (the undotted ones
are reserved for integer ops).  And that's just the start of it.

Actually, OCaml _does_ have somewhat of a C++ feel to it --
crowded syntax, lots of reasonably-coexisting (but not fully
integrated) paradigms, pretty good code generation (thanks
in part to a pretty rigid static typesystem), decent integration
with the environment.  It *IS* higher-level, and not as stable
or mature as C++, of course (with all that follows), but that
may be just a, oh, 10-years shift, maybe (the first OCaml book
came out in '95, I believe -- not too far from 10 years after
the first C++ one; there's a newer book now -- but many
things in it as marked as 'experimental, may change in future
releases').


Alex





From qrczak at knm.org.pl  Fri Dec 22 15:28:30 2000
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 22 Dec 2000 20:28:30 GMT
Subject: Good introduction to functional programming with Python?
References: <mailman.977465166.28872.python-list@python.org> <91v4m001una@news2.newsguy.com>
Message-ID: <slrn947ebd.qfb.qrczak@qrnik.zagroda>

Fri, 22 Dec 2000 09:50:56 +0100, Alex Martelli <aleaxit at yahoo.com> pisze:

> There must indeed be some Pythonic 'aura' to Haskell, I guess,
> given the somewhat surprising correlation between liking Haskell
> and liking Python among people who've tried both.

Indeed. I am very surprised how often Haskell is mentioned on the
Python's list, given quite different paradigms of these languages.

Elegance of design of a language? Significant layout? :-)

I am making a library for embedding the CPython interpreter in Haskell
programs (Glasgow Haskell only), to be prepared if I ever need to
make a scriptable program.

Handling basic Python objects already works. Includign conversion of
Haskell functions to Python bound method objects, with "self" being
a CObject, whose finalizer frees the dynamically created C function
pointer and thus removes implicit references to free variables of
the original function closure... Exceptions from either language
are going to be propagated through both languages, with appropriate
wrapping and unwrapping. Unfortunately data structures with cycles
across languages will not be freed.

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


From alain at onesite.org  Thu Dec 21 17:11:39 2000
From: alain at onesite.org (Alain TESIO)
Date: Thu, 21 Dec 2000 23:11:39 +0100
Subject: Right tools for python info space mapping project?
References: <91b7d4$9ml$1@nnrp1.deja.com>
Message-ID: <14v44tgs7bf2but90lns5r64r5rqobtof0@4ax.com>

On Thu, 14 Dec 2000 19:34:59 GMT, oivvio at polite.se wrote:

>1. RDBMS
>MySQL, or PostgreSQL. Which one has better python bindings?

Both bindings are fine, you can find PyGreSQL and MySQLdb on
http://www.vex.net/parnassus

Compare the engine feature and performance. PostgreSQL has more features but is slower,
MySQL is very quick, great for primarily read-only databases, but you can have concurrency
problems with an heavy load and a lot of simultaneous read and update accesses.
MySQL has full-text indexing in 3.23, lookup FULLTEXT in the manual.

> 2. Full text indexing.
> As neither of the free RDBMS have full text indexing this will have to
> been done separately. Sourceforge hosts Wordindexer (a perl thingy that
> works with MySQL), mifluz (a c++ word indexer), and pyMiflus (python
> bindings for mifluz). What are you're experiences with these and
> similar tools.

I'm using udmsearch (the new name is mnogosearch, http://search.mnogo.ru/) to index web
sites (live example here : http:/onesite.org/dipsearch/search.php3), you can also index a
database (look at htdb in its manual), no python frontend however.

Alain



From ngps at post1.com  Tue Dec 12 10:47:09 2000
From: ngps at post1.com (Ng Pheng Siong)
Date: Tue, 12 Dec 2000 23:47:09 +0800
Subject: corba, soap, com, etc...
In-Reply-To: <3A3500E6.7C8802FC@tismer.com>; from tismer@tismer.com on Mon, Dec 11, 2000 at 06:29:26PM +0200
References: <pKsY5.58762$yR4.1496678@news1.rdc1.tx.home.com> <3QzY5.9825$dg6.637046@pouncer.easynews.com> <912ls3$bn8$1@violet.singnet.com.sg> <3A3500E6.7C8802FC@tismer.com>
Message-ID: <20001212234709.A1431@madcap.dyndns.org>

On Mon, Dec 11, 2000 at 06:29:26PM +0200, Christian Tismer wrote:
> OmniORB comes with a GPL license, btw.

The code is GPL. The library can be used per LGPL, which is
an excellent decision! Kudos to the developers.

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




From emile at fenx.com  Fri Dec 22 09:33:36 2000
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 22 Dec 2000 06:33:36 -0800
Subject: Newbie Question
References: <05m64tkqev0aavea80ttjse1sj2gd447hg@4ax.com>
Message-ID: <91vomv$5i9dg$1@ID-11957.news.dfncis.de>

Do you intend to be invoking the unbound class method?  Or
should

os.path.walk( self.szDir, RmwExpireCheck.__CheckFile, self )

be

os.path.walk( self.szDir, self.__CheckFile, self )

--

Emile van Sebille
emile at fenx.com
-------------------


"Bob W." <removethis_kranki2 at earthlink.net> wrote in message
news:05m64tkqev0aavea80ttjse1sj2gd447hg at 4ax.com...
> I am trying to write the following class, but it fails in
the Run()
> method reference to __CheckFile and I am not sure how to
get it to
> work.  Thanks in advance for your help!
>
> class RmwExpireCheck:
>     def __init__( self, szFileName, szDir ):
>         self.szFileName = szFileName
>         FileStatList = os.stat( szFileName )
>         if FileStatList:
>             self.iCheckDate = FileStatList[8]
>         self.szDir = szDir
>         self.szFileList = []
>
>     def __getitem__(self, i):
>         line = self.szFileList[i]
>         if line:
>             return line
>         else:
>             raise IndexError
>
>     def __CheckFile( self, szDirName, szFileNames ):
>         for szFileName in szFileNames:
>             szPathName = os.path.join( szDirName,
szFileName )
>             if os.path.isfile( szPathName ):
>                 iFileTime = os.stat( szPathName )[8]
>                 if iFileTime > self.iCheckDate:
>                     self.szFileList.append( szPathName )
>
>     def Run( self ):
>         os.path.walk( self.szDir,
RmwExpireCheck.__CheckFile, self )
>
>
>




From grey at despair.rpglink.com  Thu Dec 28 14:25:58 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Thu, 28 Dec 2000 19:25:58 -0000
Subject: how do i check if a file exists with python?
References: <RlM26.98766$yR4.2639444@news1.rdc1.tx.home.com>
Message-ID: <slrn94n4u6.9am.grey@teleute.rpglink.com>

On Thu, 28 Dec 2000 19:18:09 GMT, Scott Hathaway <slhath at home.com> wrote:
>I have searched Deja, looked in two pyton books, and searched the python.org
>site trying to find out this simple question.

>Can someone please give me the call and the module it lives in?

    Well, 2 minutes and one Beazley book later.

os.path.exists()
os.path.isfile()

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From sill at localhost.kitenet.net  Fri Dec 15 16:45:28 2000
From: sill at localhost.kitenet.net (Oldayz)
Date: Fri, 15 Dec 2000 21:45:28 GMT
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <915jcf0c6o@news2.newsguy.com> <9162p3$7t2$1@animus.fel.iae.nl> <MBWZ5.334$h67.37748@newsread1.prod.itd.earthlink.net> <3A397C74.6C8906E1@cosc.canterbury.ac.nz> <91d5520m6j@news1.newsguy.com>
Message-ID: <Yhw_5.10621$h67.670217@newsread1.prod.itd.earthlink.net>

On Fri, 15 Dec 2000 14:08:15 +0100, Alex Martelli <aleaxit at yahoo.com> wrote:
>"Greg Ewing" <greg at cosc.canterbury.ac.nz> wrote in message
>news:3A397C74.6C8906E1 at cosc.canterbury.ac.nz...
>> Oldayz wrote:
>> >
>> > I thought something like forever: would be even more lucid
>>
>> If there is a break in the loop somewhere, it's
>> not really forever, though, is it?
>>
>> So perhaps it should *really* be
>>
>>    for a while:
>>       ...
>>
>> :-)
>
>Why not have a 'built-in constant' it_lasts, of value 1,
>so we can write:
>
>    while it_lasts:
>        ...
>
>This requires no change to existing Python syntax.  For
>the incurable romantics, besides allowing a 'for' statement
>form with a single parameter so cynics can write
>
>    for awhile:
>        ...
>
>and romantics can write:
>
>    for ever:
>        ...
>
>we should also introduce an 'until' statement, and a
>death_do_us_part built-in constant, of course.
>
>
>Alex

Ideally, to beat perl at its own game, we
should be able to write poems in python, i.e.:

quoth(the_raven):
	nevermore

I don't know what this block would do, though :/
Print 'nevermore' in caps?

-- 

	Andrei


From grante at visi.com  Wed Dec 20 14:12:58 2000
From: grante at visi.com (Grant Edwards)
Date: Wed, 20 Dec 2000 19:12:58 GMT
Subject: Python in embedded systems?
Message-ID: <_w706.11237$TU6.1069571@ptah.visi.com>

Has anybody heard of a Python interpreter being used in an
embedded system to provide user scripting capabilities?

At some point in the future, I might be looking for a small,
safe, interpreted langauge to put into an embedded system. It
needs to be something easy for a non EE/CS to grasp.
Resource-wise, Forth would be a good choice except I think it's
a difficult first language for non programmers.

I'd love to use Python, but I don't know if it will fit into
the embedded system environment -- I probably can't allocate
more than about 1MB of memory for the scripting feature.

I've seen some nice compact Scheme implimentations, but that
too is probably not a good choice for something that will be
used by somebody with no CS background.

I'm really hoping to avoid using some sort of BASIC, the Web
server I'm running as a JavaScript subset interpreter, so maybe
I can figure out how to use that for non-web stuff.

-- 
Grant Edwards                   grante             Yow!  Are we on STRIKE yet?
                                  at               
                               visi.com            


From krimsly at home.com  Sat Dec  9 22:38:30 2000
From: krimsly at home.com (North Krimsly)
Date: Sun, 10 Dec 2000 03:38:30 GMT
Subject: Python 2.0 Windows installer hangs
Message-ID: <WUCY5.129444$U46.4375622@news1.sttls1.wa.home.com>

Hello all.

When I run the Python 2.0 Windows installer on Windows 2000, I get to the
second screen (where it asks which components I want to install).  Then I
click on next and it just hangs, never to return.  I have disabled my
anti-virus utility as well as my software firewall.  No other applications
were running.  I'm stumped.  Any ideas on what might be causing the
installer to hang?  Sorry if this is the wrong venue for posting.  If so
please point me to the right place to post.

Thanks in advance.

-North




From glenfant at equod.com.nospam  Fri Dec  8 05:40:00 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Fri, 8 Dec 2000 11:40:00 +0100
Subject: "**" operator in method/function interfaces ??
Message-ID: <90qdi3$ar6$1@reader1.imaginet.fr>

Hi,

What's the use of "**" in such a construct ?
<sample>
...
class MyClass:
    def __init__(self, adata, **anotherdata)
...
</sample>

I found such construct in some examples of the Tkinter tutorial at
pythonware, and in some other documents.
I have gone around all books and web language tutorial I know and did not
find anything about it.

Is there an expert here who can explain what's the use of the ** operator in
that case ?

Thanks in advance.

Gilles Lenfant




From dalke at acm.org  Wed Dec  6 23:49:03 2000
From: dalke at acm.org (Andrew Dalke)
Date: Wed, 6 Dec 2000 21:49:03 -0700
Subject: Bug: Readline getting stuck on Linux and Solaris
References: <lxk8zqwupy2.fsf@ioasun3.epfl.ch> <90m53k$ovh$1@nnrp1.deja.com>
Message-ID: <90n4r0$5f1$1@slb0.atl.mindspring.net>

sragsdale at my-deja.com wrote
>I'm having similar problems with Python 2.0 hanging in Irix.  What's
>most annoying is that this happens on the Python command line when you
>hit control-C.  In Python 1.5.2 this is a KeyboardInterrupt event.  In
>Python 2.0 this hangs the interpreter which starts using 100% of the
>processor.

I had the same problem on IRIX  but with 1.5.2.  It was a
readline library problem (obviously :) and not part of Python
that I could tell.  After killing it to get the core dump, I
looked at the stack trace - it was stuck in a recursive sigaction
loop.

I don't know why you wouldn't have had the same problem with 1.5.2
as with 2.0.  Perhaps you compiled it against a different version
of the readline library?

                    Andrew
                    dalke at acm.org





From scarblac at pino.selwerd.nl  Mon Dec 18 12:25:18 2000
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 18 Dec 2000 17:25:18 GMT
Subject: search for pattern in StringList
References: <3a3e1935.5461298@news.skynet.be>
Message-ID: <slrn93shus.4tp.scarblac@pino.selwerd.nl>

Daniel <Daniel.Kinnaer at AdValvas.be> wrote in comp.lang.python:
> As a Python newbie, I was trying a very simple exercise, but got
> stuck. Is there someone who can help me complete this code, please?
> 
> 
> 
> MyList = ['an','been','call','door']
> 
> print "Scanning a StringList"
> 
> for i in range(0, len(MyList)):
>     print i, MyList[i]
> 
> #try to show every item in MyList which contains an 'a'
> #that is 'an' and 'call'.   In Pascal :
> #for i:=0 to MyList.Items.Count -1 do
> #    if pos('a',MyList.Items[i])>0 then writeln(MyList.Items[i]) 
> #in Python : ???

A string is a list of characters. To see if the list contains 'a', you
can use the 'in' operator:

>>> 'a' in 'bla'
1

Then the for statement becomes:

for s in MyList:
   if 'a' in s: 
      print s

If you're looking for strings longer than a single character, you can't
use 'in', you have to use the find() method:

for s in MyList:
   if s.find('ab') != -1:
      print s
 
This is a very common thing to do, select some items from a list based
on some criterium. filter() is a built-in function that does this:

print filter(MyList, lambda x: 'a' in x)

It "filters" the list according to the simple function given by the lambda,
that tests if 'a' in x.

Since Python 2.0, there is yet another way to write this:

print [s for s in Mylist if 'a' in s]

This is derived from set notation in math; it is 'the list of s, where
s comes from MyList and only those where "a" in s'.

For this simple example, these other options may be overkill, but
it's not a bad thing to know about them ;).


-- 
Remco Gerlich


From israelt at optushome.com.au  Wed Dec 13 08:28:14 2000
From: israelt at optushome.com.au (israel raj thomas)
Date: Wed, 13 Dec 2000 13:28:14 GMT
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de>
Message-ID: <402f3t4fn4gsvi05nrjdogf9lq88ba4vts@4ax.com>

>: Would it perhaps not be better to use Ruby ?
>
>Not for me, for the following reasons:
>a) I don't know Ruby well enough

Fair enough..

>b) Ruby is "even more" unknown in The Real World(tm) than Python

That is changing rapidly.


>d) I don't see any of the features Ruby has over Python as 'killer'
>   features.

Well, it touts itself as the successor to Smalltalk with a dash of
Eiffel's design by contract and a smidgeon of Clu/Sather.  However, I
agree that they are not "killer " features.


From auto103690 at hushmail.com  Tue Dec 26 02:32:36 2000
From: auto103690 at hushmail.com (Jeff)
Date: Tue, 26 Dec 2000 07:32:36 GMT
Subject: Season's greetings (was Re: Some Python 2.1 ideas)
References: <mailman.977765348.13376.python-list@python.org> <928otj02l39@news1.newsguy.com>
Message-ID: <oQX16.71486$df5.2622172@news1.crdva1.bc.home.com>

A non-text attachment was scrubbed...
Name: not available
Type: text/plain  charset=iso-8859-1
Size: 1834 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20001226/36841d3f/attachment.bin>

From mjackson at wc.eso.mc.xerox.com  Wed Dec 13 12:38:28 2000
From: mjackson at wc.eso.mc.xerox.com (Mark Jackson)
Date: 13 Dec 2000 17:38:28 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <9187tq02qlf@news2.newsguy.com>
Message-ID: <918c6k$90a$2@news.wrc.xerox.com>

"Alex Martelli" <aleaxit at yahoo.com> writes:
> "Steve Williams" <sandj.williams at gte.net> wrote in message
> news:3A379CB4.405992F8 at gte.net...
> > Alex Martelli wrote:
> >
> > > [snip]
> >
> > > Maybe they took solidity _for granted_, because, in their
> > > (Renaissance) times and in their (Architecture) calling,
> > > compromises regarding solidity were simply unthinkable.
> > > Well, we're not so lucky, in the software field, today; the
> > > Firmitas of *by far* most software around is imperfect.
> >
> > [snip]
> >
> > I understand the Romans required an architect to stand under his arch as
> > the falsework was being removed.
> 
> Neat idea (whether true or not -- that hardly matters!-), and
> gives good (albeit mythical) background for Vitruvius' order
> of priority.
> 
> The modern equivalent, in most cases, would be to require the
> appropriate *managers* connected with a sw-development project
> 'stand under the arch' (or whatever we can come up with as a
> fitting metaphor!).

I'm sure I remember reading, sometime in 1999, that the top managers of
the Chinese national airline had been told to be onboard an in-air
flight at the Y2K transition.  I have no idea whether the tale was
true, however.

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	I know nothing about this subject but I do have
	prejudices, which I am more than happy to share with you.
					- Leon Botstein




From grey at despair.rpglink.com  Sun Dec 31 09:50:11 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Sun, 31 Dec 2000 14:50:11 -0000
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <slrn94smds.150.grey@teleute.rpglink.com> <FOt36.17$vd1.2224@e420r-atl2.usenetserver.com> <slrn94t1vj.150.grey@teleute.rpglink.com> <mqw36.50$9s1.10042@e420r-atl2.usenetserver.com> <slrn94tqf3.150.grey@teleute.rpglink.com> <3A4F3671.AAA18BFF@stroeder.com>
Message-ID: <slrn94uht3.150.grey@teleute.rpglink.com>

On Sun, 31 Dec 2000 14:36:49 +0100, Michael Str?der <michael at stroeder.com>
wrote:
>While you're at it: Can you please *plonk* me too? We might disagree
>in the future and we can avoid flame wars in advance and save work.
>And it's a great honor for me to be in the same kill-file as the
>Eff-bot.

    I don't plonk people for disagreements.  I plonk them for stupidity.
There's a difference.  Since you have a moderate wit that shows at least some
intelligence.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From wsryu at is05.fas.harvard.edu  Wed Dec 13 08:55:31 2000
From: wsryu at is05.fas.harvard.edu (William Ryu)
Date: 13 Dec 2000 13:55:31 GMT
Subject: PIL binary for Python 2.0 for Windows ?
References: <3a2f472f$0$235@hades.is.co.za> <KZOX5.2015$Kd1.294314@newsb.telia.net>
Message-ID: <917v4j$it3$1@news.fas.harvard.edu>

I'm also having trouble installing PIL. While I can import Imaging.py just
fine and load the test image "lena.ppm," I get the dreaded:

"ImportError: The _imaging C module is not installed"

when I try to display the image with the show() method.

I've downloaded the PIL Python Modules from eefbot
(http://w1.132.telia.com/~u13208596/pil/index.htm) and have loaded the
_imaging20.dll into my DLLs folder.

I've also tried placing the DLL in my main Python20 folder and the PIL
folder, all of which are listed in PYTHONPATH.

I would really appreciate some help.

Thanks,


-wr


From fredrik at effbot.org  Wed Dec 13 15:06:37 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 13 Dec 2000 20:06:37 GMT
Subject: time.strptime(...) Python 1.5.2/Win32 missing ?
References: <918ken$pju$1@reader1.imaginet.fr>
Message-ID: <hFQZ5.3840$Kd1.440879@newsb.telia.net>

Gilles Lenfant wrote:
> Where's this function???
> >>> atime = time.strptime('10:20:30', '%H:%M:%S')
> Traceback (innermost last):
>   File "<pyshell#71>", line 1, in ?
>     atime = time.strptime('10:20:30', '%H:%M:%S')
> AttributeError: strptime
>
> Seems that there's no time.strptime function! Or did I miss something ?

http://www.python.org/doc/current/lib/module-time.html

    "strptime (string[, format])
    Parse a string representing a time according to a format.

    ...

    Note: This function relies entirely on the underlying
    platform's C library for the date parsing, and some of
    these libraries are buggy. There's nothing to be done
    about this short of a new, portable implementation of
    strptime().

    Availability: Most modern Unix systems."

If you're not using a "modern Unix system", chances are that
you don't have it.

There are several pure-Python implementations of this function
(search the comp.lang.python archives if you need one).

</F>




From fredrik at effbot.org  Sat Dec  9 04:51:50 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sat, 09 Dec 2000 09:51:50 GMT
Subject: Unit Testing tool
References: <90s69s$f85$1@troll.powertech.no>
Message-ID: <WgnY5.2710$O%.427485@newsc.telia.net>

Syver Enstad wrote:
> What is the Python unit testing tool preferred by readers/writers from
> comp.lang.python?

here's my favourite:

http://www.mems-exchange.org/exchange/software/python/quixote/unittest.html

there are other implementations of Kent Beck's smalltalk framework,
but the one above feels more "pythonic". ymmv, of course...

also see Tim Peters' doctest.py (sorry, not pointer).

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From paulp at ActiveState.com  Fri Dec  8 12:21:56 2000
From: paulp at ActiveState.com (Paul Prescod)
Date: Fri, 08 Dec 2000 09:21:56 -0800
Subject: The secrets of operator.add
References: <g08zpq51vy.fsf@scumbag.ecs.soton.ac.uk>
Message-ID: <3A3118B4.A6815DE@ActiveState.com>

Jacek Generowicz wrote:
> 

>    ncalls  tottime  percall  cumtime  percall filename:lineno(function)
>     49999    0.840    0.000    0.840    0.000 <string>:1(<lambda>)
>         1    0.720    0.720    1.560    1.560 <string>:1(?)
>         0    0.000             0.000          profile:0(profiler)
>         1    0.040    0.040    1.600    1.600 profile:0(reduce(lambda x,y:x+y, range(N)))
> 
> ...
>    ncalls  tottime  percall  cumtime  percall filename:lineno(function)
>         1    0.060    0.060    0.060    0.060 <string>:1(?)
>         0    0.000             0.000          profile:0(profiler)
>         1    0.000    0.000    0.060    0.060 profile:0(reduce(add, range(N)))
> 
> It looks as if multiple calls to operator.add are avoided somehow
> . . . or am I failing to understand the profiler's output ?

Operator.add is a builtin, not a Python function so calling it is very
quick and it doesn't really show up in the profiler.

> Is there a way of getting similar advantages in
> 
> run ('map(lambda x:x+1, range(N))')  ?

How about:

run('[x+1 for x in range(N)]')

> Is there a Python way of currying add?

lambda is the way to do currying. It may not be as efficient as you like
but that's how it works...

 Paul Prescod



From syver at NOSPAMcyberwatcher.com  Wed Dec  6 08:16:52 2000
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Wed, 6 Dec 2000 14:16:52 +0100
Subject: Global object in ASP (calling other scripting languages in ASP pages)
References: <90judk$edp$1@troll.powertech.no> <3A2D904F.1040905@ActiveState.com>
Message-ID: <GWqX5.1198$9B.36408@news1.oke.nextra.no>

> I guess that would be a JS implemted function?
Yes, thats correct.

> Try printing the repr() of your object - you may find it is not a string
> at all, but itself a COM object.  You may be able to just call this
object!

<% @Language=Python %>
<HTML>
<BODY>
<script language=JScript RunAt=Server>
function JscriptFunc()
{
 Response.Write("JScript called");
}
</script>


<%
Response.Write(str(dir()))
Response.Write('<br>')

Response.Write(ScriptingNamespace.JscriptFunc)
Response.Write('<br>')
Response.Write(repr(ScriptingNamespace.JscriptFunc))
Response.Write('<br>')
Response.Write(ScriptingNamespace.JscriptFunc())
Response.Write('<br>')
Response.Write(repr(ScriptingNamespace.JscriptFunc()))
Response.Write('<br>')

%>
</BODY>
</HTML>

Gives the output:

['ASPGLOBALTLB', 'Application', 'ObjectContext', 'Request', 'Response',
'ScriptingNamespace', 'Server', 'Session', '__builtins__', '__doc__',
'__name__', 'ax']
function JscriptFunc() { Response.Write("JScript called"); }
>
function JscriptFunc() { Response.Write("JScript called"); }
u'function JscriptFunc()\015\012{\015\012Response.Write("JScript
called");\015\012}'

>
> Try writing the test function in a different language to see if that
helps.

It does VbScript seems to work fine. Very nice, the downside is that the
site I am working with is written mostly in Jscript.





From kbass1 at nc-rr.com  Wed Dec 13 05:57:11 2000
From: kbass1 at nc-rr.com (Kevin B)
Date: Wed, 13 Dec 2000 10:57:11 GMT
Subject: Python and databases
Message-ID: <bCIZ5.23937$_y5.4911010@typhoon.southeast.rr.com>

Are there any resources that have information about Python and database
access? I have reviewing several book such as "Programming Python" and
"Learning Python" and have not found even a chapter on Python and database
access.

Thanks, Kevin




From teyc at bigfoot.com  Tue Dec 19 17:43:47 2000
From: teyc at bigfoot.com (Chui Tey)
Date: Wed, 20 Dec 2000 08:43:47 +1000
Subject: xml-rpc module question
References: <dmC%5.75166$yR4.2151642@news1.rdc1.tx.home.com>
Message-ID: <91ooea$9qb$1@bunyip.cc.uq.edu.au>

Not tested code:

from xmlrpcserver import RequestHandler

class my_handler(RequestHandler):

    def CubeIt(self, x):
        return x ** 3

    def call(self, method, params):
        if method == 'CubeIt':
            return self.CubeIt( params[0] )

if __name__ == '__main__':
    server = SocketServer.TCPServer(('', 8000), my_handler)
    server.serve_forever()


Chui







From patrick.fichou at galactica.it  Sat Dec  9 02:55:21 2000
From: patrick.fichou at galactica.it (Patrick)
Date: Sat, 9 Dec 2000 08:55:21 +0100
Subject: Newbie : Unable to use Jython under Win98
Message-ID: <OylY5.47189$%j3.534830@news6.giganews.com>

Hi,

I can't use Jython (Jython-20a1.class) under windows whith the JRE from Sun
(Jre1_3_0_win). With the outdated Microsoft virtual machine (Version
5.00.3319) it seems to run, but with an "old" Java 1.x engine !
Is there something important to do before I can use Jython ?

* Autoexec
* Path
* Install in a special directory (I've tried \Jython and \Windows\Jython)
* Reg modification
*...

Thanks





From scarblac at pino.selwerd.nl  Fri Dec 29 12:02:58 2000
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 29 Dec 2000 17:02:58 GMT
Subject: Need help with string method
References: <92if8v$iu1$1@nnrp1.deja.com>
Message-ID: <slrn94pgo7.jbc.scarblac@pino.selwerd.nl>

jvanar1 at my-deja.com <jvanar1 at my-deja.com> wrote in comp.lang.python:
> I have a pretty dumb question here.  I can't seem to get any string
> methods working from *within* a script.  here is a super-simple example
> that yields an error
> 
> >>import sys
> >>import string
> >>theString = 'This is a string'
> >>print theString.find('string')
> 
> now this works fine from the python command line but gives an error
> when run as a script. Here is the error message:
> 
> Traceback (innermost last):
>   File "./clean", line 18, in ?
>     print theString.find('string')
> AttributeError: 'string' object has no attribute 'find'

Version mismatch. I think you tried it with a 2.0 command line but the
script invokes the 1.5.2 interpreter (or similar). Try printing sys.version
from the script, and check the #! line (if it's on unix).

-- 
Remco Gerlich


From johngrayson at home.com  Tue Dec 19 07:15:09 2000
From: johngrayson at home.com (John Grayson)
Date: Tue, 19 Dec 2000 12:15:09 GMT
Subject: tkinter event binding to mouse double click
References: <91n7l6$7ij$1@news.nuri.net>
Message-ID: <91njgc$l4q$1@nnrp1.deja.com>

In article <91n7l6$7ij$1 at news.nuri.net>,
  "June Kim" <junaftnoon at nospamplzyahoo.com> wrote:
> I'm using Python 2.0 on Win32.
> In the code following, the binding doesn't seem to work; it can't
detect
> mouse double-clicking at all. (Maybe a platform-specific problem, I
reckon)
>
> from Tkinter import *
>
> root = Tk()
>
> def callback(event):
>     print "clicked at", event.x, event.y
>
> frame = Frame(root, width=100, height=100)
> frame.bind("<Double-Button-1>", callback)
> frame.pack()
>
> root.mainloop()
>
>

The problem is that you have to click on the *frame* not the (empty)
space in the frame.

Try this:

frame = Frame(root, bg='red', width=100, height=100)


John Grayson


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


From aleaxit at yahoo.com  Mon Dec 11 09:07:38 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 11 Dec 2000 15:07:38 +0100
Subject: Pascal -> Python 'record' structures
References: <2y4Z5.7634$KY1.20696@news1.rivrw1.nsw.optushome.com.au>
Message-ID: <912n3q0b88@news2.newsguy.com>

"deadmeat" <root@[127.0.0.1]> wrote in message
news:2y4Z5.7634$KY1.20696 at news1.rivrw1.nsw.optushome.com.au...
> Can anyone tell me how to read in file records that were created by Pascal
> and C programs? Will I have to write a wrapper and parse the raw data
> manually?

The python 'struct' module handles the task of converting such
binary 'struct'/'record' data to Python thingies.  You do have
to open the file for reading in binary mode and read the right
number of bytes at a time, but then struct will nicely unpack
them for you.

Suppose for example that the file is made up of records each
of which is a (binary little-endian) IEEE-double precision
(64-bit) number followed by two signed 4-byte integers.

    import struct

    format = "<dll"
    reclen = struct.calcsize(format)
    try:
        fileob = open("file.dat","rb")
        while 1:
            record = fileob.read(calcsize)
            if not record: break;
            myfloat, myint1, myint2 = struct.unpack(format, record)
            # now work with myfloat, myint1, myint2 ...
    finally:
        fileob.close()


Alex





From syver at NOSPAMcyberwatcher.com  Sat Dec 23 16:43:36 2000
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Sat, 23 Dec 2000 22:43:36 +0100
Subject: About WebBrowser
References: <91vbr9$eo1q2@SGI3651ef0.iddeo.es> <9235bj$6e6$1@troll.powertech.no>
Message-ID: <9236f3$6rb$1@troll.powertech.no>

"Syver Enstad" <syver at NOSPAMcyberwatcher.com> wrote in message
news:9235bj$6e6$1 at troll.powertech.no...
>
> "Juan Huertas" <jhg at galdon.com> wrote in message
> news:91vbr9$eo1q2 at SGI3651ef0.iddeo.es...
> > It's possible to print the content of a Html page using the webbrowser
> > module into a python module?
>
> I am not sure what you're thinking about here. But I'll try, if it's
> possible to print with the webbrowser COM component from the dispatch
> interface it presents it is possible to do it from Python, but is that
what
> you're asking for?
>
Okay so I had to find out, is this what you're looking for?

import win32com.client
from win32com.client import constants
browser = win32com.client.Dispatch('InternetExplorer.Application')
browser.Visible = -1
browser.Navigate('www.python.org')
browser.ExecWB(constants.OLECMDID_PRINT, constants.OLECMDEXECOPT_PROMPTUSER)

PS: Remember to use the makepy utility to generate wrapper classes and
constants for you.




From aleaxit at yahoo.com  Tue Dec 12 06:07:41 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 12 Dec 2000 12:07:41 +0100
Subject: Function points per lines of code
References: <3A35791C.5930A7FA@ndh.net>
Message-ID: <9150u402qau@news2.newsguy.com>

"Stefan Schwarzer" <s.schwarzer at ndh.net> wrote in message
news:3A35791C.5930A7FA at ndh.net...
> Hello,
>
> for an effort/schedule estimation I look for values of lines of code per
> function point for Python. I have searched Deja.com and Google.com but to
> no avail.

Starting from Lutz Prechelt's study, e.g. at
http://ncstrl.ubka.uni-karlsruhe.de:8080/Dienst/Repository/2.0/Body/ncstrl.u
bka_cs%2firatr-2000-5/text

one sees Capers Jones' table for a few languages:

language LL LOC/FP
C     3.5 91
C++   6   53
Java  6   53
Perl  15  21
Python ?  ?
Rexx   7  46
Tcl    5  64

and Prechelt sensibly assumes that the value for Tcl is
a typo (it has much higher language-level, and lower
LOC/FP, than, say, Java!).

The rest of Prechelt's study appears to strongly indicate
semantic equivalence (on his example problems) between
Python and Perl; so, taking Capers Jones' estimate as
good for the latter, it seems that around 20 LOCs/FP is
a reasonable approximation.

Having worked extensively in all of these languages, I
do NOT intuitively concur that C++ or Java are only twice
as powerful (in terms of LOC/FP) as C, nor that Rexx
takes more LOC/FP than them, and the estimate for Perl
(and Python) appear 'compressed downwards' on this scale,
too.  But then, I have NOT conducted actual counts and
measurements!  So, if my intuition tells me that, taking
C at about 100, C++ and Java must be closer to 30 (about
1:3 ratio) than to 50 (about 1:2) LOC/FP; that Rexx and
Tcl may be, oh, 25 (lower LOC/FP than C++ or Java); and
that Perl and Python must be definitely below 20 (15
sounds better); then this is just my own intuition and
experience.  Maybe it reflects the way *I* code in these
various languages, rather than the way they're 'meant'
to be used, for example.


Alex





From schliep at Octopussy.MI.Uni-Koeln.DE  Tue Dec 12 08:15:47 2000
From: schliep at Octopussy.MI.Uni-Koeln.DE (Alexander Schliep)
Date: 12 Dec 2000 14:15:47 +0100
Subject: Splash screen with Tkinter
References: <90m8oi$c3e1@news.cis.okstate.edu>
Message-ID: <tq9zoi1ixvg.fsf@Octopussy.MI.Uni-Koeln.DE>

"Fred Schroeder" <fredws_snb at ionet.net> writes:

> Could somebody tell me how to put up a splash screen with Tkinter? or if it
> is very involved tell me where to look, I have tried several searches, but
> have had no luck.
> TIA
> Fred

There is a splash screen class in Gato. Dokumentation at
http://www.zpr.Uni-Koeln.DE/~gato/Documentation/Internal/GatoDialogs.py_SplashScreen.html

You can download Gato at http://www.zpr.Uni-Koeln.DE/~gato/
Sorry I dont have  time to make it standalone.

Yours,
Alexander

-- 
Alexander Schliep                    schliep at zpr.uni-koeln.de
ZPR/ZAIK                             Tel: +49-221-470-6040 (w) 
University of Cologne                FAX: +49-221-470-5160
Weyertal 80                          http://www.zpr.uni-koeln.de/~schliep
50931 Cologne, Germany               


From fredrik at effbot.org  Tue Dec 19 11:46:56 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 19 Dec 2000 16:46:56 GMT
Subject: tkinter event binding to mouse double click
References: <91n7l6$7ij$1@news.nuri.net> <vpL%5.459$AH6.36619@newsc.telia.net> <91o31f$vi$1@news.nuri.net>
Message-ID: <4iM%5.468$AH6.37352@newsc.telia.net>

June Kim wrote:
> I'm using Win98 with Korean char set.

is this a custom version of Windows 98 (with e.g. custom
input methods, and not just another 8-bit character set)?

if so, this might help:

    root = Tk()
    root.call("tk", "useinputmethods", "1")

(if it doesn't, try asking on comp.lang.tcl.  there's nothing
wrong with your program, but it may be a problem with Tk
on your platform)

</F>




From Armin_member at newsguy.com  Fri Dec 15 07:43:57 2000
From: Armin_member at newsguy.com (Armin Steinhoff)
Date: 15 Dec 2000 04:43:57 -0800
Subject: Python for REAL Projects ?
References: <3a380d1e.409238@bendigo.news.origin.net.au>
Message-ID: <91d3md0171h@drn.newsguy.com>

In article <3a380d1e.409238 at bendigo.news.origin.net.au>, dragon at hitech.net.au
says...
>
>Hello Python People !
>I have been looking at, and trying out Python (under Win NT) and
>although Python SEEMS interesting can it "cut the mustad" on a real
>client project ??
>
>Curently, I develop using DELPHI + Interbase SQL ( or Borland BDE
>tables) because , in the main, DELPHI offers "mostly" easy to setup
>forms and access to database tables ( not perfect but usable).
[ clip .. ]
>Can this type of project be sucessfully tackled with Python + Tkinter
>+ MySQL ??

No ... but with Python + Tkinter (wxWindows) + Interbase bindings ...

Armin



From roy at panix.com  Mon Dec 25 11:36:13 2000
From: roy at panix.com (Roy Smith)
Date: Mon, 25 Dec 2000 11:36:13 -0500
Subject: Some Python 2.1 ideas
References: <925tl301ub0@news1.newsguy.com> <mailman.977590932.19335.python-list@python.org> <Cw616.23054$qN6.3308205@typhoon.we.rr.com> <3A452881.7686E717@uswest.net> <%vb16.187233$U46.5801214@news1.sttls1.wa.home.com> <3A464D5D.B541C04E@uswest.net> <mailman.977688488.25438.python-list@python.org> <20001225150250.6FA06A84F@darjeeling.zadka.site.co.il> <mailman.977760789.10224.python-list@python.org>
Message-ID: <roy-EA8C3B.11361325122000@news.panix.com>

"Bob Alexander" <bobalex at home.com> wrote:
> Finally, I'd be disappointed if the name "chomp" actually survived! I 
> just used it in my initial post as a placeholder until a better name 
> came along, since some folks would likely recall it from Perl.

We've already got the string methods strip, lstrip, and rstrip.  Why not 
just add an optional second argument which is the class of characters to 
strip.  Then chomp just becomes string.rstrip (input_string, '\n').

Even better, an optional argument to readlines to have it do the 
chomping itself.


From wsonna at attgloball.net  Sat Dec 30 17:59:48 2000
From: wsonna at attgloball.net (William Sonna)
Date: 30 Dec 2000 22:59:48 GMT
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com>
Message-ID: <05C6FUhLDNUU-pn2-tuTnzOCokxem@bill>

On Sun, 30 Dec 3900 03:00:05, henry_crun3583 at my-deja.com wrote:

> Can someone point me at an explanation of the logic for not having
> do:
> until

Its a design flaw in Python.

There is no rational reason not to have it. Anyone who has worked with
Pascal or Rexx (or even regular expressions) knows that "one iteration
plus zero or more iterations" is NOT as good as "one or more 
iterations", because the resulting code is sloppier.

So while there are workarounds for this bit of illogic, they are, 
nonetheless workarounds.



From thelocust at my-deja.com  Mon Dec  4 15:54:37 2000
From: thelocust at my-deja.com (thelocust at my-deja.com)
Date: Mon, 04 Dec 2000 20:54:37 GMT
Subject: Converting Perl Regex to Python!
Message-ID: <90h0a8$jgm$1@nnrp1.deja.com>

I'm working on porting a program from Perl to Python, and i'm to a
point where I need to convert their Perl regex stuff over to python.
Now, I understand what is going on in the Perl if statement below, but
i haven't seen to many good examples of regex stuff for Python, so i'm
hurting for an translation

The string i need to grab the info from looks like this:

MSG notify timecode [00:00:00:00]

if ($msg =~ /^MSG notify timecode \[(\d\d:\d\d:\d\d):\d\d\]/) {
    $cajun->{state_clock}=$1;
    }

So, how exactly do i convert this to python?


Sent via Deja.com http://www.deja.com/
Before you buy.


From fredrik at effbot.org  Thu Dec  7 19:53:00 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 08 Dec 2000 00:53:00 GMT
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu> <slrn92tu7t.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com> <3A2FE630.5C113AB@bioeng.ucsd.edu> <oOVX5.95$U4.1605@newsread1.prod.itd.earthlink.net>
Message-ID: <MhWX5.2529$O%.394850@newsc.telia.net>

Oldayz wrote:
> in other web searches you'd probably be even less lucky, while search on
> 'history of python' came up with self-descriptive    [40]History of:
> GreatSoftware - Python.

http://wiki.cs.uiuc.edu/GreatSoftware/HISTORY/Python ?

Did you follow that link?

Did you notice the colon in the "self-descriptive" title?

(It's a revision history for a page about Python, not a
page about Python's history).

> while a question like 'guys I'm writing a paper on python,
> can you tell me where it comes from?' shows laziness

Looks like you're pretty lazy yourself, but that didn't
stop you from posting...

</F>




From jhauser at ifm.uni-kiel.de  Thu Dec  7 18:39:45 2000
From: jhauser at ifm.uni-kiel.de (Janko Hauser)
Date: 08 Dec 2000 00:39:45 +0100
Subject: Parallel Computing
References: <3A300B1D.223630F3@bioeng.ucsd.edu>
Message-ID: <874s0fpzr2.fsf@lisboa.ifm.uni-kiel.de>

There are modules which wrap parts of MPI and PVM, both are available
from the starship or look at the vaults for a link to them. If you
want to have parallel communication between Python programs there are
many different ways, like Dopy, Pyro, CORBA, SOAP, XML-RPC...

HTH,
__Janko

-- 
  Institut fuer Meereskunde             phone: 49-431-597 3989
  Dept. Theoretical Oceanography        fax  : 49-431-565876
  Duesternbrooker Weg 20                email: jhauser at ifm.uni-kiel.de
  24105 Kiel, Germany


From andres at corrada.com  Sat Dec  2 00:11:41 2000
From: andres at corrada.com (andres at corrada.com)
Date: Sat, 2 Dec 2000 00:11:41 -0500
Subject: Binary vs. Text mode
In-Reply-To: <004901c05ad6$8496cb30$760214ac@PRODUT2KDAVE>; from Dave Brueck on Thu, Nov 30, 2000 at 07:04:55AM -0700
References: <NCBBLGDECCBLJILGIFMPCEKKCIAA.ben@semio.com> <004901c05ad6$8496cb30$760214ac@PRODUT2KDAVE>
Message-ID: <20001202001141.B13916@corrada.com>

As I have read in previous postings, this is not a feature of Python but
rather a feature of C. This 'feature' regularly bites people working on Macs
and PCs. On Unix, C's behavior is transparent but on Macs '\n' is converted
to '\r' while on Win32 '\n' is converted to '\r\n'.

The end result is that you don't have to make any extra efforts to write
portable code that will correctly parse text streams.

On Thu, Nov 30, 2000 at 07:04:55AM -0700, Dave Brueck wrote:
> Hi Ben,
> 
> On Windows the newline character ('\n') is silently replaced on-the-fly with
> '\r\n' when you read and write in text mode. This goofiness invalidates the
> numbers used by seek/tell. In binary mode (aka "not broken mode") no
> translation occurs... dumb, huh?
> 
> -Dave
> 

------------------------------------------------------
Andres Corrada-Emmanuel   Email: andres at corrada.com
------------------------------------------------------



From nas at arctrix.com  Fri Dec  8 07:35:15 2000
From: nas at arctrix.com (Neil Schemenauer)
Date: Fri, 8 Dec 2000 04:35:15 -0800
Subject: Namespace model
In-Reply-To: <200012081854.SAA03180@c.ict.om.org>; from jurjanm@c.ict.om.org on Fri, Dec 08, 2000 at 06:54:03PM -0000
References: <976202030.38510@egroups.com> <200012081854.SAA03180@c.ict.om.org>
Message-ID: <20001208043514.B2092@glacier.fnational.com>

You should look at PEP 227 and send your comments to Jeremey
Hylton (the author of the PEP).  If your interested in Python's
future then you should read the other PEPs as well.  PEP 001 is a
good place to start.

  Neil



From jarober at home.com  Fri Dec 15 20:09:13 2000
From: jarober at home.com (James A. Robertson)
Date: Sat, 16 Dec 2000 01:09:13 GMT
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <3bFZ5.382$Y2.6323@news>
Message-ID: <3A3AC0C4.8E13D4CC@home.com>

Chris Ryland wrote:
> 
> Russell--
> 
> Overall, good show! Some questions/observations.
> 
> "Russell E. Owen" <owen at astroNOJNK.washington.edu.invalid> wrote in message
> news:915qlh$89fa$1 at nntp6.u.washington.edu...
> > [...]
> > On the other hand, it is not as powerful, clean and regular as
> > smalltalk. Limitations include:
> > - The language is somewhat cluttered and irregular because it has
> > procedures, functions and methods. Hence arbitrary decisions had to be
> > made in some areas as to how to implement things, which can lead to some
> > hunting through the manuals.
> 
> But, as we all know, languages that give you an "on ramp" from other more
> familiar paradigms often succeed against "more clean/powerful" languages.
> (E.g., Java was an easy "reach" for many C++ programmers.)
> 
> I.e., Smalltalk's fierce O-O cleanliness may work against it for general
> acceptance. (That, and the fact that every Smalltalk vendor in the world
> (all 2 of them ;-) fell down and played dead at the first appearance of the
> (apparent) Java juggernaut.)

All two of them?

Cincom		VisualWorks, ObjectStudio, VSE
IBM		VAST
Gemstone	Gemstone/S
Object-Arts 	Dolphin
ObjectConnect	Smalltalk/MT
eXept		Smalltalk/X
QKS		SmalltalkAgents, SmallScript (.NET)

And Disney has Squeak open sourced

Somewhat more than 2.



--
James A. Robertson
Technical Product Manager (Smalltalk), Cincom
jarober at mail.com

<Talk Small and Carry a Big Class Library>


From larkas at eden.rutgers.edu  Wed Dec 20 16:42:16 2000
From: larkas at eden.rutgers.edu (Larry Kasoff)
Date: Wed, 20 Dec 2000 16:42:16 -0500
Subject: Wanted: Python White Papers
References: <91oj8v$i0r$1@nnrp1.deja.com>
Message-ID: <3A4127B8.E79D60F9@eden.rutgers.edu>

Redhat also uses Python. In fact, Python is installed in the base setup
on Redhat Linux.

noahspurrier at my-deja.com wrote:

> Is anyone aware of any white papers to explain the use of Python
> at any big, high-profile companies?
> I'm trying to use a "look-how-these-cool-companies-use-python"
> argument to defend the use of Python for a project. That won't be my
> only argument, of course :-) But currently Python is not winning many
> management converts here based solely on its technical merits.
> Many still say, "What's Python? Never heard of it...". It helps
> a lot when I say "Google uses it!", but it would help more if I
> knew HOW it is used and WHY.
>
> Actually, the only high-profile company that uses Python that I
> can think of is Google. Is anyone familiar with their architecture?
> How does Python fit into their design?
>
> Is there anyone out there from Google that can represent Google's use
> of Python? What are Google's plans for the future of Python in their
> architecture? I heard that they might be switching to Java.
>
> Yours,
> Noah Spurrier
>
> Sent via Deja.com
> http://www.deja.com/



From dbrueck at edgix.com  Thu Dec 14 15:53:13 2000
From: dbrueck at edgix.com (Dave Brueck)
Date: Thu, 14 Dec 2000 13:53:13 -0700
Subject: Capturing stdout from a console process?
In-Reply-To: <91b8q3$aqu$1@nnrp1.deja.com>
Message-ID: <01a901c0660f$e0d6a5c0$760214ac@PRODUT2KDAVE>

Hi Sam,

Use the os.popen* functions:

import os
w,r = os.popen4('xcopy c:*.* d: /v')
print r.read() # Prints stdout and stderr results of your command

-Dave

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Sam Schulenburg
> Sent: Thursday, December 14, 2000 12:59 PM
> To: python-list at python.org
> Subject: Capturing stdout from a console process?
> 
> 
> I am using <os.system("xcopy c:*.* d: /v")> within a python program to
> copy files from one directory to another under windows NT. When the
> program executes, a new console window is created with the results of
> the xcopy command being displayed in the console window. Is 
> their a way
> to capture this output by python so I can log error messages?
> 
> Sam Schulenburg
> 
> 
> Sent via Deja.com
> http://www.deja.com/
> -- 
> http://www.python.org/mailman/listinfo/python-list



From joconnor at mail.cybermesa.com  Mon Dec  4 13:31:54 2000
From: joconnor at mail.cybermesa.com (Jay O'Connor)
Date: Mon, 04 Dec 2000 18:31:54 GMT
Subject: Alternative Conversation
Message-ID: <200012041831.LAA09777@mail.cybermesa.com>

An embedded and charset-unspecified text was scrubbed...
Name: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20001204/9aab0792/attachment.ksh>

From fredrik at effbot.org  Tue Dec 19 15:08:02 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 19 Dec 2000 20:08:02 GMT
Subject: Python in "math through programming" curriculum
References: <2g2t3t06r2agrrjmfnr41eu4lc2u39a8qi@4ax.com> <3A3EFF47.67DE69CF@gte.net> <513v3toslkjkggh9oumlf20qc9bi1gdtif@4ax.com> <3A3FC677.FDB19937@gte.net>
Message-ID: <CeP%5.521$AH6.45315@newsc.telia.net>

Clifford J. Nelson wrote:
> I'm sure I don't attach ANY weight to any of your opinions.
> I've seen your web sites about Synergetics and programming!

clifford, please keep your personal problems off
comp.lang.python.

</F>




From sholden at holdenweb.com  Wed Dec 20 14:10:39 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 20 Dec 2000 14:10:39 -0500
Subject: dynamic function call
References: <Ko606.141$G66.2664@vixen.cso.uiuc.edu>
Message-ID: <Fw706.1805$WA2.28069@e420r-atl2.usenetserver.com>

Hwanjo Yu <hwanjoyu at uiuc.edu> wrote in message
news:Ko606.141$G66.2664 at vixen.cso.uiuc.edu...
> Hi,
>
> Is there any way to bind a function name at runtime ?
> For instance, we don't know the function name to call, but we know the
> function name is saved in a variable, "fname".
> How to call it in this case ?
> Thanks in advance.
>
You just use the variable fname instead of the function name!  For example:

>>> def myfun(x):
...  print "x is", x
...
>>> yourfun = myfun
>>> yourfun('This is cool')
x is This is cool

The "def" statement simply binds the function name to its compiled code
body.  If you "assign" the functin to another variable, the same call
mechanism can be used on that binding too.

regards
 Steve






From trotts at llnl.gov  Sun Dec 31 16:41:21 2000
From: trotts at llnl.gov (Issac Trotts)
Date: Sun, 31 Dec 2000 13:41:21 -0800
Subject: How can I conveniently print an expression?
In-Reply-To: <3A4F74AC.25FB45D8@mail.com>; from jdries@mail.com on Sun, Dec 31, 2000 at 07:02:20PM +0100
References: <20001230231017.A30045@llnl.gov> <3A4F74AC.25FB45D8@mail.com>
Message-ID: <20001231134121.A4884@llnl.gov>

Thanks for the clever trick.  It works!

Issac


On Sun, Dec 31, 2000 at 07:02:20PM +0100, Jan Dries wrote:
> 
> 
> Issac Trotts wrote:
> > 
> > In Python, one might naively try something like
> > 
> > def printexpr(expr_string):
> >   expr = eval(expr_string)
> >   print expr_string+' = '+`expr`
> > 
> > which would be used to say things like
> > 
> > printexpr('len(sys.argv)')
> > 
> > but of course this does not always work because when Python executes
> > eval(expr_string), it assumes that the elements of the expression
> > specified by expr_string are to be found in the global scope. 
> >
> > If there were some mechanism that allowed us to access the variables
> > of a calling function from within a called function, then the problem
> > would be solved.  Does anyone know of such a mechanism?
> 
> I once came up with a solution for this. The following should do the
> trick for you:
> 
> from sys import exc_info
> 
> def caller_symbols():
>     try:
>         raise StandardError
>     except StandardError:
>         t = exc_info()[2].tb_frame
>         return (t.f_back.f_back.f_globals,t.f_back.f_back.f_locals)
> 
> def printexpr(expr_string):
>     caller_globals,caller_locals = caller_symbols()
>     expr = eval(expr_string,caller_globals,caller_locals)
>     print expr_string+' = '+`expr`
> 
> 
> Regards,
> Jan



From loewis at informatik.hu-berlin.de  Sat Dec  9 05:52:47 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 09 Dec 2000 11:52:47 +0100
Subject: reloading modules question
References: <90scdf$hpe$1@troll.powertech.no>
Message-ID: <p6qsnnxamtc.fsf@informatik.hu-berlin.de>

"Syver Enstad" <syver at NOSPAMcyberwatcher.com> writes:

> Is it safe to loop through sys.modules.items()
> reloading each module, as long as I catch the ImportError?

What do you mean by "safe"? Python will continue to work, if that is
what you mean.

> Am I guaranteed to have all python modules reloaded if I do the loop twice?

No. Reloading a builtin module does not do anything.

> The reason I am asking is that IIS webserver keeps Python in memory
> so that it won't use the new versions of python modules I deploy if
> don't stop and start the webserver. I can't afford to stop the
> webserver all the time, so I am considering ways to reload the
> python modules while it's running.

Which of your modules require reloading when they are reused? I
recommend to redesign the modules so that they act as libraries,
instead of keeping session state.

Regards,
Martin


From wware at world.std.com  Thu Dec  7 21:01:51 2000
From: wware at world.std.com (Will Ware)
Date: Fri, 8 Dec 2000 02:01:51 GMT
Subject: Music synthesizer in Python/C
Message-ID: <G588B3.6A5@world.std.com>

http://www.deja.com/getdoc.xp?AN=702341030&fmt=text
I have been tinkering with a few iterations of a software emulation
of a music synthesizer (having tinkered with the hardware versions
in the 70s-80s). I had previously tried direct discrete-time
implementation of the various synthesizer modules that used to be
popular: VCOs, VCAs, filters, envelope generators... This time around,
discrete-time-sampled signals have been replaced by lists of coefficients
for complex exponentials used to generate sums of sinusoids, so all
the schlepping and processing is done on the coefficients. Still plenty
of work to do on it, but it's already amusing to play with.
-- 
# - - - - - - - - - - - - - - - - - - - - - - - -
# Resistance is futile. Capacitance is efficacious.
# Will Ware	email:    wware @ world.std.com


From xmnemonic at aol.com  Mon Dec 25 03:18:52 2000
From: xmnemonic at aol.com (Xmnemonic)
Date: 25 Dec 2000 08:18:52 GMT
Subject: ready for C or C++?
Message-ID: <20001225031852.04296.00003766@ng-fp1.aol.com>

If I've learned Python and become proficient with it, am I ready to learn C or
C++?


From aahz at panix.com  Sat Dec 30 12:10:15 2000
From: aahz at panix.com (Aahz Maruch)
Date: 30 Dec 2000 09:10:15 -0800
Subject: Lazy Python usage (hack request)
References: <Pine.LNX.4.10.10012301443330.13756-100000@kev.ioc.ee>
Message-ID: <92l4tn$raf$1@panix3.panix.com>

In article <Pine.LNX.4.10.10012301443330.13756-100000 at kev.ioc.ee>,
Pearu Peterson  <pearu at ioc.ee> wrote:
>
>Is it possible to achive the following behavior in Python?
>
>>>> import lazy
>>>> a = b
>No name 'b' defined, using default 777
>>>> print a
>777
>
>Note that changing the python session above is not an option. (Also 'No'
>for the answer;-)

Well, if you're going to stick to it not being an option to make any
changes, the answer is "no".  However, you can make a little hack that
*does* work:

>>> from lazy import lazy
>>> a = lazy.b
No name 'b' defined, using default 777
>>> print a
777

Inside lazy.py, you define a class called lazy with a __getitem__()
method.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

Ninth Virtual Anniversary: Tomorrow!


From pyth at relax.dk  Fri Dec 15 10:12:33 2000
From: pyth at relax.dk (pyth at relax.dk)
Date: Fri, 15 Dec 2000 16:12:33 +0100 (CET)
Subject: Python for REAL Projects ?
In-Reply-To: <3eq_5.5732$n51.113513@e420r-atl1.usenetserver.com>
Message-ID: <Pine.LNX.4.21.0012151601390.13213-100000@relax.dk>

[snip]
> I know that many systems have been built around MySQL, and I'll probably get
> flamed for saying this by MySQL advocates, but my impression is that MySQL
> is a toy system which will not compete in a true multi-user production
> environment with commercial products. Not that there's anything wrong with
> that, and if you accept its limitations MySQL can be effective.
[snip]

I am a developer on a system that uses a mysql db, and have this to say
about your comment :o)

MySQL is fast and reliable, as long as your are not storing large chunks
of data, don't have the need for transactions support, though they have
made something that almost works as transactions, and in many cases,
it's not a good idea to use MySQL in a "multi-db-access" environment,
unless you really know what you're doing, and makes sure nothing can go
wrong, and that is usually the problem.... the program/scripts has to do
the checking of db-access :o/

So... You're right... We need real transactions!
But other than that, MySQL is REALLY fast and reliable... for small chunks
of data :o)

/Dan Larsen




From smnordby at yahoo.com  Mon Dec 25 09:40:03 2000
From: smnordby at yahoo.com (Steve)
Date: Mon, 25 Dec 2000 06:40:03 -0800
Subject: ready for C or C++?
References: <20001225031852.04296.00003766@ng-fp1.aol.com>
Message-ID: <3A475C43.A82D059@yahoo.com>

Xmnemonic wrote:
> 
> If I've learned Python and become proficient with it, am I ready to learn C or
> C++?

I began playing with Python about a year ago and just started learning C
for the heck of it.  It's going well so far, but its lower level and you
have to learn and get comfortable with pointers.  Other than that, my
only trouble is forgetting {}'s and ;'s (Python spoils you).

-Steve-


From erno-news at erno.iki.fi  Sun Dec 10 21:44:47 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 11 Dec 2000 04:44:47 +0200
Subject: Multiple Threads & IRC
References: <3A3415E0.1080607@bellsouth.net>
Message-ID: <kud7ez8yn4.fsf@lasipalatsi.fi>

In article <3A3415E0.1080607 at bellsouth.net>, Wilbur
<willybur at bellsouth.net> writes:

| It seems that I will need multiple threads running for this client.  One 
| thread will have to be constantly checking my socket to see if the 
| server has sent anything.

it is probably easier to use the Tkinter.tkinter.createfilehandler().
that will let you pass in a callback function that will get called
once there is input on a socket.

| The server sends data in packets, and it sends whenever it can.  How can 
| I get my client to be able to receive data from the server without 
| knowing when it will send and what amount of data will be sent?  I only 
| know how to tell Python to wait until it receives, say, 1024 bytes.  If 
| the server doesn't send any more, how can I tell Python to stop waiting? 

you have to employ some buffering to make the pieces of data arriving
from the server break into lines nicely. something like
(untested, off the top of my head)

def input_callback(self, sock):
        data = sock.read(1024)
        # if there's more than 1024 bytes of data coming,
        # we'll just get called again.
        if not data:
                # eof, handle...
        else:      
                self.buf = self.buf + data
                while '\n' in self.buf:
                        line, self.buf = string.split(self.buf, '\n', 1)
                        self.handle_line(line)



you could do it with threads too, i think, but i'm not familiar with
tkinter <-> threads interaction.

  -- erno


From moshez at zadka.site.co.il  Sun Dec 31 10:21:31 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Sun, 31 Dec 2000 17:21:31 +0200 (IST)
Subject: how do i check if a file exists with python?
In-Reply-To: <92hi3h01f0d@news2.newsguy.com>
References: <92hi3h01f0d@news2.newsguy.com>, <RlM26.98766$yR4.2639444@news1.rdc1.tx.home.com> <phU26.31$KD3.4767@typhoon.aracnet.com>
Message-ID: <20001231152131.7CECDA84F@darjeeling.zadka.site.co.il>

On Fri, 29 Dec 2000 09:30:38 +0100, "Alex Martelli" <aleaxit at yahoo.com> wrote:

> When somebody asks how to determine whether a file exists, it IS quite
> likely that what they mean is something different
...
> It's unlikely, although of course possible, that 'actual existence'
> is exactly what one needs to determine

In that case, on a typical OS (even Windows-like, although it's more
pronounced on UNIX) one *can't* know the answer, in general. Consider
stuff like unreadable directories, for example.

files-are-like-schroedinger-cat-so-you-better-be-ready-for-anything-
when-you-open-the-box-ly y'rs, Z.
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From jrfsousaREMOVE00592425 at esoterica.pt  Fri Dec 22 19:49:31 2000
From: jrfsousaREMOVE00592425 at esoterica.pt (José Rui Faustino de Sousa)
Date: Sat, 23 Dec 2000 00:49:31 -0000
Subject: Refreshall from Python 1.3
References: <977505346.331315@osiris.esoterica.pt> <XeM06.7892$3O1.86960@e420r-atl1.usenetserver.com>
Message-ID: <977532570.388660@osiris.esoterica.pt>

"Steve Holden" <sholden at holdenweb.com> wrote in message news:XeM06.7892
> Hope this helps your wallet!
>

It does ;-)

Best regards
Jos? Rui


--
========================================================================
he influence of Darwin, or more broadly of the concept of evolution, has
had effects more truly profound [than other aspects of civilization]. It
has literally led us into a different world. - George Gaylord Simpson.
========================================================================
Iam://Jos? Rui Faustino de Sousa http://homepage.esoterica.pt/~jrfsousa/
mailto://jrfsousaREMOVE00592425 at esoterica.pt phone://+351-239444940
address://rua Carlos A. Pinto de Abreu n? 30C, 1? 3040 Coimbra Portugal
========================================================================
Real Programmers do sig blocks in Fortran 95.
========================================================================





From root at rainerdeyke.com  Thu Dec 14 19:51:35 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Fri, 15 Dec 2000 00:51:35 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <r0pzoi11lhz.fsf@black.cs.washington.edu> <5jvZ5.3588$Kd1.418770@newsb.telia.net> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com> <kuu28971kp.fsf@lasipalatsi.fi> <3A380E4E.8537A0FB@san.rr.com> <91a1vr02b02@news2.newsguy.com> <Co9_5.25465$x6.13756579@news2.rdc2.tx.home.com> <91bbq102mtk@news1.newsguy.com>
Message-ID: <rWd_5.26820$x6.14119694@news2.rdc2.tx.home.com>

"Alex Martelli" <aleaxit at yahoo.com> wrote in message
news:91bbq102mtk at news1.newsguy.com...
> "Rainer Deyke" <root at rainerdeyke.com> wrote in message
> news:Co9_5.25465$x6.13756579 at news2.rdc2.tx.home.com...
> > "Alex Martelli" <aleaxit at yahoo.com> wrote in message
> > news:91a1vr02b02 at news2.newsguy.com...
> > > There is no value *IN* x.
> > >
> > > A variable *REFERS TO* 'a value' (aka 'an object').
> >
> > If somebody really wants mutable integer objects, they're not that hard
to
> > create.
> >
> > class MutableInt(UserInt.UserInt):
> >   def set(self, value):
> >     self.data = value
>
> UserInt is not in the Python 2 docs, though it's nice to hear it
exists:-).

Actually it doesn't, but it's trivial (if tedious) to implement.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From bk at xk7.com  Thu Dec 21 11:42:03 2000
From: bk at xk7.com (Burkhard Kloss)
Date: Thu, 21 Dec 2000 16:42:03 -0000
Subject: Replay commands
References: <91rep9$tev$1@nnrp1.deja.com>
Message-ID: <977417475.798911@master.nyc.kbcfp.com>

> When using Python interactively, I need to save my command history
> for editing and reuse. Is that possible?
If you are using a Unix style platform, you can easily do that via the
script command.





From aleaxit at yahoo.com  Fri Dec 29 18:49:28 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 30 Dec 2000 00:49:28 +0100
Subject: 2001 Enchancement Wishlist
References: <3A4CEACB.220998A0@javanet.com> <Ro736.2511$Qb7.273102@newsb.telia.net>
Message-ID: <92j83r13gh@news1.newsguy.com>

"Fredrik Lundh" <fredrik at effbot.org> wrote in message
news:Ro736.2511$Qb7.273102 at newsb.telia.net...
    [snip]
> (fwiw, singleton classes are pretty unpythonic too.  python

Amen, Hallelujah.

> already has a framework for creating exactly one instance of
> something: the module namespace).

Pity that isn't as flexible as an object-instance -- can't define
special methods to determine how it will print, how it will
let usercode loop on its items, etc.

Lightweight-class instances, delegating all state to their
module, can implement whatever special methods may be
handy, and still avoid any need for the Singleton design
pattern as per GoF (my least-favourite GoF DP).


Alex





From gregj at pobox.com  Tue Dec 19 01:43:12 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Tue, 19 Dec 2000 06:43:12 GMT
Subject: Super Database Tool in Python for Ever !!!
References: <t3satn4tam43f9@corp.supernews.com>
Message-ID: <4sD%5.166657$U46.5343892@news1.sttls1.wa.home.com>

sounds like something idealab! would be interested in. www.idealab.com.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From samschul at pacbell.net  Tue Dec  5 17:29:57 2000
From: samschul at pacbell.net (Sam Schulenburg)
Date: Tue, 05 Dec 2000 22:29:57 GMT
Subject: Tkinter frame items?
References: <90jlgp$okn$1@nnrp1.deja.com>
Message-ID: <90jq8q$t03$1@nnrp1.deja.com>

Thanks for the help Mattthew,and Fredrik. As usual If I had read the
whole manual I wold have found it. Your quick responce has saved me a
lot of time. It always supprises me how simple a solution to GUI
programing is under Python versis Windows C++ and MFC.

Sam


Sent via Deja.com http://www.deja.com/
Before you buy.


From mfletch at tpresence.com  Tue Dec  5 04:17:05 2000
From: mfletch at tpresence.com (Mike Fletcher)
Date: Tue, 5 Dec 2000 04:17:05 -0500 
Subject: How do packages and .pth files work under Windows (Py 2.0)
Message-ID: <CC59868A0A76D311B1F50090277B321C1F3713@VRTMAIL>

Okay, here is the basic deal. You want to make your directory look something
like this...

SomeDirectoryOnPythonPath
	mx
		__init__.py
		TextTools
			__init__.py

You _don't_ want mx to be in your Python path if you want to do
	from mx import TextTools

To see what your current Python path is...

p:\>d:\bin\lang\py2\python
Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
>>> import sys
>>> sys.path
['', 'd:\\bin\\lang\\py2\\pythonwin', 'd:\\bin\\lang\\py2\\win32',
'd:\\bin\\lan
g\\py2\\win32\\lib', 'd:\\bin\\lang\\py2', 'd:\\bin\\lang\\py2\\dlls',
'd:\\bin\
\lang\\py2\\lib', 'd:\\bin\\lang\\py2\\lib\\plat-win',
'd:\\bin\\lang\\py2\\lib\
\lib-tk', 'd:\\bin\\lang\\py2\\numeric', 'd:\\bin\\lang\\py2\\pil']
>>>

As you can see, I don't have lib/site-packages in my Python path (or even on
my disk, for that matter).  For what most people seem to do on Win32 (to the
best of my knowledge), is to use the base Python directory for all
"well-behaved package-structured" modules.  That is, my directory structure
looks something like (for my 1.5.2 version of TextTools):

python
	mx
		TextTools

And without any .pth files, that works fine.  Similarly, had I put the same
structure in:

python
	lib
		mx
			TextTools

(lib being on the Python path), it should also work, again, without the use
of .pth files.  That's actually the nice thing about properly packaged
systems, you don't need to muck around with reconfiguring your Python path,
you just drop them into an existing directory.

Hope this helps,
Mike



-----Original Message-----
From: Jonathan Gilligan [mailto:jonathan.gilligan at vanderbilt.edu]
Sent: Tuesday, December 05, 2000 2:20 AM
To: python-list at python.org
Subject: How do packages and .pth files work under Windows (Py 2.0)


I am having problems understanding how .pth files are supposed to work
under windows. The Python 2.0 documentation has several inconsistent
descriptions.

In one place, it says that under Windows and Mac, .pth files are only
interpreted if they are in the python root. Elsewhere the
documentation indicates that they are interpreted if they are in the
lib\site-packages directory.

Here is the basic problem. I have packages bar, far, and gar, all of
which I want to be subpackages in the package foo, so I can write
"import foo.bar as FB". If bar is a working package, the Python
documentation suggests that II should be able to do what I want if I
put the directory bar (containing the bar package, which has a valid
__init__.py) inside a directory "site-packages\foo" and put a file
"foo.pth" in  the site-packages directory with the single line "foo".

However, this does not work for me. Somehow, Python does not seem to
understand the foo.pth file, and python complains "ImportError: No
module named foo"

Here are the details:

I am trying to get the mx tools installed on my computer and have
python 2.0 built from cvs sources in D:\Sources\python\python\dist\src
and am trying to get mxTextTools installed somewhere so that I can use
the following line (from the Martel genome parsing package):

    from mx import TextTools as TT

I have installed mxTextTools in
D:\Sources\Python\Python\dist\src\lib\site-packages\mx\TextTools. If I
set PYTHONPATH to
"D:\Sources\\Python\Python\dist\src\lib\site-packages\mx" then I can
"import TextTools" but if I set PYTHONPATH to
"D:\Sources\Python\Python\dist\src\lib\site-packages" then I cannot
"import mx.TextTools". I keep getting an error message:

    "ImportError: No module named mx"

This occurs even if I create a "mx.pth" file in "site-packages" with
the single line "mx". I have also tried putting the "mx" directory and
the "mx.pth" file in the python root
(D:\Sources\python\python\dist\src).

I have the registry correctly set up:

"HKLM\Software\Python\PythonCore\2.0" has subkeys

"InstallPath" with default value "D:\Sources\Python\Python\dist\src"
"PythonPath" with default value
"D:\Sources\Python\Python\dist\src\lib;D:\Sources\Python\Python\dist\s
rc\lib\site-packages;d:\Sources\python\python\dist\src\pcbuild;D:\Sour
ces\Python"

Can anyone help me figure out how to get packages to work in the way I
am trying to?

It would help me enormously if there were a way to get Python to print
out a list of where it looks when it tries to resolve an "import". Is
this possible?

Thanks for any help that anyone can give me.

Jonathan Gilligan


-- 
http://www.python.org/mailman/listinfo/python-list



From zeitlin at seth.lpthe.jussieu.fr  Sat Dec  9 19:41:19 2000
From: zeitlin at seth.lpthe.jussieu.fr (Vadim Zeitlin)
Date: 10 Dec 2000 00:41:19 GMT
Subject: wxPython & floating point numbers ?
References: <m2puj18q81.fsf@hobbes.dyn.dhs.org>
Message-ID: <slrn935k9f.6os.zeitlin@seth.lpthe.jussieu.fr>

On 09 Dec 2000 18:22:06 +0100, Ueli Schlaepfer <u_schlaepfer at bluewin.ch> wrote:
>Hi everyone,
>
>when using the otherwise excellent wxPython package, I observe the
>following strange behaviour:
>
>>>> print 0.1
>0.1
>>>> import wxPython.wx
>>>> print 0.1
>0.0
...
>Has anybody seen the same error?

 No, but I've seen a similar one which was due to the fact that GDK (used by
GTK+ used by wxGTK used by wxPython) changes the current locale by itself.
And if in the new locale the decimal separator symbol is not period any more
this might cause problems (although I still fail to see how can this lead to
the result above to be honest).

 To check if this guess is correct, just unset LANG and all LC_XXX environment
variables and repeat the test above.

>What is the solution, if there is any?

 If the problem is due to this, just do "LANG=C python" instead of just
"python" (assuming you use a ksh-derived shell).

 Regards,
VZ

-- 
GCS/GM d? H+ s++:-- p2 au--- a- w+ v C+++ UBLS+++ P- L++ N++ E--- W++++ M? V--
-po+ R++ G`` !tv b+++ D--- e++++ u++ h--- f+ r++ n- y?


From tcmay at got.net  Mon Dec 18 14:06:54 2000
From: tcmay at got.net (Tim May)
Date: Mon, 18 Dec 2000 11:06:54 -0800
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <3a36c85e.4409136@News.CIS.DFN.DE> <owen_nospam-1312001131590001@cs204-90.fsmodem.washington.edu> <3A37DD31.C9A5F18C@pacbell.net> <91bhlh$c64u$1@nntp6.u.washington.edu> <3A395197.CB2974D5@home.com> <L_r_5.7453$n51.136546@e420r-atl1.usenetserver.com> <steve976915950@serf.org> <tcmay-E53C56.15191615122000@news.got.net> <qEs%5.39707$n51.728728@e420r-atl1.usenetserver.com>
Message-ID: <tcmay-FD4431.11065418122000@news.got.net>

In article <qEs%5.39707$n51.728728 at e420r-atl1.usenetserver.com>, "Steve 
Holden" <sholden at holdenweb.com> wrote:

> Tim May <tcmay at got.net> wrote in message
> news:tcmay-E53C56.15191615122000 at news.got.net...
> > In article <steve976915950 at serf.org>, steve at serf.org (A SERFer) wrote:
> >
> >
> > > Indeed we are!  We keep hearing a circular argument:
> > > 1. Cincom are committed to Smalltalk in the long term.
> > > 2. We must charge these prices if Smalltalk is to survive within 
> > > Cincom.
> > >
> > > Its like a sort of quantum mechanical uncertainty thing... each
> statement
> > > on its own makes sense, but put together, they are meaningless.
> >
> > Really? Translate it to this:
> >
> > 1. Intel is committed to the Pentium in the long term.
> > 2. We must charge these prices if the Pentium is to survive within 
> > Intel.
> >
> In fact, this is quite different from what Intel say.  They are fairly
> openly charging what the market will bear, more or less along the lines 
> of
> (John Erlichman's?) "When you've got them by the balls, their hearts and
> minds will surely follow.  Intel claim to be maximizing shareholder 
> return
> by setting their pricing at the level which optimizes that parameter.

Precisely my point. That is, "what the market will bear."


> > Or translate it to any number of other softare tools, such as chip
> > layout software, PCB software, design rule checking tools. Or factory
> > management tools. Or data bases.
> >
> > 1. Oracle is committed to its database products in the long term.
> > 2. We much charge these prices if these products are to survive withing
> > Oracle.
> >
> Again the same philosophy would seem to be in place at Oracle, who have
> never claimed to me (not that I have much to do with them) that their 
> prices
> *NEED* to be that high.  It's just that demand for the products is so 
> high
> they can afford to ignore those who do not choose (or cannot afford) tp 
> pay
> them.

Again, precisely my point.

Cincom charges what it charges because it _can_. If enough purchasers 
think the price is too high, and they don't buy, then sales will decline.

Arguments that some price is "too high" rarely matter...what matters is 
whether sales continue.

I cited Intel and Oracle to illustrate the point. 


--Tim May

-- 
Timothy C. May         tcmay at got.net        Corralitos, California
Political: Co-founder Cypherpunks/crypto anarchy/Cyphernomicon
Technical: physics/soft errors/Smalltalk/Squeak/agents/games/Go
Personal: b.1951/UCSB/Intel '74-'86/retired/investor/motorcycles/guns




From daniel at cs.washington.edu  Tue Dec 12 14:35:04 2000
From: daniel at cs.washington.edu (Daniel Wood)
Date: Tue, 12 Dec 2000 19:35:04 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl>
Message-ID: <r0pzoi11lhz.fsf@black.cs.washington.edu>

I'm glad this discussion is happening because I've been learning
python and I'm worried that the bloom is coming off of our romance.
Perhaps if I can be more understanding, our relationship can survive
. . .

scarblac at pino.selwerd.nl (Remco Gerlich) writes:

> Integers, longs, complex numbers, strings, functions, tuples, types, floats
> and None are immutable. Most of these are obvious; the number 3 is the
> number 3, why would you change that?!

Here is what I wanted to do.  I have some random parameters, lambda
and beta.  My calculation uses them like so:

plot( lambda, beta )

I'm using pygtk as a widget set.  I have some sliders.  I wanted to
make one of the sliders control lambda and one control beta.  The
control should be scaled by some parameter etc, etc.  Now I could have
a function 

def lambdaCB( val ):
 global lambda
 lambda = val / constant

but instead I thought that I would have a simple class

class ScaleCB:
  def __init__( self, variable, scale ):
    self.variable = variable
    self.scale = scale

  def __call__( self, val )
    self.variable = val / self.scale

(then, I can do lots of parameters with nice simple code instead of
bunches of stupid helper functions.

This obviously didn't work.

I *think* that the "pythonic" sol'n appears to be lambda = [ value ]
or some such.  Clearly, one person's hack is another person's idiom,
but I can't convince myself that this doesn't suck.  It seems like
what I want is a better idiom for a "mutable integer".  I guess it
would be something just like an int but with a set method.

(Excessive detail here: I'm using swig and in fact these variables are
part of swig's cvar interface.  The cvar interface appears to use some
sort of __getattr__ and __setattr__ hooks in a very un-pythonic way.
I'm not sure but I *think* that this is because SWIG is trying to
expose mutable floats, ints, etc.  It basically binds a NAME
(e.g. cvar.myvar) to a C language thing instead of some sort of object
living in cvar.myvar that corresponds to a C language thing.  I
believe that the SWIG notes mentions something about this and
basically SWIG wants you to be able to do "cvar.myvar = 10" and get
the C style effect.  This has the sad effect that you can't take the C
language bindings and use them in any way other than with the official
NAME "cvar.etcetc".)

Clues gratefully accepted,
Daniel.


From jessw at loop.com  Tue Dec 19 12:53:13 2000
From: jessw at loop.com (Jesse W)
Date: Tue, 19 Dec 2000 09:53:13 -0800
Subject: Gif Transparencies in Pil
Message-ID: <3A3FA089.FE1@loop.com>

How does the transparency support work for Gif files using Pil?  I
looked in the handbook and searched the website, but I only came up with
hints about having the 'transparency' attribute in the info dictionary. 
Does anyone have any help with this?

				Thank you for your time,
						Jesse W


From neelk at alum.mit.edu  Thu Dec 21 19:35:35 2000
From: neelk at alum.mit.edu (Neelakantan Krishnaswami)
Date: 22 Dec 2000 00:35:35 GMT
Subject: 2 question from a python newbie!
References: <91tl43$rko$1@neptunium.btinternet.com> <slrn944q0t.919.scarblac@pino.selwerd.nl> <91tq58$cq6$1@uranium.btinternet.com>
Message-ID: <slrn945c6l.p2p.neelk@alum.mit.edu>

On Thu, 21 Dec 2000 20:45:18 -0000, G. Willoughby <sab at NOSPAM.freeuk.com>
wrote:
>
> I'm not a total programming newbie just a python newbie  :)
> I'll check out that site and them, thar books.

In that case, I really recommend the book _Learning Python_ from
O'Reilly. It's very good. Also, ask questions on this newsgroup when
you hit something confusing. There are a great many knowledgeable
people who are willing to explain the various bits of Python arcana
on this ng.


Neel


From jasonic at nomadicsltd.com  Sun Dec 31 12:27:06 2000
From: jasonic at nomadicsltd.com (Jason Cunliffe)
Date: Sun, 31 Dec 2000 12:27:06 -0500
Subject: ML & Python question
Message-ID: <t4uqtfi3fv7sf2@corp.supernews.com>

A very well respected programmer friend was yesterday trying to tell me
about SML [ML].
Part of his encouragment to me was simply to go deeper into programming
abstraction and develop better [safer] techniques.

I argued feebly that although I love studying programming, I am effectively
an incompetent novice programmer - and not too confident in this lifetime
that I will attain great fluency. Which is why I love Python - it
immediately makes me sane and comfortable. It's fun and gives me hope that I
and will seriously improve over time, with practice and concentration.

He mentioned problems which languages like ML avoid. I replied that my
impression was that Python seems to avoid a lot of confusing code and that
others appeared to love it for its inherent clarity etc..  Among the many
reasons I gave for liking Python, this news group and the quality of its
community are high on my list.

>>> Please, Can anyone here provide a intro to what SML is and how Python
relates to it?

This is really not intended as Python vs. flame bait question - I am
interested to expand my progamming horizons and am hoping people here might
help me to do so.

Happy New Year and best wishes in 2001
+ thanks very much to you all for contributing so wonderfully to cpl.

- Jason
___________________________________________________________
Jason CUNLIFFE = NOMADICS['Interactive Art and Technology']




From max at alcyone.com  Tue Dec 12 12:04:41 2000
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 12 Dec 2000 09:04:41 -0800
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <29lZ5.138780$U46.4638547@news1.sttls1.wa.home.com>
Message-ID: <3A365AA9.4C02CA46@alcyone.com>

Greg Jorgensen wrote:

> The C/C++ loop idioms are:
> 
>   for (; ;) { ... }
> 
> and less common:
> 
>   while (1) { ... }

Not to start another pointless argument, I've seen while (1) [or while
(true) in C++] to be a much more common construct than for (;;).  Both
work perfectly well, of course.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ An undevout astronomer is mad.
\__/ Edward Young
    Alcyone Systems / http://www.alcyone.com/
 Alcyone Systems, San Jose, California.


From smoriano at sicon.net  Mon Dec 11 12:10:01 2000
From: smoriano at sicon.net (Sergio Boix Moriano)
Date: Mon, 11 Dec 2000 18:10:01 +0100
Subject: A Newbie Question: How To Call Other Python Programs From A Main 
 Python Program.
References: <20001211162202.CE9E81CEE7@dinsdale.python.org>
Message-ID: <3A350A69.AE65867@sicon.net>

             A Newbie Question: How To Call Other Python Programs From A Main Python Program.
        Date:
             Mon, 11 Dec 2000 16:01:52 GMT
       From:
             jbranthoover at my-deja.com
 Organization:
             Deja.com - Before you buy.
         To:
             python-list at python.org
 Newsgroups:
             comp.lang.python



Hello All,
        I have written several small Python programs.  These programs
are not just functions but complete programs with their own menuing
systems.

        My question is,  how do I (and is it a good idea) call these
independent Python programs from one main program that selects which
program to run from a main menu?  I tried to use the import command.
This seems to work OK except that the programs only run once.  If the
same program is selected from the menu,  nothing happens.  It seems to
return immediately without running the program.

        Any information that you can give me will be greatly
appreciated.  Thank you for your time.

      If you?re working under windowsx, nt and  sample code will be

            import win32con, win32api
            guion='executable arguments'
            #    The executable will be either the program or the python executable
            #  with  the .py file as an argument. Tis is:
            # guion='mypy.exe args'
            # or guion='python.exe mypy.py args'

           show=win32con.SW_SHOWNORMAL
           cad1 = win32api.WinExec(guion, show)

        You can use this way

                    import os
                    os.popen('executable','r')

        With this code you have a multiplattform code (it works in winblows, linux, mac, etc)

    Regards




From aleaxit at yahoo.com  Mon Dec 25 19:08:55 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 26 Dec 2000 01:08:55 +0100
Subject: classes (was Re: Same again please for OOP)
References: <92708d0g4g@news1.newsguy.com>, <mailman.977623149.8184.python-list@python.org> <9257mf0sl1@news1.newsguy.com> <roy-E5B925.13324224122000@news.panix.com> <925khi019hs@news1.newsguy.com> <kubsu1t81x.fsf@lasipalatsi.fi> <mailman.977739488.28224.python-list@python.org>
Message-ID: <928nls02k55@news1.newsguy.com>

"Moshe Zadka" <moshez at zadka.site.co.il> wrote in message
news:mailman.977739488.28224.python-list at python.org...
    [snip]
> I've learned the hard way not to write __getattr__ methods. Currently,
none
> of my code contains them. True gurus, like Gordon McMillan, can use
__getattr__
> (e.g., in filecmp.py), to get lazy evaluation of attributes, but I tend to
> fear those things.

I'm no guru (not a _Python_ one, I mean; at COM and C++, I'm somewhat
more experienced) yet I've happily used __getattr__ for some lazy evaluation
tasks since quite early on.  Actually, __getattr__ (for lazy evaluation, as
opposed to implicit delegation, which is another cool use) tends to be a
single,
simple thingy that delegates attribute foo to __get_foo__, attribute bar
to __get_bar__, and so on, and comes from a mixin.


> The "computed attributes" interface needs to be improved befroe Python can
> be said to be not one of *those* languages.

I wouldn't mind getting the __get_foo__ delegation automatically (though
using Boost Python gets me that, I don't _always_ have it around... yet:-),
and I'd love a simpler and more explicit way to express implicit and
automatic delegation of certain attributes to some sub-object, but, with
some naming-conventions and some luck, I haven't been bitten yet (not
in production code -- halfway decent unit tests would show up any silly
error in this area in a prototype, I should hope).

I dearly hope the special-casing of stuff now handled from __getattr__
doesn't get out of hand -- with complex structures and whatever.  I
think 'officializing' the specific-attribute getter/setter naming convention
might solve 80% of the cases at little conceptual cost, and something
like a sequence of (directory, delegate) pairs somewhat parallel to the
current __class__.__bases__ (but, per-object, and with 'filtered'
delegation, rather than 'delegate whatever the delegate can handle')
might add a further 10+%, relegating 'bare __getattr__' to very rare
needs indeed.  I wouldn't mind as a Python user, I'd rejoice as a
teacher of Python to beginners:-).

But __getattr__ ain't too bad... with suitable mixins, I can have even
beginners using it quite productively without real risks of accidents!-)

(__setattr__, now, THAT is admittedly something else again!-)


Alex





From rturpin at my-deja.com  Sat Dec 23 17:07:12 2000
From: rturpin at my-deja.com (rturpin at my-deja.com)
Date: Sat, 23 Dec 2000 22:07:12 GMT
Subject: One Python 2.1 idea
References: <mailman.977590932.19335.python-list@python.org>
Message-ID: <9237md$orm$1@nnrp1.deja.com>

Personally, I thought Python 1.5x was so slick, it needed
little more from a language perspective. That said, list
comprehension is a nice enhancement. For 2.1, I think
there are three things Python needs to make it more broadly
useful, and more broadly used: speed, speed, and speed.

I can already hear the protests that Python is pretty fast
compared to some of the alternative, that it is fast enough
for whatever you are using it for, that one can always
optimize after the case, that often it is the glued pieces
that cause performance problems, and that critical code can
be rewritten in C/C++. All I can say is this: speed is to
scripting languages what location is to real estate. One of
the nice things about list comprehensions is that it
improves performance by abstracting an existing language
mechanism.

Russell


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


From tim.one at home.com  Thu Dec 21 19:53:52 2000
From: tim.one at home.com (Tim Peters)
Date: Thu, 21 Dec 2000 19:53:52 -0500
Subject: A Suggestion for Python Colon Syntax
In-Reply-To: <91trir$52l$1@hecate.umd.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEAOIFAA.tim.one@home.com>

[William Djaja Tjokroaminata]
> I saw many replies that give the reasons why the colons are
> *needed*.

Not mine:  it told you why the colons are *there*.  It's a human-factors
thing, not a tech thing.  Desirable, not necessary.

> My suggestion is to make it *optional*, like the semicolon at
> the end of statements.

Understood.  Won't happen, though:  Python has been around for a decade.  Do
you imagine that Guido hasn't heard this suggestion before?  In fact there
are very few "new syntax" ideas that weren't debated and rejected before
1991 ended.  And, yes, this is one of them (ditto the whole "while 1" thread
that's been going on virtually non-stop for a decade).

> Yes, for people who cares about making it easier to read for newbies,
> they can always put colons.  Yes, for people who use emacs to
> automatically indent the next line, they can put colons too.

Guido abhors gratuitous variations in coding sytle; that's largely why
indentation is used instead of curly braces or begin/end to begin with.  A
choice about whether or not to use a colon is simply not something people
*need*.  The few places Python allows gratuitous choices today are places
Guido regrets (although not enough to lose sleep over <wink>).  For example,
Guido is unhappy that both "<>" and "!=" are accepted to mean "not equal";
the older "<>" survives only for compatibility with a very early release of
the language.

> ...
> I want opinion of you who have programmed for a long time whether
> there will be some harms if the colons are made optional.

It's Pythonic to either always require them or never allow them; the former
won 10 years ago.

> If not, probably someone can tell me how I can forward this
> suggestion to the core Python development team?

You already did <wink> -- Guido's on vacation this week, which makes me as
close as you'll get until after Christmas.  You can submit a formal feature
request via the SourceForge bug manager at:

    http://sourceforge.net/bugs/?group_id=5470

but you should expect it to be rejected.  Besides that Guido will object on
principle, the practical support costs of adding such a thing are enormous:
editing environments from the Emacs Python mode to IDLE would be broken, and
would be difficult to fix.  For a hint as to why it's difficult, study
tokenize.py in the std library.  That trailing colon makes partial parsing a
whole lot easier, and especially when parsing backwards (as e.g. the Emacs
python-mode and IDLE and PythonWin need to do routinely).

    if  (a,
         b,
         c) != \
        (d,
         e,
         f)
        (a,
         b,
         c) = \
        (d,
         e,
         f)
ly y'rs  - tim




From aleaxit at yahoo.com  Wed Dec  6 04:26:06 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 6 Dec 2000 10:26:06 +0100
Subject: Using dictionaries
References: <3a2c9264.503856587@news-server.bigpond.net.au> <90i8a001u10@news2.newsguy.com> <3a2dca59.583722939@news-server.bigpond.net.au>
Message-ID: <90l0n001tdg@news2.newsguy.com>

"sorular" <sorular at netscape.net> wrote in message
news:3a2dca59.583722939 at news-server.bigpond.net.au...
    [snip]
> The dictionary would be {'UserXXXX':'123123'}
>
> ----Configuration.py (start)----
> ...
> UserName=UserXXXX
> UserNumb="123123"
> ...

I'll assume you mean

UserName='UserXXXX'

as these quotes are really crucial.


> self.authenticate=Dict_lookup(UserName=function(UserNumb))

This passes to this __init__:

> class Dict_lookup:
>     def __init__(self,**kw):
>         self.dict=kw

a 'kw' dictionary like {'UserName': function(UserNumb)}.

In other words, in the sintax:

    callable(aname=avalue)

aname is NOT looked up anywhere: {'aname': avalue} is the
resulting keywords-dictionary (which the callable can
retrieve as **kw).  'aname' is taken as *constant* in
this syntactic form.


Try, instead (Python 2 syntax):

self.authenticate=Dict_lookup(**{UserName:function(UserNumb)})

here, you are explicitly building and passing the keywords
dictionary, rather than relying on it being built for you
by the Python compiler and interpreter; so, you can use any
expressions you desire for the keys as well as for the values,
rather than being limited to constant keys as in the name=value
keywords-syntax.

Alternatively, save yourself the double stars on both the
call and the function definition:

class Dict_lookup:
    def __init__(self,kw):
        self.dict=kw

self.authenticate=Dict_lookup({UserName:function(UserNumb)})

and the net effect will be the same (though it will also
work on older versions of Python; the double-stars syntax
on the call only works in Python 2).


Alex





From Blair.Lowe at compeng.net  Wed Dec 27 15:14:26 2000
From: Blair.Lowe at compeng.net (Blair Lowe)
Date: Wed, 27 Dec 2000 13:14:26 -0700
Subject: Loops (was while true: !!!)
In-Reply-To: <200012111656.JAA15835@mail.cybermesa.com>
References: <200012111656.JAA15835@mail.cybermesa.com>
Message-ID: <v04220801b66ffc0ee736@[192.168.11.33]>

Hi,

At 16:56 +0000 2000/12/11, Jay O'Connor wrote:
>
>Here, you have both the the start, end, and step conditions. However, i is
>not immutable in the loop, so you could conceivably have something in the
>loop effect i, and therefor have an endless loop.

Daemons (Unix server programs such as the apache web server httpd) 
are examples of programs that should have an infinite loops. There 
are all over the place. The infinite loop construct is a standard 
construct as many here have said already. Limiting a loop just to 
make it a non infinite loop construct is not a solution unless there 
is a conditional reason to do so... thus the repeat ... until 
<condition> construct. Check out "UNIX Network Programming" by 
Stevens, published by Prentice Hall.

>I feel the approach of using "while (1):" could suffer the same 
>danger. Here you are literally declaring an endless loop, counting 
>on something to happen within the loop that will cause the loop to 
>end. If you have a bug or accident and that never happens, ooops....

That's why we debug.

There loops are often used to wait for signals, and have a routine in 
them that actually falls asleep so that little processor power is 
used in the meantime.

Debugging and testing should also be done on a development machine so 
that catastrophic errors do not occur that grind a production machine 
to a halt.

Blair.
Computer Engineering Inc. http://www.compeng.net
Phone: 780 499 5687 (9 - 5 MST) Fax:   780 435 0693 (24 Hours)
All email advice that is provided for free is without warrantee: use 
at your own risk. 



From root at rainerdeyke.com  Wed Dec 13 13:43:30 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Wed, 13 Dec 2000 18:43:30 GMT
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org>
Message-ID: <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com>

"Aaron Jon Reichow" <reichowa at tcfreenet.org> wrote in message
news:Pine.BSF.4.10.10012131113170.59599-100000 at tcfreenet.org...
> On 13 Dec 2000, Bjoern Giesler wrote:
>
> > : Would it perhaps not be better to use Ruby ?
> >
> > Not for me, for the following reasons:
> > a) I don't know Ruby well enough
> > b) Ruby is "even more" unknown in The Real World(tm) than Python
> > c) I think Ruby code looks ugly (what with @, #, $, no 'return' etc.)
> > d) I don't see any of the features Ruby has over Python as 'killer'
> >    features.
>
> Proper blocks rather than awkward lambda functions, for starters.

Disclaimer: my knowledge of Ruby is limited to a recent magazine article I
read.

I intensely dislike Ruby's block system because they are inconsistent with
other method arguments and a method cannot accept more than one block as
argument.  The syntax for actually creating a block is nice, but all in all
I prefer Python.

I also dislike forced single inheritance.  If mix-ins (as a replacement for
multiple inheritance) are so powerful, why include inheritance in the
language at all?


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From victor at prodigy.net  Thu Dec 28 02:02:07 2000
From: victor at prodigy.net (Victor Muslin)
Date: Thu, 28 Dec 2000 07:02:07 GMT
Subject: fcgi.py
References: <3a4a865c.1325844015@localhost> <3a4a90ab.1328483109@localhost> <lYy26.91$9Ow6.17498496@news.randori.com>
Message-ID: <3a4aca25.98948941@newnews.prodigy.net>

Thanks, Robin. I was just thinking the same thing. Seems that for a
simple-minded fix changing

        s=socket.fromfd(sys.stdin.fileno(), socket.AF_INET,
                        socket.SOCK_STREAM)
        s.getpeername()
 
to

            s=socket.socket(socket.AF_INET, socket.SOCK_STREAM)
            s.bind(('', port))          # bind to the localhost
            s.listen(1)

did the trick. Then I added a few more tweaks to allow specification of port as
a parameter to Accept() and tweaked the _test() function to allow specification
of port as a command line parameter for testing.

For anyone who may wish to use this module on WindowsNT the full source code may
be downloaded from http://pages.prodigy.net/victor/fcgi2_py. The directory
contains two files:

	fcgi2.py - modification of original Robin Dunn's fcgi.py
	fcgi2.fcgi - a sample config file for cgi-fcgi program

fcgi2.py can be tested by running it from a command line with or without a port
parameter as follows:

(1)	python fcgi2.py

or

(2)	python fcgi2.py -port 9001

If invoked as in (1) it should run the same way as the original fcgi.py. If
invoked as in (2) it should run as a daemon listening on localhost's port 9001.
For more detailed usage information run it as:

(3)	python fcgi2.py -?

On Wed, 27 Dec 2000 20:03:10 -0800, "Robin Dunn" <robin at alldunn.com> wrote:

>"Victor Muslin" <victor at prodigy.net> wrote in message
>news:3a4a90ab.1328483109 at localhost...
>> I just found an old posting that says that socket.fromfd() is not
>> implemented on Windows platform (though Python documentation does not
>> say so). Is there a FastCGI solution for Python for WindowsNT???
>>
>
>The fcgi.py module is written such that it expects to be started from a
>"FastCGI Environment" such as from mod_fastcgi in Apache or from cgi-fcgi.
>On Unix this means that the file descriptor for stdin is a listening socket
>pre-created by the parent process that the webserver will connect to in
>order to send requests to your fastcgi application.  It's done this way so
>mod_fastcgi can create multiple processes all listening on the same socket,
>not just the same port, but the same socket.  This is definitly a
>unix-thing.
>
>On the other hand if you don't mind losing the multiple process capability,
>fcgi.py could probably be tweaked to create its own listening socket instead
>of expecting it to already be there.  It could then be started up
>independently of the webserver, and on any machine.  With creative use of
>threads you can even regain most of the benefit of having multiple processes
>running, (start a new thread, or allocate one from a thread pool, for every
>connection to the socket.)
>
>--
>Robin Dunn
>Software Craftsman
>robin at AllDunn.com
>http://wxPython.org     Java give you jitters?
>http://wxPROs.com        Relax with wxPython!
>
>
>



From aleaxit at yahoo.com  Thu Dec 14 03:46:52 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 14 Dec 2000 09:46:52 +0100
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <tcmay-FCA22F.14073713122000@news.got.net> <918v7a02scn@news1.newsguy.com> <tcmay-04E80C.16181313122000@news.got.net>
Message-ID: <91a1ed02a4j@news2.newsguy.com>

"Tim May" <tcmay at got.net> wrote in message
news:tcmay-04E80C.16181313122000 at news.got.net...
    [snip]
> Being interested in ontology (things, objects, hierarchies, structures)
> even more so than in advocating any one language over another, I would
> be fascinated to see some good examples of MI in actual use.

I can't find a source any more, but one of the most interesting
arguments I've heard for "only inherit from abstract classes"
(which is a precious, highly-practical rule of thumb for C++
and other languages, championed on practical grounds by Scott
Meyers and other C++ gurus) is actually ontology-based, and
aimed at the object-oriented analysis stage.

What classes are 'abstract' (not actually instantiated as
themselves, but only in terms of their subclasses) or 'concrete'
(directly instantiable without subclassing) is part of what your
O-O analysis is determining, and will depend on what exactly
you're trying to do.  The root of the 'only abstract classes
ever get subclassed' idea is that there is an inconsistency
of levels-of-detail in your analysis if it allows subclassing
of instantiable classes.  If your application-needs (or more
generally, your domain-modeling needs) are such that you need
to distinguish between, e.g., different kinds of 'vehicles'
(cars vs trucks vs planes vs ...), then 'Vehicle' itself, the
common superclass, cannot be instantiable in a logically
consistent way -- it would break the idea that your concrete
classes partition the universe of discourse, if it were.

This is based on the idea of subclassing as asserting IS-A.

As, say, a Plane IS-A Vehicle, it cannot be the case (if your
concrete classes partition the universe of discourse) that
a (generic) Vehicle is instantiable; else, you would have
ambiguity in what concrete class to use to model an actual
plane -- either a concrete Plane or a concrete generic Vehicle
would satisfy all expressed constraints.  In practice, object-models
which allow instantiation of a superclass (such as Vehicle) do
so by adding a NON-expressed constraint about using the most
'leaf' class an actual object can possibly be modeled as; but,
that analysis-constraint is left non-explicitly-expressed in
the object model... it CANNOT be expressed in it.  Whatever
semantic constraints we attach to Vehicle will get inherited
by Plane, so we cannot say *of Vehicle instances* 'and this
is not specifically a Plane', else, bye-bye IS-A.

The extra constraints need to be placed in _another_ subclass
of Vehicle -- OtherwiseUnclassifiedVehicle, say.  Now *THAT*
one can be concrete -- and it will not be further subclassed.

I'm not sure I'm expressing this well, and will appreciate
it if anybody recognizes the source of this (I guess I've
absorbed this from some O-O Analysis classic, but I've looked
far and wide and can't put my finger on *which* one!).


If one accepts the idea that, at the OOA level, subclassing
is always only properly performed on abstract classes, then
of course the same idea will generalize to _multiple_
inheritance -- the (multiple) superclasses of any given
(abstract or concrete) class will all be abstract ones.

Which, in turn, tends to raise a typical issue -- are we
indeed modeling IS-A, or aren't we surreptitiously doing
PLAYS-THE-ROLE-OF by the inappropriate avenue of subclassing?

And if this doesn't lead us deep enough into troubled
ontological waters, there's the sometimes-unexpressed
meta-hypothesis behind it all -- that 'this object is
an instance of this class' is a *permanent* property of
the object; which, in turn, is why playing-a-role (which
is not permanent) is badly modeled by class-membership
(and subclassing).

In an object-system where class-membership itself is
dynamic and fluid, and new classes can be built on the
fly, change their bases ditto, have objects previously
belonging to other classes switched to them, etc, the
applicability of these criteria can be very doubtful.

Python's object system is like this, for example, though
I have not seen much actual use of this potential so
far (which fits with Python's philosophy: consistently
aim for simplicity, which, at the language level, means
not introducing constraints [any extra constraint is an
extra complication], but, at the programming level,
means 'doing the simplest thing that can possibly
work' [to quote a snippet from XP philosophy] -- not
often will 'the simplest thing' include wild abandon
in switching class-memberships around, generating new
classes on the fly, changing inheritance-DAG's at
runtime, and so on:-).


Alex





From aleaxit at yahoo.com  Fri Dec 22 06:53:25 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 22 Dec 2000 12:53:25 +0100
Subject: Speed of Python 1.5.x vs 2.0
References: <4vj54t4v0hvjsnstfmtmf36q9irotp59sj@4ax.com> <mailman.977475665.3482.python-list@python.org> <dud64ts8mb7mlc41kht410q5g87ekb2nc5@4ax.com>
Message-ID: <91vfc6029r5@news2.newsguy.com>

"Steve Horne" <sh at ttsoftware.co.uk> wrote in message
news:dud64ts8mb7mlc41kht410q5g87ekb2nc5 at 4ax.com...
    [snip]
> I'd expect list comprehensions to work significantly faster than
> equivalent expressions using filter, map and lambdas - the reason
> being that the overhead of repeatedly calling the lambda should be
> avoided, not to mention avoiding the temp storage between the filter
> and map.
>
> I haven't tried this, but it might be worth a go.

You'll be disappointed, I fear... a typical simple case
produces not-very-significant differences, e.g.:

import time

def withMap(inList):
    return map(lambda x: x+1, inList)

def withLC(inList):
    return [x+1 for x in inList]

def stopWatch(function, N=100*1000):
    start = time.clock()
    function(range(N))
    stend = time.clock()
    return stend-start

print "Using map and lambda:     %f" % stopWatch(withMap)
print "Using list-comprehension: %f" % stopWatch(withMap)



D:\PySym>python tim.py
Using map and lambda:     0.499228
Using list-comprehension: 0.489680

D:\PySym>python tim.py
Using map and lambda:     0.502484
Using list-comprehension: 0.502492

D:\PySym>python tim.py
Using map and lambda:     0.491193
Using list-comprehension: 0.483273


There may be a tiny margin for the list comprehension, but
hardly anything worth writing home about.  And I remember
posts from a few months ago that identified cases in which
map was systematically faster (though, if I recall right,
not by all that much either).

Nor does adding a filtering step change much...:

import time

def withMap(inList):
    return map(lambda x: x+1, filter(lambda x: x%2, inList))

def withLC(inList):
    return [x+1 for x in inList if x%2]

def stopWatch(function, N=100*1000):
    start = time.clock()
    function(range(N))
    stend = time.clock()
    return stend-start

print "Using map and lambda:     %f" % stopWatch(withMap)
print "Using list-comprehension: %f" % stopWatch(withMap)


D:\PySym>python tim.py
Using map and lambda:     0.834212
Using list-comprehension: 0.808803

D:\PySym>python tim.py
Using map and lambda:     0.832734
Using list-comprehension: 0.818849

D:\PySym>python tim.py
Using map and lambda:     0.832896
Using list-comprehension: 0.819406


It may be 2% rather than 1%, but I strongly doubt such
differences can be considered meaningful for any real
Python use case.


Alex





From fgeiger at datec.at  Fri Dec 29 12:16:09 2000
From: fgeiger at datec.at (Franz GEIGER)
Date: Fri, 29 Dec 2000 18:16:09 +0100
Subject: popen(), ActivePython and ActiveCommodo on one machine
References: <92g4jo$61k$1@newsreaderm1.core.theplanet.net>
Message-ID: <92ih26$br0$1@newsreaderm1.core.theplanet.net>

O.k., I uninst'ed ActiveKomodo, ActivePerl and ActivePython. Then reinst'ed
ActivePython and everything's o.k. again.

Sorry for that post.

Regards
Franz

"Franz GEIGER" <fgeiger at datec.at> wrote in message
news:92g4jo$61k$1 at newsreaderm1.core.theplanet.net...
> Did anyone experience incompatibilities between scripts using
>
> f = popen("net view")
> f.read()
> f.close()
>
> and PythonWin (ActivePython)?
>
> I am not sure, but at about the same time as I use these commands I
> installed ActiveCommodo 1.0 Beta 1. So probably this could be a problem
now.
>
> For known reasons (Win NT 4.0 Svr) I'm a bit afraid of
> uninstalling/installing this and that.
>
> Any experinces so far?
>
>
> Thanks in advance and
> Best regards
> Franz GEIGER
>
>
>




From clarence at netlojix.com  Fri Dec 29 14:24:06 2000
From: clarence at netlojix.com (Clarence Gardner)
Date: Fri, 29 Dec 2000 11:24:06 -0800
Subject: Question re: eval()
References: <978109654.339714327@news.silcom.com> <92iivt$6uk0r$1@ID-11957.news.dfncis.de>
Message-ID: <978118012.3281608@news.silcom.com>

On Fri, 29 Dec 2000, Emile van Sebille wrote:
>What are you trying to do that using a dict doesn't address?

The lazy evaluation.  Not only will the values that I want to supply
require large database accesses to retrieve, but I don't even know
with what names they will be requested!

Clarence Gardner
clarence at netlojix.com


From magnus.heino at rivermen.se  Wed Dec 13 09:03:39 2000
From: magnus.heino at rivermen.se (Magnus Heino)
Date: Wed, 13 Dec 2000 14:03:39 +0000
Subject: I dont remember the name of the project...
Message-ID: <917s2u$ibr$1@merlin.rembrandtstr.bocholt.de>


Hi.

A few days ago, I visited the website of a project that provided a 
framework like CORBA, but much more lightweight, and written in python.

I dont remember the name of the project, anyone know what I may be refering 
to??

/Magnus


From ullrich at math.okstate.edu  Fri Dec 29 09:02:25 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Fri, 29 Dec 2000 14:02:25 GMT
Subject: how do i check if a file exists with python?
References: <fBN26.99045$yR4.2640173@news1.rdc1.tx.home.com>, <RlM26.98766$yR4.2639444@news1.rdc1.tx.home.com> <92g5v8$r5k$1@nnrp1.deja.com> <mailman.978078157.26485.python-list@python.org>
Message-ID: <3a4c95dc.88792787@nntp.sprynet.com>

On Fri, 29 Dec 2000 16:54:56 +0200 (IST), Moshe Zadka
<moshez at zadka.site.co.il> wrote:

>[Scott Hathaway, on how to find a file]
>> Thanks for both responses!
>
>Unfortunately, I think they did you more harm then good by telling you
>the answer: I doubt you need it. Now, it may be that I'm wrong, in which
>case just go ahead an use os.path.isfile, but if you're doing something
>like
>
>def foo():
>	if os.path.isfile("foo"):
>		return open("foo")
>	else:
>		return None
>
>Or something to that effect, *don't*. It's evil, and will come back
>to bite you (unreadable existing files, files changing between isfile
>and open, etc.).
>
>Instead, simple do:
>
>def foo():
>	try:
>		return open("foo")
>	except IOError:
>		return None
>
>In other words, don't try to avoid exceptions -- know where you want
>to deal with them, and do so.

Hmm. Probably isfile should be removed from the language,
since it's considered harmful, eh?

It's certainly true the the second foo is better than the first,
in fact yes the first is evil. But there are other reasons a person 
might want to know whether a file exists (although probably
things like foo are the most likely explanation.)

I sometimes make cumulative backups of files containing
important stuff. I sort of need to determine that file.001,
file.002 and file.003 already exist and file.004 does not
before deciding what to name the current backup file.

(When I tell people they're asking the wrong question I
get in trouble for it...)

>Moshe Zadka <sig at zadka.site.co.il>
>This is a signature anti-virus. 
>Please stop the spread of signature viruses!
>



From adjih-spam at crepuscule.com  Fri Dec 29 07:04:17 2000
From: adjih-spam at crepuscule.com (Cedric Adjih)
Date: 29 Dec 2000 12:04:17 GMT
Subject: unix pipes + python
References: <m4iu27odrr4.fsf@macquarie.com.au>
Message-ID: <92huk1$oe8$1@ites.inria.fr>

Timothy Docker <timd at macquarie.com.au> wrote:

> I must be doing something silly here...
>
> | import os
> | 
> | def child( infd  ) :
> |     "Run a child process"
> |     pid = os.fork()
> |     if pid == 0:
> |         if infd != None: os.dup2( infd, 0 )
> |         os.execv( '/usr/bin/cat', ['cat']  )
> |     else:
> | 	return pid
> | 
> | toPipe,fromPipe = os.pipe()
> | child( fromPipe )
> | os.write( toPipe, "hello\nthere\n" )
> | os.close( toPipe )
>
> This script behavious almost as expected. I end up with the text
> echoed back, but the 'cat' subprocess is left running - I was
> expecting it to exit when the write side of the pipe was closed in the 
> parent, but it is still reading from fd 0.
>
> Any suggestions on what's wrong?

According to Python it should be:
  fromPipe,toPipe = os.pipe()
You are reading and writing on the wrong descriptor of the pair.

Also closing the unused descriptor might be good style
(closing `fromPipe' in the parent, and `toPipe' in the child).

-- Cedric.


From Gareth.McCaughan at pobox.com  Fri Dec 29 17:14:36 2000
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: Fri, 29 Dec 2000 22:14:36 +0000
Subject: 2001 Enchancement Wishlist
References: <3A4CEACB.220998A0@javanet.com>
Message-ID: <slrn94q36c.8rr.Gareth.McCaughan@g.local>

Raymond Hettinger wrote:

> 4.  Allow dictionaries to respond to sequence operations as if .keys() were
>      in the list object.  Examples:
> 
>     knights = { 'lancalot':'good', 'gallahad':'brave', 'robin':'coward' }
>     for name in knights:
>            print name, knights[name]
>     print knights.len()
>     if 'arthur' in knights:
>            print 'Not the King'
>     print knights[0]              # may be any one depending on the hash table order
>     try:
>         print knights[5]
>     except IndexError:
>         print 'Too few knights for the ladies in the castle'
>     del knights['lancelot']      # regular mapping operations left unchanged.

Nope.

  | >>> knights = {}
  | >>> knights[0] = 'foo'
  | >>> knights[0]
  | 'foo'
  | >>> knights[1] = 'bar'
  | >>> knights.keys()
  | [1, 0]

Uh-oh.

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
sig under construc


From fgeiger at datec.at  Sun Dec 31 03:37:27 2000
From: fgeiger at datec.at (Franz GEIGER)
Date: Sun, 31 Dec 2000 09:37:27 +0100
Subject: Running Boa Contructor with the latest versions of Python & wxPython
References: <92ld65$4un$1@news.nuri.net>
Message-ID: <92mrdn$um3$1@newsreaderg1.core.theplanet.net>

I went back to wxPython-2.2.1-Py20b1.EXE, and everything was fine again.

Reagrds
Franz

"June Kim" <junaftnoon at nospamplzyahoo.com> wrote in message
news:92ld65$4un$1 at news.nuri.net...
> I tried to run BC on my windows machine installed with the lastest
versions
> of Python(2.0) and wxPython(2.2.2) and got the following error.
>
> Any roundabout ways to run it besides downgrading the Python and wxPython?
>
> Traceback (most recent call last):
>   File "boa.py", line 119, in ?
>     main()
>   File "boa.py", line 112, in main
>     app = BoaApp(0)
>   File "boa.py", line 49, in __init__
>     wxApp.__init__(self, redirect)
>   File "c:\python20\wxPython\wx.py", line 1646, in __init__
>     _wxStart(self.OnInit)
>   File "boa.py", line 58, in OnInit
>     self.main = Palette.BoaFrame(None, -1, self)
>   File "Palette.py", line 157, in __init__
>     wxMenu(), self.componentSB, app)#palettePage.menu
>   File "Editor.py", line 135, in __init__
>     self.addShellPage()
>   File "Editor.py", line 310, in addShellPage
>     self.tabs.AddPage(ShellEditor.ShellEditor(self.tabs, -1), 'Shell')
>   File "ShellEditor.py", line 26, in __init__
>     PythonStyledTextCtrlMix.__init__(self, wId, -1)
>   File "Views\StyledTextCtrls.py", line 137, in __init__
>     self.SetKeywords(0,
> AttributeError: 'ShellEditor' instance has no attribute 'SetKeywords'
>
>
>




From teyc at bigfoot.com  Tue Dec 19 05:31:03 2000
From: teyc at bigfoot.com (Chui Tey)
Date: Tue, 19 Dec 2000 20:31:03 +1000
Subject: Python interpreter in python?
Message-ID: <91ndgd$g8d$1@bunyip.cc.uq.edu.au>

Is there a python implementation of the python VM? Basically for
experimenting?

Chui





From wware at world.std.com  Wed Dec 20 21:44:29 2000
From: wware at world.std.com (Will Ware)
Date: Thu, 21 Dec 2000 02:44:29 GMT
Subject: Greetings from the Interpreter!!!
References: <mailman.977314029.15733.python-list@python.org> <slrn942oun.eoe.grey@teleute.rpglink.com>
Message-ID: <G5wCy5.783@world.std.com>

Steve Lamb (grey at despair.rpglink.com) wrote:
> >import string,time,os
> >print (string.join((lambda x:x[:10]+x[8:])(map(lambda x:
> >string.center("*" * (lambda x: ((x < 24) * (x - 3)) + 3)(x) ,24),
> >range(1,28,2))),os.linesep), "Please Wait...")[time.time()<977612400]

>     Alright, maybe I am missing something but that does nothing at all.

The portion "print string.join(.....os.linesep)" prints a Christmas
tree. The use of parentheses allows you to violate the normal requirement
of whitespace alignment. Voila, obfuscated Python. Obfuscating C or Perl
is no challenge since they're 90% obfuscated already by design.

IIRC there are a few examples of Perl poetry in the Camel book, and one
is particularly good, dark, and Poe-like. Alas my copy is not handy, and
I don't easily find it on the web, which surprises me.
-- 
import string,time,os;print string.join((lambda x:x[:10]+x[8:])(map(
lambda x:string.center("*"*(lambda x:((x<24) ### Seasons Greetings, Will Ware
*(x-3))+3)(x),24),range(1,28, 2))),"\n") ################ wware at world.std.com


From fredrik at effbot.org  Thu Dec 14 13:17:38 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Thu, 14 Dec 2000 18:17:38 GMT
Subject: Source Code Management
References: <fb4f3tsgffnmvme64aol1cn397te7ln349@4ax.com> <91aq96$tba$1@nnrp1.deja.com>
Message-ID: <698_5.3930$Kd1.455411@newsb.telia.net>

gtalvola at NameConnector.com wrote:
> If you don't mind paying money, I believe Perforce fits all your
> requirements.  http://www.perforce.com/  We use it and have nothing
> but good things to say about it.

Couldn't agree more.

(PythonWorks 1.1 can talk to Perforce, of course...)

</F>




From vbusch at gmx.de  Thu Dec 28 12:16:56 2000
From: vbusch at gmx.de (Volkmar Busch)
Date: Thu, 28 Dec 2000 17:16:56 GMT
Subject: Python for prototyping of the user interface
Message-ID: <20001228.17165633@mis.configured.host>

Hi,

I took my first steps in using Python, and I really like the language. I 
read a lot about using Python for prototyping. As a basis for discussion 
with our users we use GUI prototypes. Building complex user interfaces in 
Python to me seems quite complicated, because

- I am missing a Gui builder
- I am missing controls like notebooks and tables

So I'd like to asked the experienced Python users how they solve this 
problem.

Thanks in advance,
Volkmar


From amk at mira.erols.com  Fri Dec 22 14:28:00 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 22 Dec 2000 19:28:00 GMT
Subject: PEP 222 draft
Message-ID: <slrn947bj1.1rs.amk@207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com>

I've completed a draft of PEP 222 (sort of -- note the XXX comments in
the text for things that still need to be resolved).  This is being
posted to python-dev, python-web-modules, and
python-list/comp.lang.python, to get comments on the proposed
interface.  I'm on all three lists, but would prefer to see followups
on python-list/comp.lang.python.

--amk

Abstract

    This PEP proposes a set of enhancements to the CGI development
    facilities in the Python standard library.  Enhancements might be
    new features, new modules for tasks such as cookie support, or
    removal of obsolete code.

    The intent is to incorporate the proposals emerging from this
    document into Python 2.1, due to be released in the first half of
    2001.


Open Issues

    This section lists changes that have been suggested, but about
    which no firm decision has yet been made.  In the final version of
    this PEP, this section should be empty, as all the changes should
    be classified as accepted or rejected.

    cgi.py: We should not be told to create our own subclass just so
    we can handle file uploads. As a practical matter, I have yet to
    find the time to do this right, so I end up reading cgi.py's temp
    file into, at best, another file. Some of our legacy code actually
    reads it into a second temp file, then into a final destination!
    And even if we did, that would mean creating yet another object
    with its __init__ call and associated overhead.

    cgi.py: Currently, query data with no `=' are ignored.  Even if
    keep_blank_values is set, queries like `...?value=&...' are
    returned with blank values but queries like `...?value&...' are
    completely lost.  It would be great if such data were made
    available through the FieldStorage interface, either as entries
    with None as values, or in a separate list.

    Utility function: build a query string from a list of 2-tuples

    Dictionary-related utility classes: NoKeyErrors (returns an empty
    string, never a KeyError), PartialStringSubstitution (returns 
    the original key string, never a KeyError)


    
New Modules

    This section lists details about entire new packages or modules
    that should be added to the Python standard library.

    * fcgi.py : A new module adding support for the FastCGI protocol.
      Robin Dunn's code needs to be ported to Windows, though.

Major Changes to Existing Modules

    This section lists details of major changes to existing modules,
    whether in implementation or in interface.  The changes in this
    section therefore carry greater degrees of risk, either in
    introducing bugs or a backward incompatibility.

    The cgi.py module would be deprecated.  (XXX A new module or
    package name hasn't been chosen yet: 'web'?  'cgilib'?)

Minor Changes to Existing Modules

    This section lists details of minor changes to existing modules.
    These changes should have relatively small implementations, and
    have little risk of introducing incompatibilities with previous
    versions.


Rejected Changes

    The changes listed in this section were proposed for Python 2.1,
    but were rejected as unsuitable.  For each rejected change, a
    rationale is given describing why the change was deemed
    inappropriate.

    * An HTML generation module is not part of this PEP.  Several such
      modules exist, ranging from HTMLgen's purely programming
      interface to ASP-inspired simple templating to DTML's complex
      templating.  There's no indication of which templating module to
      enshrine in the standard library, and that probably means that
      no module should be so chosen.

    * cgi.py: Allowing a combination of query data and POST data.
      This doesn't seem to be standard at all, and therefore is
      dubious practice.

Proposed Interface

    XXX open issues: naming convention (studlycaps or
    underline-separated?); need to look at the cgi.parse*() functions
    and see if they can be simplified, too.

    Parsing functions: carry over most of the parse* functions from
    cgi.py
    
    # The Response class borrows most of its methods from Zope's
    # HTTPResponse class.
    
    class Response:
        """
        Attributes:
        status: HTTP status code to return
        headers: dictionary of response headers
        body: string containing the body of the HTTP response
        """
        
        def __init__(self, status=200, headers={}, body=""):
            pass
    
        def setStatus(self, status, reason=None):
            "Set the numeric HTTP response code"
            pass
    
        def setHeader(self, name, value):
            "Set an HTTP header"
            pass
    
        def setBody(self, body):
            "Set the body of the response"
            pass
    
        def setCookie(self, name, value,
                      path = '/',  
                      comment = None, 
                      domain = None, 
                      max-age = None,
                      expires = None,
                      secure = 0
                      ):
            "Set a cookie"
            pass
    
        def expireCookie(self, name):
            "Remove a cookie from the user"
            pass
    
        def redirect(self, url):
            "Redirect the browser to another URL"
            pass
    
        def __str__(self):
            "Convert entire response to a string"
            pass
    
        def dump(self):
            "Return a string representation useful for debugging"
            pass
            
        # XXX methods for specific classes of error:serverError, 
	# badRequest, etc.?
    
    
    class Request:
    
        """
        Attributes: 

        XXX should these be dictionaries, or dictionary-like objects?
        .headers : dictionary containing HTTP headers
        .cookies : dictionary of cookies
        .fields  : data from the form
        .env     : environment dictionary
        """
        
        def __init__(self, environ=os.environ, stdin=sys.stdin,
                     keep_blank_values=1, strict_parsing=0):
            """Initialize the request object, using the provided environment
            and standard input."""
            pass
    
        # Should people just use the dictionaries directly?
        def getHeader(self, name, default=None):
            pass
    
        def getCookie(self, name, default=None):
            pass
    
        def getField(self, name, default=None):
            "Return field's value as a string (even if it's an uploaded file)"
            pass
            
        def getUploadedFile(self, name):
            """Returns a file object that can be read to obtain the contents
            of an uploaded file.  XXX should this report an error if the 
            field isn't actually an uploaded file?  Or should it wrap
            a StringIO around simple fields for consistency?
            """
            
        def getURL(self, n=0, query_string=0):
            """Return the URL of the current request, chopping off 'n' path
            components from the right.  Eg. if the URL is
            "http://foo.com/bar/baz/quux", n=2 would return
            "http://foo.com/bar".  Does not include the query string (if
            any)
            """

        def getBaseURL(self, n=0):
            """Return the base URL of the current request, adding 'n' path
            components to the end to recreate more of the whole URL.  
            
            Eg. if the request URL is
            "http://foo.com/q/bar/baz/qux", n=0 would return
            "http://foo.com/", and n=2 "http://foo.com/q/bar".
            
            Returned URL does not include the query string, if any.
            """
        
        def dump(self):
            "String representation suitable for debugging output"
            pass
    
        # Possibilities?  I don't know if these are worth doing in the 
        # basic objects.
        def getBrowser(self):
            "Returns Mozilla/IE/Lynx/Opera/whatever"
    
        def isSecure(self):
            "Return true if this is an SSLified request"
            

    # Module-level function        
    def wrapper(func, logfile=sys.stderr):
        """
        Calls the function 'func', passing it the arguments
        (request, response, logfile).  Exceptions are trapped and
        sent to the file 'logfile'.  
        """
        # This wrapper will detect if it's being called from the command-line,
        # and if so, it will run in a debugging mode; name=value pairs 
        # can be entered on standard input to set field values.
        # (XXX how to do file uploads in this syntax?)

    
Copyright
    
    This document has been placed in the public domain.



From jp at ulgo.koti.com.pl  Wed Dec 27 20:04:35 2000
From: jp at ulgo.koti.com.pl (Jacek =?iso-8859-2?Q?Pop=B3awski?=)
Date: 28 Dec 2000 01:04:35 GMT
Subject: split hangs
References: <slrn94l1pu.4so.jp@localhost.localdomain> <92e3lm$pni$1@panix3.panix.com>
Message-ID: <slrn94l4d3.550.jp@localhost.localdomain>

Aahz Maruch wrote:
>What is the regex?

ourregex=re.compile(r'(<(?:[^<>"]+(?:"[^"]*")?)+>)')

-- 
Every time I climb the mountain 
And it turned into a hill
I promised me that I'd move on
And I will.                              "Strange Highways" - Ronnie James Dio


From MarkH at ActiveState.com  Wed Dec 13 17:24:56 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Wed, 13 Dec 2000 22:24:56 GMT
Subject: Problems With socket Under NT & PythonWin 2.0
References: <910jtt$nnh$1@nnrp1.deja.com> <fMQY5.12943$hp.1457332@typhoon.nyroc.rr.com> <915k4g$kcc$1@nnrp1.deja.com> <8QzZ5.1711$7g.340689@typhoon.nyroc.rr.com>
Message-ID: <3A37F5FA.3020209@ActiveState.com>

Darrell wrote:

> I was guessing that your app was sharing an important thread with PythonWin.

That would be the main thread ;-)

> PythonWin could have started your app in another thread to avoid this.

Exactly.  In general, this problem will exist for all IDEs that run code 
in the same process.  For this reason, most IDEs are moving to 
out-of-proc debugging.  Komodo has only out-of-proc debugging (and works 
very nicely too)

Mark.



From grey at despair.rpglink.com  Sun Dec 31 12:27:42 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Sun, 31 Dec 2000 17:27:42 -0000
Subject: why no "do : until"?
References: <3A4E95FD.D6C737A7@engcorp.com>, <92jj3j$ff7$1@nnrp1.deja.com> <slrn94shbe.150.grey@teleute.rpglink.com> <X8s36.2731$Qb7.305339@newsb.telia.net> <slrn94smds.150.grey@teleute.rpglink.com> <FOt36.17$vd1.2224@e420r-atl2.usenetserver.com> <slrn94t1vj.150.grey@teleute.rpglink.com> <mailman.978282497.8457.python-list@python.org>
Message-ID: <slrn94ur4e.150.grey@teleute.rpglink.com>

On Mon,  1 Jan 2001 00:49:25 +0200 (IST), Moshe Zadka
<moshez at zadka.site.co.il> wrote:
>> The advice that runs
>> contrary to your 10+ years seems very reasonable to my 20+ years of
>> programming experience, and apparently many others with a variety of
>> experiences agree.  

>oooooh, oooooh, my dick is longer.
>I've been programming with it for, like, 100+ yrs. I haven't really
>got an opinion about /F's advice -- I just wanted to show off.

    If you're 100+ years old I seriously doubt your dick is anything but a
shriveled stack o' dimes.  :)

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From jp at ulgo.koti.com.pl  Thu Dec 21 19:27:06 2000
From: jp at ulgo.koti.com.pl (Jacek =?iso-8859-2?Q?Pop=B3awski?=)
Date: 22 Dec 2000 00:27:06 GMT
Subject: reading numbers from file
Message-ID: <slrn9457uq.9tl.jp@localhost.localdomain>

I have file with numbers:

23 41 24 0 2 42 532 53 234
34 234 53 23 4 323 53 (etc...)

1.How to read one line of this file and write it to list of numbers?
  (like: a=[23,41,24...])
2.How to read only 5 numbers, write it to list, then another 5, and write
  it to second list?
3.What will change if I want to add comments (#) and empty lines to my file?

-- 
Don't care which God you follow 
Whose promises you swallow
Time and again  
We must meet at the end              "Lord of the Last Day" - Ronnie James Dio


From peter at schneider-kamp.de  Thu Dec 28 09:12:15 2000
From: peter at schneider-kamp.de (Peter Schneider-Kamp)
Date: Thu, 28 Dec 2000 15:12:15 +0100
Subject: Good introduction to functional programming with Python?
References: <LNBBLJKPBEHFEDALKOLCAEKPIFAA.tim.one@home.com>
Message-ID: <3A4B4A3F.99A4E018@schneider-kamp.de>

Tim Peters schrieb:
> 
> if-we-don't-act-fast-i'm-afraid-python-may-lose-the-industrial-
>     control-market-ly y'rs  - tim

You'd have to write program verification tools for distributed
python applications...

maybe-that's-why-erlang-is-erlang-ly y'rs
Peter



From nospam at no.spam  Fri Dec 15 12:54:47 2000
From: nospam at no.spam (Carsten Geckeler)
Date: Fri, 15 Dec 2000 18:54:47 +0100 (MET)
Subject: Scopes and exceptions
In-Reply-To: <3A39F7CA.F8A@hvision.nl>
Message-ID: <Pine.LNX.4.10.10012151850350.657-100000@erde.kosmos.all>

On Fri, 15 Dec 2000, Hans Nowak wrote:

> Magnus Heino wrote:
> > 
> > Could someone please explain this to me:
> > 
> > [magnus at daysleeper magnus]$ cat e.py
> > 
> > a = 'test'
> > b = 'test'
> > 
> > print 'a      = %s' % id(a)
> > print 'b      = %s' % id(b)
> > 
> > def func():
> >         print "'test' = %s" % id('test')
> > 
> > func()
> > 
> > try:
> >         test, '%s' % (a,)
> > except NameError, e:
> >         print 'e      = %s' % id(e)
> >         print 'a      = %s' % id(a)
> >         print 'b      = %s' % id(b)
> >         print "'test' = %s" % id('test')
> > 
> > [magnus at daysleeper magnus]$ python e.py
> > a      = 136227368
> > b      = 136227368
> > 'test' = 136227368
> > e      = 135997204
> > a      = 136227368
> > b      = 136227368
> > 'test' = 136227368
> > [magnus at daysleeper magnus]$
> 
> I don't understand what is so special about this? Python stores a string
> 'test' in memory... a points to this string, then b, so id(a), id(b) and
> id('test') all yield the same result. This is also the case after the
> exception. Variable e has a different id, because it's a different
> variable; not a string, but an instance of exceptions.NameError. 

It _is_ a string, but it's not 'test' but
	"There is no variable named 'test'"
therefore the ids are different.

But the point of Magnus' mail was if
	id(a) == id('%s' % (a,))
and it is *not*.

Cheers, Carsten
-- 
Carsten Geckeler:  carsten dot geckeler at gmx dot de
To get proper email-address replace `dot' and `at' by the corresponding symbols.




From bradlosh at my-deja.com  Tue Dec 12 16:17:52 2000
From: bradlosh at my-deja.com (Brad)
Date: Tue, 12 Dec 2000 21:17:52 GMT
Subject: HPUX install
References: <913ett$u0b$1@nnrp1.deja.com>
Message-ID: <9164lp$3p7$1@nnrp1.deja.com>

I used gcc, and got the following error after typing make:

  (cd Modules; make -f Makefile.pre Makefile) + cp ./Setup.in Setup  echo "#
Edit this file for local setup changes" >Setup.local  rm -rf
../libpython2.0.a  /bin/sh ./makesetup Setup.config Setup.local Setup making
Makefile in subdirectory .  CONFIG_FILES=Makefile CONFIG_HEADERS= /bin/sh
config.status creating Makefile making Makefile in subdirectory Parser
`Makefile' is up to date. making Makefile in subdirectory Grammar `Makefile'
is up to date. making Makefile in subdirectory Objects	(cd ..;
CONFIG_FILES=Objects/Makefile CONFIG_HEADERS= \  /bin/sh config.status)
creating Objects/Makefile making Makefile in subdirectory Python `Makefile'
is up to date. making Makefile in subdirectory Modules `Makefile' is up to
date.  (rm -f Modules/hassignal; cd Modules; make hassignal)  rm -f hassignal
 for i in threadmodule.o  gcmodule.o  regexmodule.o regexpr.o  pcremodule .o
pypcre.o  posixmodule.o  signalmodule.o  _sre.o  arraymodule.o	cmathmodule.o
 mathmodule.o  stropmodule.o  structmodule.o  timemodule.o  operator.o 
_codecs module.o  unicodedata.o unicodedatabase.o  ucnhash.o  _localemodule.o
 fcntlmodu le.o  pwdmodule.o  grpmodule.o  errnomodule.o  selectmodule.o 
socketmodule.o	m mapmodule.o  md5module.o md5c.o  shamodule.o	rotormodule.o
 newmodule.o  binasc ii.o  parsermodule.o  cStringIO.o	cPickle.o config.o
getpath.o main.o getbuildi nfo.o; do \	if test "$i" = "signalmodule.o"; then
\  echo yes >hassignal; break; \  fi; \  done  cd Parser ; make OPT="-g -O2
-Wall -Wstrict-prototypes" VERSION="2.0" \  prefix="/usr/local"
exec_prefix="/usr/local" all  gcc -g -O2 -Wall -Wstrict-prototypes
-I./../Include -I.. -DHAVE_CONFIG_H  -c pgenmain.c In file included from
../Include/pgenheaders.h:24,  from pgenmain.c:16: ../Include/pymem.h:7:
pyport.h: No such file or directory *** Error exit code 1

Stop.
*** Error exit code 1

Stop.


thanks for the help

In article <913ett$u0b$1 at nnrp1.deja.com>,
  Brad <blosh at tsbrass.com> wrote:
> We are running HPUX 10.20 and I was wondering why I cant get python to
> install.
>
> ./configure seems to work
>
> make does not...it gives the following error
>
>   cd Parser ; make OPT="-O" VERSION="2.0" \  prefix="/usr/local"
> exec_prefix="/usr/local" all  cc -Ae -O -I./../Include -I.. -DHAVE_CONFIG_H
> -c pgenmain.c (Bundled) cc: warning 480: The -A option is available only with
> the C/ANSI C pro duct; ignored. (Bundled) cc: warning 480: The -O option is
> available only with the C/ANSI C pro duct; ignored. cpp: "pymem.h", line 7:
> error 4036: Can't open include file 'pyport.h'. *** Error exit code 1
>
> Stop.
> *** Error exit code 1
>
> Stop.
>
> any help is appreciated
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
>


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


From aahz at panix.com  Mon Dec 18 15:00:52 2000
From: aahz at panix.com (Aahz Maruch)
Date: 18 Dec 2000 12:00:52 -0800
Subject: while true: !!!
References: <mailman.977163362.30384.python-list@python.org>
Message-ID: <91lqdk$hba$1@panix6.panix.com>

In article <mailman.977163362.30384.python-list at python.org>,
Tim Peters <tim.one at home.com> wrote:
>
>while "forever, because stupid Guido won't define 'true'":
>    wink

Traceback (innermost last):
  File "<stdin>", line 2, in ?
NameError: wink
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

This is a signature anti-virus.
Please stop the spread of signature viruses!


From pbjorn at uswest.net  Fri Dec 15 18:05:27 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Fri, 15 Dec 2000 16:05:27 -0700
Subject: Source Code Management
References: <fb4f3tsgffnmvme64aol1cn397te7ln349@4ax.com> <91aq96$tba$1@nnrp1.deja.com>
Message-ID: <3A3AA3B7.3AC3F4BF@uswest.net>

Do you have a link to how to get the result in Python marshalled dictionary
format? (I looked at the perforce website but couldn't find anything...)

And to the original poster, Perforce is great. Perhaps not as full featured
as e.g. ClearCase, but infinitely simpler to use and maintain, and very
useable over a network.

-- bjorn

gtalvola at NameConnector.com wrote:

> If you don't mind paying money, I believe Perforce fits all your
> requirements.  http://www.perforce.com/  We use it and have nothing
> but good things to say about it.
>
> And it has explicit support for being scripted in Python -- you can get
> the results of its commands in Python marshalled dictionary format,
> which makes scripting Perforce using Python pretty easy.
>
> In article <fb4f3tsgffnmvme64aol1cn397te7ln349 at 4ax.com>,
>   dale at out-think.NOSPAM.co.uk wrote:
> > I'm looking for a source code management system for looking after
> Python, Java, Javascript and
> > possibly other languages too.
> >
> > Does anyone know of such a system that matches or comes close?
> >
> > Essential features:
> > * Runs on Windows 2000
> > * Works over the Internet (ftp would be good)
> > * Secure (ftp level security would pobably suffice)
> > * Module level version control with deltas or compressed archives
> > * Package level (multiple modules) sign in/sign out
> > * System level release control. (ie, this is now release x, note all
> module versions for this
> > release.)
> >
> > Desirable features:
> > * Extendible - preferably scriptable with Python
> > * Manage arbitrary files including documentation
> > * GUI or command line interfaces
> > * Low maintenance
> > * Library stored as plain files and not in some odd database format
> >
> > That's not too much to ask, is it!  :-)
> >
> > Thanks
> > --
> > Dale Strickland-Clark
> > Out-Think Ltd
> > Business Technology Consultants
> >
>
> Sent via Deja.com
> http://www.deja.com/



From see at below  Mon Dec  4 21:17:50 2000
From: see at below (Paul Foley)
Date: 05 Dec 2000 15:17:50 +1300
Subject: "inherited" keyword in Python?
References: <3A2B9EFA.3A64C13B@schlund.de> <90gevs02uu1@news2.newsguy.com>
Message-ID: <m2elznipbl.fsf@mycroft.actrix.gen.nz>

On Mon, 4 Dec 2000 16:58:18 +0100, Alex Martelli wrote:

> "Carsten Gaebler" <cg at schlund.de> wrote in message
> news:3A2B9EFA.3A64C13B at schlund.de...
>> Hi there!
>> 
>> Does Python know something like the "inherited" keyword from Borland
>> Pascal? In Pascal it is used to call a method of a class's superclass,
>> i.e. instead of saying "superclass.method()" you say "inherited
>> method()". Can be quite helpful if you change the name of the superclass
>> or just forgot it. :-)

[...]
> Metaprogramming magick was deemed inappropriate here (as I
> read the group consensus).  One *might* have a global function
> called super(), which tried to gauge the appropriate class
> object to use (in single-inheritance cases, only, of course),

Why "of course"?  If you can find out the name of the method which is
calling super() (which can be done, but I don't remember how), and the
class from which it's being called (I can think of ways to avoid
needing this, if it can't be done), it shouldn't be very hard to make
a super() function that works with multiple inheritance.  Use
cpl(class) [see below] to get a properly ordered list of base classes,
and search for the first one that implements the named method.  Call
that.  [You need to know which class the current method belongs to in
order to get the right cpl, otherwise it'll infloop...but that isn't
obtainable, you could maybe have super() maintain a stack to tell it
where it's at, or something...]


def cpl(klass):
    if klass.__dict__.has_key('__cpl__'):
        return klass.__cpl__
    classes = get_heirarchy(klass)
    constraints = []
    for i in map(local_precedence_ordering, classes):
        constraints = constraints + i
    constraints = remove_duplicates(constraints)
    klass.__cpl__ = topological_sort(classes, constraints, cpl_tie_breaker)
    return klass.__cpl__

def topological_sort(elements, constraints, tie_breaker):
    elements = elements[:]
    constraints = constraints[:]
    result = []
    while 1:
        rhs = map(lambda x: x[1], constraints)
        elts = filter(lambda x, rhs=rhs: x not in rhs, elements)
        if not elts:
            if not elements:
                return result
            raise ValueError, "Inconsistent constraints in topological_sort"
        if len(elts) == 1:
            elt = elts[0]
        else:
            elt = tie_breaker(elts, result)
        result.append(elt)
        elements.remove(elt)
        constraints = filter(lambda x, elt=elt: x[0] != elt and x[1] != elt,
                             constraints)

def cpl_tie_breaker(elements, current):
    cpl = current[:]
    cpl.reverse()
    for klass in cpl:
        for i in klass.__bases__:
            if i in elements:
                return i

def local_precedence_ordering(klass):
    if not klass.__bases__:
        return []
    result = []
    classes = [klass] + list(klass.__bases__)
    for i in range(0, len(classes)-1):
        result.append([classes[i], classes[i+1]])
    return result

def get_heirarchy(klass):
    result = [klass]
    for i in map(get_heirarchy, klass.__bases__):
        result = result + i
    return remove_duplicates(result)

def remove_duplicates(list):
    result = []
    for i in list:
        if i not in result:
            result.append(i)
    return result

-- 
C/C++/Java/Perl/etc are for people who want to make things that work.
Common Lisp is for people who want to make things that don't break.
                                                           -- Erik Naggum
(setq reply-to
  (concatenate 'string "Paul Foley " "<mycroft" '(#\@) "actrix.gen.nz>"))


From jepler at inetnebr.com  Mon Dec 11 08:30:35 2000
From: jepler at inetnebr.com (Jeff Epler)
Date: Mon, 11 Dec 2000 13:30:35 GMT
Subject: [ANNOUNCE] PyHTML 0.5 -- Indentation-structured HTML generator
References: <slrn92v6h6.6vl.jepler@potty.housenet>
Message-ID: <slrn939lnr.6bm.jepler@potty.housenet>

PyHTML Indentation-structured HTML generator with all the power of Python
                                   Version 0.5, released 10 December 2000
					 Jeff Epler <jepler at inetnebr.com>


PyHTML is an extension of the Python grammar, plus a modified compiler
which permits the use of HTML-style tags as Python blocks, and which makes
these tags and expression-statements write their values as output, making
it easy to write HTML document generators with the block-indentation
style of Python.

This is a rough-cut version of pyhtml, intended to get a reaction from
the Python community.  Please let me know your thoughts, either in the
comp.lang.python newsgroup or via direct mail.  As updated versions
become available, they will be placed at
	http://incolor.inetnebr.com/jepler/pyhtml/

Thanks to the authors of Quixote: A.M. Kuchling, Neil Schemenauer,
and Greg Ward for both the inspiration of an augmented Python language
for HTML generation and for some of the code in that fine package.

Thanks to the author of SPARK, John Aycock.  His parser toolkit and Python
grammar relieved me of the need to modify core Python to parse pyhtml's
extended syntax.  (However, making SPARK emit asts that could be understood
by the compiler was still a bit of work!  And it still speaks only the
Python 1.5 dialect)

Thanks to everyone who has been involved with the development of Python,
it's the most pleasant language I work in.


Example:
#---- cut here for example1.pyh ----
# This program writes out the lines in README in preformatted HTML,
# calling the htmlesc() function to prevent interpretation of any
# special characters, such as <, >, and &.
#
# execute as follows:
#   python pyhtml.py example1.pyh > example1.html
# and then load example1.html in your browser
from htmlesc import htmlesc

<HTML>
  <HEAD>
    <TITLE> "PyHTML Makes It Easy"
  <BODY>
    <PRE>
      for i in open("README").readlines():
        htmlesc(i)
print str(__output) # Current cruft necessary to see output
#---- example1.pyh ends ----

History:
0.5  10 December, 2000
  SPARK version of Python grammar (Python 1.5 subset) completed and
  included.
  PyHTML requires no modifications to the core Python grammar, and
  all external code is included (from SPARK, Quixote, and Tools/compiler) 

0.1  6 December, 2000
  First version publicly released

Syntax:
compound_stmt: ... | tag_stmt
tag_stmt: '<' NAME [tag_args] '>' suite
tag_args: NAME '=' expr (',' NAME '=' expr)* [',']

How it works:
The grammar for Python is modified to include a "tag statement" which is
valid in all the same places as a 'for' statement (compound_stmt).  
The compiler is extended to emit code both for the tags (my work) and for
expression-statements (Taken from Quixote) to write output to a special object.
The following code:
	<UL>
		for i in range(10):
		<LI> str(i)
would output something like
<UL><LI>0</LI><LI>2</LI>....<LI>9</LI></UL>


From auto103690 at hushmail.com  Fri Dec 29 21:31:38 2000
From: auto103690 at hushmail.com (Jeff)
Date: Sat, 30 Dec 2000 02:31:38 GMT
Subject: Is 100,000 entries big for a dictionary?
References: <92iug4$vqb$1@nnrp1.deja.com> <92j58k$55c$1@nnrp1.deja.com>
Message-ID: <eOb36.78936$df5.2813761@news1.crdva1.bc.home.com>

A non-text attachment was scrubbed...
Name: not available
Type: text/plain  charset=iso-8859-1
Size: 1342 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20001230/8c047269/attachment.bin>

From neelk at alum.mit.edu  Sun Dec 31 17:11:04 2000
From: neelk at alum.mit.edu (Neelakantan Krishnaswami)
Date: 31 Dec 2000 22:11:04 GMT
Subject: ML & Python question
References: <t4uqtfi3fv7sf2@corp.supernews.com>
Message-ID: <slrn94vfjg.4av.neelk@alum.mit.edu>

On Sun, 31 Dec 2000 12:27:06 -0500, Jason Cunliffe <jasonic at nomadicsltd.com>
wrote:
>
> Please, Can anyone here provide a intro to what SML is and how
> Python relates to it?
>
> This is really not intended as Python vs. flame bait question - I am
> interested to expand my progamming horizons and am hoping people
> here might help me to do so.

The very short answer is that Python is a very high level, dynamically
typed OO language, and that standard ML is a very high level,
statically typed functional language. Both are IMO incredibly fun to
program in.

The high-level part is straightforward. In both languages, the
language takes care of details that you would have to manage by hand
in lower-level languages. For example, memory is garbage-collected in
Python and SML, and neither language makes you manually track
pointers, array bounds, and the like -- all that is handled
automatically

The OO versus functional part is also pretty straightforward. In
functional languages programmers tend to avoid using assignment and
mutable state, whereas in OO languages one uses state, but
encapsulates it inside objects. 

The statically typed and dynamically typed part is a bit harder to
explain, and IMO the most fundamental difference between SML and
Python. A couple of examples might be in order. In Python, every value
has a type associated with it:

>>> type("foo")
<type 'string'>
>>> type(3)
<type 'int'>

Whenever you evaluate an expression, Python checks the types at
runtime, and then complains if the types don't work together:

>>> 3 + 5
8
>>> string.join(["foo", "bar"], '')
'foobar'
>>> 3 + "foo"
Traceback (innermost last):
  File "<stdin>", line 1, in ?
TypeError: number coercion failed

In SML you might see something like this, 

- 3 + 5;
> val it = 8 : int

- String.concat ["foo", "bar"];
> val it = "foobar" : string

- 3 + "foo";
! Toplevel input:
! 3 + "foo";
!     ^^^^^
! Type clash: expression of type
!   string
! cannot have type
!   int

This looks almost the same, but there's an important difference: in
Python, the type error is detected at runtime, and in SML, the type
error is detected at compile time, and the compiler quits before ever
trying to evaluate the expression. You simply can't write a type
incorrect SML program.

This is why your friend said that ML is safer -- there is a whole
category of errors that are caught earlier in the programming process.
In languages with very powerful type systems (like SML), very complex
errors can be caught with the type system.

However, the flip side is that Python is a bit more flexible; for
example, the following code in Python works fine:

try:
    b = 3 + "foo"
except TypeError:
    b = "something_odd"

This is a goofy example, but if you need to load and unload complex
objects (such as pieces of code) from and into files and databases,
then dynamic typing makes life much easier.


In short, learning both languages is a good idea. A tutorial in
standard ML (at about the difficulty level of the Python tutorial) can
be found at:

  http://www.cs.cmu.edu/afs/cs/usr/rwh/public/www/introsml/

Standard ML has a number of excellent implementations. One of the 
best is Standard ML of New Jersey, which can be found at:

  http://cm.bell-labs.com/cm/cs/what/smlnj/

Most importantly, have fun. :)


Neel


From tim.one at home.com  Mon Dec 11 16:09:31 2000
From: tim.one at home.com (Tim Peters)
Date: Mon, 11 Dec 2000 16:09:31 -0500
Subject: global __dict__
In-Reply-To: <Pine.LNX.4.10.10012111601001.487-100000@erde.kosmos.all>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEFNIDAA.tim.one@home.com>

[posted & mailed]

[Carsten Geckeler]
> I was wondering if there is a __dict__ dictionary for the global
> namespace as for classes and modules.

Your terminology is a bit off:  you're *always* executing in a module in
Python.  "global namespace" has no other meaning than the namespace of the
module you're currently executing in.

> I was trying the following
>
> >>> var = 1
> >>> __dict__
>
> and expected that __dict__ would contain something like {..., 'var': 1},
> but there is no global __dict__.  Where can i find the global
> dictionary?

You need the module's __dict__.  In an interactive shell, you're running in
a module Python creates for you, named "__main__".  So:

>>> var = 1
>>> import __main__
>>> __main__.__dict__
{'__main__': <module '__main__' (built-in)>,
 '__doc__': None,
 'var': 1, # <------------------------------------- here's your vrbl "var"
 '__name__': '__main__',
 '__builtins__': <module '__builtin__' (built-in)>}
>>>

Somewhat easier is to use the globals() builtin function (which returns the
__dict__ for the module you're currently in):

>>> __main__.__dict__ is globals()
1
>>>





From syver.enstad at sensewave.com  Sun Dec  3 12:30:44 2000
From: syver.enstad at sensewave.com (Syver Enstad)
Date: Sun, 3 Dec 2000 18:30:44 +0100
Subject: Python programs (or scripts) as drop targets
References: <hv5k2tg9cdf33j8v333qpkk7jo45ht596n@4ax.com>
Message-ID: <90h0gq$cu0$1@troll.powertech.no>

> So is there anyone out there with the VC++ skills sufficiently interested
in this idea to code the
> necessary interfaces?

If there isn't support for doing this in python I could look into it. I've
made a couple of shell extensions before that sounds pretty similar so it
shouldn't be too much work.

Can't you contact me Dale if we can't find a way to do this in Python?




From aleaxit at yahoo.com  Fri Dec 29 03:17:27 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 29 Dec 2000 09:17:27 +0100
Subject: Python vs. PHP (& Java?)
References: <69BC4C6AE83ED311BC9400805FA7B14AF84D2C@il93exp01.css.mot.com> <mailman.978059571.14515.python-list@python.org>
Message-ID: <92hhak01edr@news2.newsguy.com>

"Moshe Zadka" <moshez at zadka.site.co.il> wrote in message
news:mailman.978059571.14515.python-list at python.org...
> On Thu, 28 Dec 2000, Kemp Randy-W18971 <Randy.L.Kemp at motorola.com> wrote:
>
> > 1.  Could we argue that English is a better language to use around the
world
> >     then French or Spanish?
>
> Yes. Personally, I think that would be wrong -- English has a very
> hard to learn spelling system.

...but it partly compensates with simpler grammar.  Not quite 'a wash',
IMHO, but these things count...

When a given communications-set is only-written, or only-spoken, the
complexities in written<->spoken correspondence-laws matter little.

This is a limit case, sure, but it does happen: you'll likely never
learn just how exotic my English pronunciation is, as long as our
interactions remain confined to email, news, &c; if my spelling is
reasonable, even if I'm unsure about how most of the vowel clusters
are to be spoken, you won't know about my doubts.  Vice versa, I am
unlikely to have any idea about how strong Woody Allen's spelling
skills may be, despite my familiariy with his pronunciation, since I
don't get to see his written materials before copy-editing.

Grammar is more pervasive.  With many English verbs, for example,
you get away with as little as four separate spellings/word forms -
'dream', 'dreams', 'dreamed', 'dreaming' (plus auxiliary utility
words such as 'to' for the infinite, 'will/shall' for the future,
and so on, but those are shared between different verbs).  It is
not _all_ that hard to learn and apply the rules, e.g., "the -s
form is for third person singular present-tense, only" -- they are
also shared between verbs.  Contrast with the situation in Latin
derived languages, including French, Spanish, Italian: typically
six forms (depending on 1st/2nd/3rd person, singular/plural) per
tense -- "sogno, sogni, sogna; sognamo, sognate, sognano" -- and
that is just the present indicative tense of 'to dream' in Italian.

Here (as in Latin, and differently from French) subject-pronoun is
mostly omitted, so you have to recognize the verb form in order to
even understand WHO I'm saying is dreaming -- 'sognamo' (we dream)
and 'sognano' (they dream) being a well-known source of difficulty
for foreign-mother-tongue speakers, particularly in the written
form (the distinction relies just about totally on syllable stress,
'sOgnano' for the 3rd person plural vs 'sognAmo' for the 1st person
plural -- people whose mother-tongue doesn't give syllable stress
much or any phonematic significance have a particularly hard time).


OK, just one simple example.  It is one defensible thesis that
each natural language tends to evolve to some (vaguely defined)
"constant overall complexity" -- if historical evolution leads to
some simplification somewhere, compensating complexities appear
elsewhere (although I guess _causation_ could be the other way
'round!-).  Think of it as a linguistic version of Parkinson's
law about the time it takes for any given task tending to expand
or contract to fit just about the time that is available for it...
for example, modern living languages tend to have many examples
of simplification (along various different axes) compared to
what we know of their ancestor-forms -- one possibility is that
this is compensated by a vastly expanded vocabulary (particularly,
passive-vocabulary) of today's typical speaker when compared with,
say, the typical speaker of 1000 or 2000 years ago (here I'm
threading on *particularly* contentious ground, so I strongly
urge all readers NOT to be content with my glib summaries, but
to read up on their own about these issues, if they have any
interest for them).


Alex





From loewis at informatik.hu-berlin.de  Fri Dec  8 15:55:27 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 08 Dec 2000 21:55:27 +0100
Subject: Link problems embedding python in Linux. HELP!
References: <90r5kf$q9j$1@nnrp1.deja.com>
Message-ID: <p6qr93iiqf4.fsf@informatik.hu-berlin.de>

jvanar1 at my-deja.com writes:

> I've read every resource available, but nothing tells me what to link
> with. here is my compilation line:
> 
> gcc -o go -lpython1.5 -L/usr/lib/python1.5 go.c

The order of libraries on the linker line is relevant; place
-lpython1.5 after go.c.

Hope this helps,
Martin


From rturpin at my-deja.com  Mon Dec 25 10:45:08 2000
From: rturpin at my-deja.com (rturpin at my-deja.com)
Date: Mon, 25 Dec 2000 15:45:08 GMT
Subject: One Python 2.1 idea
References: <mailman.977590932.19335.python-list@python.org> <9237md$orm$1@nnrp1.deja.com> <924bjt0e3n@news1.newsguy.com> <925baq$5sj$1@nnrp1.deja.com> <mailman.977737448.26875.python-list@python.org>
Message-ID: <927q24$ond$1@nnrp1.deja.com>

In article <mailman.977737448.26875.python-list at python.org>,
  Peter Schneider-Kamp <peter at schneider-kamp.de> wrote:
> I see ome good reasons for enhancing the "for" statement:
>
> - loop counter iteration will give a speedup by avoiding
>   construction of a list and looping over that list.
> - looping a certain amount of times is a frequent construct

Yes, this is a good performance optimization. BUT it
requires NO CHANGE in how loops are written. The WRONG
way to do this is to provide another construct in the
language, making it more complex, but giving it no
more expressive power. The RIGHT way to do this is
for the compiler to recognize the common construct:

    for i in range(<exp1>: <exp2>):

The compiler then generates efficient byte code that
does precisely what you describe. It can generate a
minimal "ghost" of a list for the iteration, because
there is no way that future code can do anything else
with that list.

Russell


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


From aleaxit at yahoo.com  Mon Dec 11 14:58:12 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 11 Dec 2000 20:58:12 +0100
Subject: Python2 distutils question: how to best "autoconfig"...?
References: <9130e60obc@news2.newsguy.com> <82aZ5.3458$Kd1.395204@newsb.telia.net>
Message-ID: <913br4012r8@news2.newsguy.com>

"Fredrik Lundh" <fredrik at effbot.org> wrote in message
news:82aZ5.3458$Kd1.395204 at newsb.telia.net...
> Alex Martelli wrote:
> > The specific issue is "how many bits of significance
> > does a Python float have".  (Or is there somewhere in
> > the Python headers a #define for this that I missed...?)
>
> why not just do things on the Python side?  if not else,
> it's a sure way to avoid optimization head-aches:

Good point, which Tim Peters also did in his mail!  I can
only claim a brainstorm on my part, I guess:-).


Alex





From quinn at dinar.ugcs.caltech.edu  Sat Dec 16 21:06:55 2000
From: quinn at dinar.ugcs.caltech.edu (Quinn Dunkan)
Date: 17 Dec 2000 02:06:55 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <r0pzoi11lhz.fsf@black.cs.washington.edu> <5jvZ5.3588$Kd1.418770@newsb.telia.net> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com> <kuu28971kp.fsf@lasipalatsi.fi> <3A380E4E.8537A0FB@san.rr.com> <91a1vr02b02@news2.newsguy.com> <Co9_5.25465$x6.13756579@news2.rdc2.tx.home.com> <91bbq102mtk@news1.newsguy.com> <rWd_5.26820$x6.14119694@news2.rdc2.tx.home.com> <91d5el0me7@news1.newsguy.com> <rOr_5.31120$x6.15241856@news2.rdc2.tx.home.com>
Message-ID: <slrn93o7so.tu0.quinn@dinar.ugcs.caltech.edu>

On Fri, 15 Dec 2000 16:38:47 GMT, Rainer Deyke <root at rainerdeyke.com> wrote:
>Yes, in-place modification should be supported.  I had something like this
>in mind:
>
>class UserInt:
>  def __init__(self, value):
>    self.data = value
>  def __add__(self, other):
>    if isinstance(other, UserInt):
>      return self.__class__(self.data + other.data)
>    else:
>      return self.__class__(self.data + other)
>  __radd__ = __add__
>  def __iadd__(self, other):
>    if isinstance(other, UserInt):
>      self.data += other.data
>    else:
>      self.data += other
>  # Repeat for all other operations
>
>UserInt as written is more properly UserNumber, since it supports longs ints
>and floats.

Just as a matter of style, would it be superior to add a __coerce__ method
that knows how to coerce ints, longs, and floats into UserInts, or make an
explicit check on each method?  __coerce__ saves a lot of typing, but I'm
generally wary of implicit things like coercion.  When I implemented a class
like this (a vector, actually), I used __coerce__, but after trying to wrap my
brain around the control flow, I was thinking maybe I wouldn't next time, even
though it seems to work fine.

Is this fear unfounded?  Is there a better way?


From z3penguin at penguinpowered.com  Thu Dec  7 20:29:47 2000
From: z3penguin at penguinpowered.com (Z 3 Penguin)
Date: Thu, 7 Dec 2000 20:29:47 -0500
Subject: problems with urllib in Python20?
Message-ID: <90pdir$k7$1@bob.news.rcn.net>

whenever i try to use urllib.urlopen() in Python20/Win32, i get this error:
-START-
Fatal Python error: Interpreter not initialized (version mismatch?)

abnormal program termination
-END-
it also gives the same thing with urllib2.  any ideas?
-z




From loewis at informatik.hu-berlin.de  Sat Dec  9 05:55:13 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 09 Dec 2000 11:55:13 +0100
Subject: sgmlop binaries
References: <90sg8k$7cc$1@bunyip.cc.uq.edu.au>
Message-ID: <p6qn1e5ampa.fsf@informatik.hu-berlin.de>

"Chui Tey" <teyc at bigfoot.com> writes:

> Are there any procompiled binaries for sgmlop for python 1.52 on the net?

There are precompiled binaries included in PyXML 0.6.2, which is
available from sourceforge.net/projects/pyxml.

Regards,
Martin



From erikd at zip.com.au  Sun Dec 31 02:32:22 2000
From: erikd at zip.com.au (Erik de Castro Lopo)
Date: Sun, 31 Dec 2000 07:32:22 GMT
Subject: tftplib ?
References: <Pine.LNX.4.21.0012310129140.9173-100000@saturn.ispol.com>
Message-ID: <3A4EDF22.30F29C1@zip.com.au>

Gregory Trubetskoy wrote:
> 
> Before I write my own -
> 
> is there by a chance a tftplib.py or something that implements tftp in
> Python out there?

What tftp and how does it differ from ftp?

Erik
-- 
+-------------------------------------------------+
     Erik de Castro Lopo     erikd at zip.com.au
+-------------------------------------------------+
"There is no reason why anyone would want a computer in their home"   
Ken Olson, DEC, 1977


From dbrueck at edgix.com  Tue Dec 19 14:04:06 2000
From: dbrueck at edgix.com (Dave Brueck)
Date: Tue, 19 Dec 2000 12:04:06 -0700
Subject: pow(-1,0) = 1, -1**0 = -1, why?
In-Reply-To: <7lav3t4ptquheuah74s5ff10ghg54u2dhf@4ax.com>
Message-ID: <020301c069ef$696a0570$760214ac@PRODUT2KDAVE>

Operator precedence:

>>> (-1)**0
1

>>> -(1**0)
-1

-Dave

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Kirby Urner
> Sent: Tuesday, December 19, 2000 11:44 AM
> To: python-list at python.org
> Subject: pow(-1,0) = 1, -1**0 = -1, why?
> 
> 
> 
>  Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)] on win32
>  Type "copyright", "credits" or "license" for more information.
>  IDLE 0.6 -- press F1 for help
>  
>  >>> -1**0
>  -1
>  >>> pow(-1,0)
>  1
>  >>> -10**0
>  -1
>  >>> pow(-10,0)
>  1
> 
> 
> These are inconsistent.  I think pow() is giving the right answer.
> 
> Kirby
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list



From syver.enstad at sensewave.com  Mon Dec  4 20:35:33 2000
From: syver.enstad at sensewave.com (Syver Enstad)
Date: Tue, 5 Dec 2000 02:35:33 +0100
Subject: Using COM objects as dictionaries in win32com.
References: <90c9gj$nkn$1@troll.powertech.no> <3A2ACBE7.7030402@ActiveState.com>
Message-ID: <90i1rc$o2v$1@troll.powertech.no>

<M. Hammond>
so it is in!
</M.Hammond>

Cool!

I fiddling with an implementation of items() in CDispatch, so that one could
use f.ex
the Session.Contents object in ASP like this:

for key, value in Session.Contents.items:
    #do stuff

There some more work to do there as strange things happen when I apply the
method to objects that don't really support it. I guess another problem is
that I've implemented as a method in CDispatch (in dynamic.py) and then
won't work on make_py generated classes won't it.

Anyway, my fiddlings so far has produced this, that works with
Session.Contents and a couple of other objects, but hangs the web server
when applied to the Request object (works fine on Request.QueryString and
Request.Form though)

    def items(self):
        ret = []
        for each in self:
            ret.append( each, self[each]) )
        return tuple(ret)





From aleaxit at yahoo.com  Wed Dec 27 17:23:50 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 27 Dec 2000 23:23:50 +0100
Subject: classes (was Re: Same again please for OOP)
References: <mailman.977623149.8184.python-list@python.org> <9257mf0sl1@news1.newsguy.com> <roy-E5B925.13324224122000@news.panix.com> <925khi019hs@news1.newsguy.com> <kubsu1t81x.fsf@lasipalatsi.fi> <92708d0g4g@news1.newsguy.com> <mailman.977952911.9681.python-list@python.org>
Message-ID: <92dqb0020hk@news2.newsguy.com>

"D-Man" <dsh8290 at rit.edu> wrote in message
news:mailman.977952911.9681.python-list at python.org...
    [snip]
> However, I can't get Alex's example code to work.  I either get
> AttributeError or an infinitely recursive __getattr__.  The reason is
> if I make a class Foo that has a function __get_value, the name is
> mangled to _Foo__get_value.  If I reduce the leading underscores to
> only 1 underscore, it works fine.
>
> Have I missed something here Alex?

No, sorry, my fault -- two leading underscores without two trailing
ones do of course get mangled that way!  'Just one' works.


Alex





From frank.neuber at gmx.de  Wed Dec  6 14:44:30 2000
From: frank.neuber at gmx.de (Frank Neuber)
Date: Wed, 6 Dec 2000 20:44:30 +0100
Subject: setsockopt and urllib
References: <slrn92moue.hc.frank.neuber@mars.private.de> <3A2DD408.D4E0B8E3@pehr.net>
Message-ID: <slrn92t5ou.tf.frank.neuber@mars.private.de>

In article <3A2DD408.D4E0B8E3 at pehr.net>, pehr anderson wrote:

Hi all,
thanks for this link below ...
>http://www.timo-tasi.org/python/
This is exactly what i want :-)

I only did a little change or add in the Timeout(Exception) class to provide 
an easily access to the timeout message.

class Timeout(Exception):
   def __init__(self, message):
     self.message=message

I think this was necessary for me to use and reuse the message text in 
an error popup window.

>> Hi group,
>> I have to set special timeouts on receive and send calls in conjunction with
>> urllib-calls.
>> Is there a simple solution to do this without modifying the urllib itself.
>> For Example:
>> I wish to set the timeout to 60 s for the following command:
>> (tmpfile, hdr) = urllib.urlretrive(urlname)
>> It is possible to to this with the socks module call
>> setsockopt(SOL_SOCKET,SO_RCVTIMEO, ttime) over the urllib module.

-- 
     _/_/_/_/ _//   _/ Frank Neuber
    _/       _/_/  _/  frank.neuber at gmx.de (private)
   _/_/_/   _/ _/ _/
  _/       _/  _/_/    neuber at opensource-systemberatung.de
 _/       _/    // http://www.opensource-systemberatung.de


From gdd0 at gte.com  Fri Dec 15 13:49:55 2000
From: gdd0 at gte.com (Gary D. Duzan)
Date: 15 Dec 2000 18:49:55 GMT
Subject: Is Stackless and uthread.py SMP friendly?
References: <91br9v$qa4$1@nnrp1.deja.com> <ku1yva5ocq.fsf@lasipalatsi.fi>
Message-ID: <91dp4j$5fj$1@news.gte.com>

In article <ku1yva5ocq.fsf at lasipalatsi.fi>,
Erno Kuusela  <erno-news at erno.iki.fi> wrote:
>In article <91br9v$qa4$1 at nnrp1.deja.com>, jbalding at hotmail.com writes:
>
>| will Stackless and uthread.py make my SMP
>| experience with Python a happy one?
>
>it should work just fine, since microthreads don't use os threads.

   True, but I believe the real question was whether microthreads
could run on different CPUs, to which I believe the answer is "no".
It will work fine, but only on one CPU per process.

					Gary Duzan
					Verizon Laboratories




From schaffer at optonline.net  Tue Dec 26 10:47:52 2000
From: schaffer at optonline.net (Les Schaffer)
Date: 26 Dec 2000 10:47:52 -0500
Subject: Can't load jpeg image with PIL (win32)
References: <9241ls$ks3$1@news1-2.kornet.net> <874rzsh2g9.fsf@optonline.net> <92acsd$4ni$1@news1-2.kornet.net>
Message-ID: <87y9x3w5ef.fsf@optonline.net>

>     raise IOError, "cannot identify image file"
> IOError: cannot identify image file

hmmmm...

1.) i tested the same functionality on my system and it works fine

2.) are you sure 02.jpg is a __jpeg__ image???  ;-)  

    try opening lena.jpg

    


les schaffer


From aleaxit at yahoo.com  Sun Dec 24 14:53:02 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 24 Dec 2000 20:53:02 +0100
Subject: classes (was Re: Same again please for OOP)
References: <mailman.977623149.8184.python-list@python.org> <9257mf0sl1@news1.newsguy.com> <roy-E5B925.13324224122000@news.panix.com>
Message-ID: <925khi019hs@news1.newsguy.com>

"Roy Smith" <roy at panix.com> wrote in message
news:roy-E5B925.13324224122000 at news.panix.com...
> "Alex Martelli" <aleaxit at yahoo.com> wrote:
> > For example, suppose our program will need to model CD's.  A CD
> > object has several attributes -- depending on our purposes, we may
> > want to model some appropriate subset of them; for example, a
> > "title", a "price" in Euros, and a sequence of "tracks" (each of which
> > has its own attributes, say a "title" and a "duration" in seconds).
>
> That's good as far as it goes, but I think you missed the real power of
> classes, which is subclassing.

I disagree that subclassing is 'the real power' -- it's just one handy
way, one of many, of reusing some code.  You may be confusing
subclassing with *polymorphism*, which IS indeed _awesomely_
powerful... but, in Python, need not have anything to do with the
'subclassing' implementation technique (an easy confusion to make
if coming from languages, such as C++ and Eiffel, which almost
equate polymorphism with subclassing... Java, which singles out
interfaces as its preferred road to polymorphism, may lead to some
more conceptual clarity about this distinction!).


'As far as it goes' were 324 lines, and I couldn't do justice to the
_behavior_ concept at all within those confines, let alone the key
case of _polymorphic_ behavior.  Particularly because ideal Python
style, IMHO, needs _much_ clarity about the state/behavior
distinction, one which is often missing, witness...:

> The mp3 class might return a URL for its location() method, and the CD
> or Vinyl classes might return a string, "Third shelf from the bottom,

Looks like a typical case of state mismodeled as behavior.  Why should
'location' be a method -- what does it *DO*?!  It doesnt _do_ anything;
rather, it's an 'accessor' for a part of the object's state.  A typical
'implementation' would be, say:
    def location(self):
        return self.the_location
where we can't give that part-of-state its 'normal' name, which would
be just 'location', since the method's name preempts it.  In the crucial
spirit of "do the simplest thing that can possibly work", modeling state
as state (attributes) is often best in Python -- with __getattr__ and
__setattr__ in reserve if we later refactor it in more sophisticated ways
(to meet some specific application need) and want to insulate client
code from the effects of such changes.

Hardly a suitable showcase for polymorphism -- or, actually, for
behavior *at all*.

One can do better, of course.  But, didactically, I think that it's much
better to start by ensuring a good understanding of class instances
as handy holders of *state*, with 'behavior' initially limited to that
connected with initialization, before moving on to more general issues
of behavior, and, from there, to polymorphism (and to subclassing in
various uses -- the purely utilitarian use of mixin inheritance is, I
believe, a suitable introduction, _before_ proceeding with subtler and
more refined issues related to the interplay of inheritance and
polymorphism; it's nice that Python lends itself so cleanly to a
presentation of these issues as separate -- in fact, it's part of what
makes it such an ideal language for _teaching_ OO, quite apart from
the fact that it's also ideal for _doing_ it:-).

And I wouldn't present such instruction in posts of more than a
few hundreds lines at a time, because even the simplest way to
present things still needs the reader to digest each step properly,
before moving on...


Alex





From bayinnaung at my-deja.com  Fri Dec 15 16:44:12 2000
From: bayinnaung at my-deja.com (bayinnaung at my-deja.com)
Date: Fri, 15 Dec 2000 21:44:12 GMT
Subject: Python advocacy
References: <91bln3$lmr$1@nnrp1.deja.com> <3A39AC3E.FA75C4FF@engcorp.com>
Message-ID: <91e3bb$kua$1@nnrp1.deja.com>



In article <3A39AC3E.FA75C4FF at engcorp.com>,
  Peter Hansen <peter at engcorp.com> wrote:

Regarding language advocacy...
>There is no reason to focus on the fact that people
>appear to take this sort of thing personally.
>That's probably just human nature,
>and likely not even a Bad Thing.

As far as it constitutes *blindness* it does seem like a *bad thing*:

Take for instance continuations, coroutines, and micro-threads,
a hot topic in the Python world.
There are small implementations with thorough descriptions
in the Scheme world that can be experimented with immediately:
   Essentials of Programming Languages by Friedman, Wand, and Haynes
   Lisp in Small Pieces by Christian Queinnec

The book on compiling continuations is written in ML:
   Compiling with Continuations by Andrew J. Appel

In the web design world Flash "movie clips" are a co-routine-like
potential gold mine of examples for future Python.

The canvas and text widget interfaces of tkinter (the tk library)
provide a great guide to what can be done with JFC classes.
Much of Topley's advanced JFC book is trivial in tkinter.

What about Lisp's meta-object protocol? There are echoes of that
in the C++ design pattern world.

What about Microsoft's .NET separation of run-time and language
(e.g. regexes), the designer himself goes as far to say that
syntax/language might be construed as a programmer preference
in the future:
   http://windows.oreilly.com/news/hejlsberg_0800.html

Taking things personally makes it difficult to think
in a cross-language sort of way.

>the author is oblivious to the huge investment people
>put into developing systems in a particular language
>(not to mention the investment in *learning* that language)
>and the fact that as long as the language stays popular,
>improves, and continues to serve their needs well,
>they get a payoff from that investment.

How could opening one's eyes to what's happening in
the programming world outside one's language
decrease payoff?...

...it might even help prevent extinction ala Lisp.

What about MS .NET's split between language and runtime.
or the actual changing of the instruction set of the virtual machine
to support generic programming?

These are bound to change the way scripting languages are put together
even though a true language advocate would probably put his fingers in
his ears when he heard the word "Microsoft" , ".NET", or "C#" .

The thing that struck me most about that paper was how
the same people who are logically exacting in their trade:
programming in a specific language, are not logically exacting when
listening to someone talking about something bigger, more inclusive
than their trade, namely *programming in general*.

Jon Fernquest
bayinnaung at hotmail.com


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


From greg at cosc.canterbury.ac.nz  Mon Dec 11 19:14:00 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Tue, 12 Dec 2000 13:14:00 +1300
Subject: Dumb Newbie Question
References: <3A327D01.9841EF98@drew.edu>
Message-ID: <3A356DC8.7837AF38@cosc.canterbury.ac.nz>

Scott Weisel wrote:
> 
> sys.path = sys.path + "c:\\python16\\programs

You're trying to add a list and a string.
You want to either do

   sys.path = sys.path + ["c:\\python16\\programs"]

or, more compactly,

   sys.path.append("c:\\python16\\programs")

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


From entropiamax at jazzfree.com  Thu Dec 28 09:36:30 2000
From: entropiamax at jazzfree.com (entropia)
Date: Thu, 28 Dec 2000 15:36:30 +0100
Subject: Interesting behaviour of the assignment
References: <92f4gk$qk6$2@news.nuri.net>
Message-ID: <3A4B4FED.E2397AE4@jazzfree.com>

On Python2.0-win I obtain the same for every namber >=100

June Kim escribi?:

> >>> a=100
> >>> b=100
> >>> a is b
> 0
> >>> a=99
> >>> b=99
> >>> a is b
> 1
>
> Is this an intended/expected result? What is the lying structure beneath
> this?
>
> --
> http://www.python.org/mailman/listinfo/python-list




From jon at bezek.dstc.monash.edu.au  Mon Dec 11 22:55:12 2000
From: jon at bezek.dstc.monash.edu.au (Jonathan Giddy)
Date: 12 Dec 2000 14:55:12 +1100
Subject: Backslash escape in regular expressions
References: <jon.976507791@bezek.dstc.monash.edu.au> <3A34D69C.BAC68E27@engcorp.com>
Message-ID: <jon.976591746@bezek.dstc.monash.edu.au>

Peter Hansen <peter at engcorp.com> writes:

] Jonathan Giddy wrote:
] > 
] > According to the re module documentation, backslash either escapes special
] > characters, or signals a special sequence.  The special sequences are
] > then listed.
] > 
] > However, as this code shows, there are some special sequences (mainly
] > the whitespace characters) that are special, but aren't listed.  Is this
] > a lapse in the re implementation or the re documentation?  Can I safely
] > expect re.compile(r'\(hello\)\n') to always match '(hello)\n' (the current
] > behaviour) and not match '(hello)n' (the documented behaviour?)
] 
] The documentation I have clearly shows that \\ is a special sequence
] which turns into the backslash character itself.  

I agree that the documentation (Section 4.2.1 of the Library Reference)
states this.  But you're paying too much attention to the code <0.5 wink>.
\\ is irrelevant to the problem, as apart from the code sample, \\ 
doesn't appear in the problem description.

] With the raw-string
] form with 'r' your "current behaviour" above *is* the documented
] behaviour, isn't it?  At least, just using those strings with "print"
] shows that you don't get "(hello)n"...

Consider re.compile(r'\y').  \y is clearly never a special sequence.  Since 
this is a raw string, the re module gets a string with the two characters
'\' and 'y'.  By my reading of the re module documentation, \y should match 
a plain y, which it does.

Now consider re.compile(r'\n').  \n is normally a newline, but inside a 
raw string, it is actually the two characters '\' and 'n'.  By my reading
of the re module documentation, \n is not a "special character escape" (\*, 
\?, and so forth), since 'n' is not a special character in a regex.  In
addition, it is not a "special sequence", since it does not consist of
'\' and a character from the list in the documentation.  Therefore, the
documentation indicates that, like \y, \n should match a plain n.

So, the regex in the example should match the Python string '(hello)n',
but instead matches the Python string '(hello)\n'.  I prefer the latter,
but think the documentation should indicate that \a, \f, \n, \t, \v, and
\x are also "special sequences".




From pbjorn at uswest.net  Fri Dec 29 15:24:55 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Fri, 29 Dec 2000 13:24:55 -0700
Subject: Bug or Feature?
References: <92ij58$74k3p$1@ID-11957.news.dfncis.de>
Message-ID: <3A4CF317.9F6AB947@uswest.net>

My guess is that it's because eval is a function and exec is a statement
("print print.__doc__ doesn't work either ;-)

-- bjorn

Emile van Sebille wrote:

> Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)]
> on win32
> (and linux current cvs)
>
> This works:
>
> print eval.__doc__
>
> but this doesn't:
>
> print exec.__doc__
>
> --
>
> Emile van Sebille
> emile at fenx.com
> -------------------



From dsh8290 at rit.edu  Thu Dec 28 13:46:36 2000
From: dsh8290 at rit.edu (D-Man)
Date: Thu, 28 Dec 2000 13:46:36 -0500
Subject: Interesting behaviour of the assignment
In-Reply-To: <92f4gk$qk6$2@news.nuri.net>; from junaftnoon@nospamplzyahoo.com on Thu, Dec 28, 2000 at 07:25:14PM +0900
References: <92f4gk$qk6$2@news.nuri.net>
Message-ID: <20001228134636.A15310@westsidecnc.rh.rit.edu>

This is not a bug, but rather a side-effect.  Integers in python are
immutable.  For efficiency, at startup the small integers are created.
Then every variable that has such a value is just a reference to the
constant integer created at startup.  This way you don't have
malloc/free calls and ref-counting to handle each time you do

i = 1
i = i + 1


Apparently 99 is the threshold for the integers created at startup and
the integers created upon demand.  I did a few tests of my own:

a = 99
b = 99
a is b
a == b
id( a )
id( b )
# the same object

a = 100
b = 100
a is b
a == b
id( a )
id( b )
# different objects

list = []
list.append( 100 )
list.append( 100 )
list.append( 100 )
for i in list : 
	print id( i )
# note that each element has a different id

list = []
for i in range( 3 ) :
	list[i] = 100
for i in list :
	print id( i )
# each one is the same here


id() returns the id of the object, which is implementation-dependant.
It currently just returns the address of the object.

It appears that the 'is' keyword checks to see if 2 references refer
to the same object while the '==' operator checks to see if the
objects 2 references refer to have the same value.

I would conclude that when the assignment is made in the loop, it is
optimized to be the same actual object (since it is immutable) but
when the assignment is made independently in separate commands the
objects are created independently.

-D


PS.  I would have included the interpreter's output, but I couldn't
find an easy way to do it.  I am using vim as the editor in a remote
login session without the copy-paste features of X.


On Thu, Dec 28, 2000 at 07:25:14PM +0900, June Kim wrote:
> >>> a=100
> >>> b=100
> >>> a is b
> 0
> >>> a=99
> >>> b=99
> >>> a is b
> 1
> 
> Is this an intended/expected result? What is the lying structure beneath
> this?
> 



From kirschh at lionbioscience.com  Thu Dec 28 03:13:33 2000
From: kirschh at lionbioscience.com (Harald Kirsch)
Date: 28 Dec 2000 09:13:33 +0100
Subject: split hangs
References: <slrn94l1pu.4so.jp@localhost.localdomain> <92e3lm$pni$1@panix3.panix.com> <slrn94l4d3.550.jp@localhost.localdomain>
Message-ID: <yv2n1dhufo2.fsf@lionsp093.lion-ag.de>

jp at ulgo.koti.com.pl (Jacek Pop?awski) writes:

> Aahz Maruch wrote:
> >What is the regex?
> 
> ourregex=re.compile(r'(<(?:[^<>"]+(?:"[^"]*")?)+>)')

Yes, something is very strange with this type of regex. Run the
following program. It shows strange behaviour with 1.5.2 on Linux as
well as with 2.0 on IRIX.

The regex looks harmless. Is this a bug or do we run into some
backtracking problem here?


########################################################################
import re
import sys

boing = re.compile(r'(<(?:[^<>"]+(?:"[^"]*")?)+>)')

s = 'p < 0.05, all data in arbitrary'

for i in range(1, len(s)):
  t = s[0:i]
  boing.search(t)
  print "survived `%s'"%t


print "----------------------------------"
########################################################################

Harald Kirsch


-- 
----------------+------------------------------------------------------
Harald Kirsch   | kirschh at lionbioscience.com | "How old is the epsilon?"
LION Bioscience | +49 6221 4038 172          |        -- Paul Erd?s


From Gareth.McCaughan at pobox.com  Sat Dec 30 16:10:52 2000
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: Sat, 30 Dec 2000 21:10:52 +0000
Subject: Returning something else from __init__ -- was Re: 2001 Enchancement Wishlist
References: <3A4CEACB.220998A0@javanet.com> <Pine.LNX.4.10.10012301008410.5079-100000@kev.ioc.ee>
Message-ID: <slrn94sjqs.f6j.Gareth.McCaughan@g.local>

Pearu Peterson wrote:

> I see that many Python experts are against this proposal but anyway, I
> would like to support this idea with the following examples.
..
> That is, if imaginary part of a Complex number is zero, it's constructor
> would return Real instance:

What's wrong with

    class Complex:
      ...
    def make_complex(re,im):
      if im==0: return re
      return Complex(re,im)

and then

    a = make_complex(1,2)
    b = make_complex(1,0)

There's one thing that bothers me greatly about it:
that function ought to be less global. It would be
nice to be able to define class methods so that you
could say

    a = Complex.make(1,2)
    b = Complex.make(1,0)

I'm not sure whether the reason Python doesn't have
class methods is (1) Guido disapproves, or (2) they
wouldn't fit well with the way Python implements
instance methods. It's not clear to me how one could
fit class methods into Python now.

For now, the only feasible way to encapsulate these
things is in a module. So e.g. you could have a module
"complex" which defines a class "Complex" and a
factory function "make"; you'd probably say

    import complex

    a = complex.make(1,2)

and so on.

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
sig under construc


From joconnor at mail.cybermesa.com  Mon Dec 11 11:56:58 2000
From: joconnor at mail.cybermesa.com (Jay O'Connor)
Date: Mon, 11 Dec 2000 16:56:58 GMT
Subject: while true: !!!  (fwd)
Message-ID: <200012111656.JAA15835@mail.cybermesa.com>

An embedded and charset-unspecified text was scrubbed...
Name: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20001211/5cddb176/attachment.ksh>

From victor at prodigy.net  Wed Dec 27 19:32:19 2000
From: victor at prodigy.net (Victor Muslin)
Date: Thu, 28 Dec 2000 00:32:19 GMT
Subject: fcgi.py
Message-ID: <3a4a865c.1325844015@localhost>

Sorry for newbie questions, but I ran into a couple of problems with
fcgi.py. I am using Python 1.5.2 on WindowsNT 4.0.

1) I tried to run fcgi.py from the command line (to exercise its
_test() function) and received the following error message:

	D:/Apps/apache/cgi-bin> fcgi.py
	Exception exceptions.AttributeError: 'err' in <method
FCGI.__del__ of FCGI instance at 82a220> ignored

The traceback was:

	Traceback (innermost last):
	  File "fcgi.py", line 386, in _test
	    req = Accept()
	  File "fcgi.py", line 231, in __init__
	    _startup()
	  File "fcgi.py", line 367, in _startup
	    s=socket.fromfd(sys.stdin.fileno(), socket.AF_INET,
	AttributeError: fromfd

Seems that the socket module for WindowsNT Python 1.5.2 does not have
fromfd() method.

Here's the proof (at least for my Python installation):

>>> pprint.pprint(dir(socket))
['AF_INET',
 'INADDR_ALLHOSTS_GROUP',
 'INADDR_ANY',
 'INADDR_BROADCAST',
 'INADDR_LOOPBACK',
 'INADDR_MAX_LOCAL_GROUP',
 'INADDR_NONE',
 'INADDR_UNSPEC_GROUP',
 'IPPORT_RESERVED',
 'IPPORT_USERRESERVED',
 'IPPROTO_GGP',
 'IPPROTO_ICMP',
 'IPPROTO_IDP',
 'IPPROTO_IGMP',
 'IPPROTO_IP',
 'IPPROTO_MAX',
 'IPPROTO_ND',
 'IPPROTO_PUP',
 'IPPROTO_RAW',
 'IPPROTO_TCP',
 'IPPROTO_UDP',
 'IP_ADD_MEMBERSHIP',
 'IP_DROP_MEMBERSHIP',
 'IP_MULTICAST_IF',
 'IP_MULTICAST_LOOP',
 'IP_MULTICAST_TTL',
 'IP_OPTIONS',
 'IP_TOS',
 'IP_TTL',
 'MSG_DONTROUTE',
 'MSG_OOB',
 'MSG_PEEK',
 'SOCK_DGRAM',
 'SOCK_RAW',
 'SOCK_RDM',
 'SOCK_SEQPACKET',
 'SOCK_STREAM',
 'SOL_SOCKET',
 'SOMAXCONN',
 'SO_ACCEPTCONN',
 'SO_BROADCAST',
 'SO_DEBUG',
 'SO_DONTROUTE',
 'SO_ERROR',
 'SO_KEEPALIVE',
 'SO_LINGER',
 'SO_OOBINLINE',
 'SO_RCVBUF',
 'SO_RCVLOWAT',
 'SO_RCVTIMEO',
 'SO_REUSEADDR',
 'SO_SNDBUF',
 'SO_SNDLOWAT',
 'SO_SNDTIMEO',
 'SO_TYPE',
 'SO_USELOOPBACK',
 'SocketType',
 '__builtins__',
 '__doc__',
 '__file__',
 '__name__',
 '_fileobject',
 '_realsocketcall',
 '_socketobject',
 'error',
 'errorTab',
 'gethostbyaddr',
 'gethostbyname',
 'gethostbyname_ex',
 'gethostname',
 'getprotobyname',
 'getservbyname',
 'htonl',
 'htons',
 'ntohl',
 'ntohs',
 'socket']
>>>

socket.fromfd() method is called in the fcgi._startup() function,
which is called in the constructor of FCGI object. Since
fcgi._startup() function is catching only socket.error exceptions
instead of AttributeError exception the program dumps.

If AttributeError except clause is added immediately after the
socket.error() in fcgi._startup():

    except AttributeError:
            global _isFCGI
            _isFCGI = 0
            return

then the program works fine.

Is there something I am doing wrong or is there something wrong with
my Python installation?

2) I was wondering whether it is possible to use fcgi.py based
programs with Apache using cgi-fcgi brdige. If so, how would one
configure cgi-fcgi to know to launch a python interpreter with a
specific program to interpret? If there is an example it would be
greatly appreciated.

Thanks


From sill at localhost.kitenet.net  Sat Dec 16 09:41:56 2000
From: sill at localhost.kitenet.net (Oldayz)
Date: Sat, 16 Dec 2000 14:41:56 GMT
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <915jcf0c6o@news2.newsguy.com> <9162p3$7t2$1@animus.fel.iae.nl> <MBWZ5.334$h67.37748@newsread1.prod.itd.earthlink.net> <3A397C74.6C8906E1@cosc.canterbury.ac.nz> <91d5520m6j@news1.newsguy.com> <wvw_5.10652$h67.670217@newsread1.prod.itd.earthlink.net> <slrn93mh41.1nk.scarblac@pino.selwerd.nl>
Message-ID: <UaL_5.12595$h67.878819@newsread1.prod.itd.earthlink.net>

On 16 Dec 2000 10:34:06 GMT, Remco Gerlich <scarblac at pino.selwerd.nl> wrote:
>Oldayz <sill at localhost.kitenet.net> wrote in comp.lang.python:
>> Hey, how about if you think while 1: looks ugly
>> you do true = 1 somewhere on top and use
>> while true: ? I think it's much more obvious
>> and clean and you don't need to change syntax.
>
>Yes! And you don't need the ugly "break" statement, you can simply do
>"true = 0" and the loop will end! (Remember to set it back to 1 later)

Yeah, you're right.. Things like that are definitely bad when someone
else looks at your code, but for something that you're sure nobody will
ever see, i might try using this.. except for doing true = 0, of course 
- that just makes things less clear. Isn't there some way to make that
variable unchangeable though?

>
>
>
>#define-THIRTEEN-12-ly-yrs,
> Remco Gerlich


-- 

	Andrei


From MarkH at ActiveState.com  Thu Dec 28 07:47:21 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Thu, 28 Dec 2000 12:47:21 GMT
Subject: Problem passing COM object
References: <926o83$39e$1@nnrp1.deja.com>
Message-ID: <3A4B3533.2010501@ActiveState.com>

wmistar at cyclones.com wrote:

> The error comes out as follows:
> 
>>>> dbh = trdb.AutoLogin("", "", "", "", 0, tr)
>>> 
> Traceback (innermost last):
>   File "<interactive input>", line 1, in ?
>   File "<COMObject TRDatabase.Database>", line 2, in AutoLogin
> com_error: (-2147352571, 'Type mismatch.', None, 6)

I'm afraid this isn't win32com giving the error, but the object itself. 
  win32com is passing the COM object, but the COM object itself is 
returing a COM error code of DISP_E_TYPEMISMATCH.

> In VB, the same call is sucessfully done by doing this:
> Dim mtr As TRDatabase.Application
> 
> dbh = trdb.AutoLogin("", "", "", "", bForceLogin, mtr)
> 
> Please help,


No idea I'm afraid.  The Python COM test suite does similar things...

Mark.



From aahz at panix.com  Sat Dec 23 12:30:38 2000
From: aahz at panix.com (Aahz Maruch)
Date: 23 Dec 2000 09:30:38 -0800
Subject: A Suggestion for Python Colon Syntax
References: <mailman.977544967.21073.python-list@python.org> <CF316.26760$1t.879107@typhoon.ne.mediaone.net>
Message-ID: <922nfu$dfk$1@panix2.panix.com>

In article <CF316.26760$1t.879107 at typhoon.ne.mediaone.net>,
David Abrahams <abrahams at mediaone.net> wrote:
>"Tim Peters" <tim.one at home.com> wrote in message
>news:mailman.977544967.21073.python-list at python.org...
>>
>> 2. The NumPy people had a real problem in that
>>     x = x + y
>> could end up generating multimegabyte temp arrays while
>>     x += y
>> need not.  That is, for them it wasn't "just another way" to write the
>> former; it offers crucially different semantics.
>
>Yes! Any chance of getting this capability for extension classes?

<blink>  Um, how do you think it works in NumPy?
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

'"Crisp" is a good quality for crackers; less so for pot roast.'  --pnh


From schaffer at optonline.net  Thu Dec 21 10:39:13 2000
From: schaffer at optonline.net (Les Schaffer)
Date: 21 Dec 2000 10:39:13 -0500
Subject: PyOpenGL and Nvidia GL?
References: <tmi%5.60958$df5.2143169@news1.crdva1.bc.home.com>
Message-ID: <87hf3xss0u.fsf@optonline.net>

In the mean time, if undef GL_EXT_polygon_offset, all runs well
(unless that is you are running nvidia and SMP, in chich case you get
random hard freezes under linux/nvidia/opengl).

les schaffer

*** _openglmodule.c.orig        Thu Dec  7 04:27:32 2000
--- _openglmodule.c     Thu Dec 21 10:37:00 2000
***************
*** 53,58 ****
--- 53,60 ----
  #include <windows.h>
  #endif
  #include <GL/gl.h>
+ #include <GL/glx.h>
+ #undef  GL_EXT_polygon_offset
  #include <math.h>
  
  #ifdef NUMERIC


From dsavitsk at my-deja.com  Thu Dec 21 11:53:17 2000
From: dsavitsk at my-deja.com (dsavitsk at my-deja.com)
Date: Thu, 21 Dec 2000 16:53:17 GMT
Subject: python with VB/Delphi
References: <91opqf$nr7$1@nnrp1.deja.com> <3A412723.4778E74F@eden.rutgers.edu>
Message-ID: <91tchp$cq2$1@nnrp1.deja.com>

In article <3A412723.4778E74F at eden.rutgers.edu>,
  Larry Kasoff <larkas at eden.rutgers.edu> wrote:
> If you are using Delphi, get Python for Delphi.

thanks ... the real issue is that i have a VB project that i wrote
before i had any idea what i was doing. It compiles to an 9meg .exe.
thus i am now fixing bugs in it, and splitting it up into manageable
objects.  i thought it might make my evenings more pleasant to do some
of the re-coding in python while keeping the interface more or less as-
is.

Also, my experience is that vb presents some installation issues so i
thought that delphi for the interfaces might be a better choice?

doug


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


From grey at despair.rpglink.com  Sat Dec 30 00:00:33 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Sat, 30 Dec 2000 05:00:33 -0000
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <3A4D6A05.EEC97FA6@uswest.net>
Message-ID: <slrn94qqvg.mgg.grey@teleute.rpglink.com>

On Fri, 29 Dec 2000 21:52:21 -0700, Bjorn Pettersen <pbjorn at uswest.net> wrote:
>henry_crun3583 at my-deja.com wrote:
>> do :
>>     R=Themistor.ReadResistance()
>>     T=R_to_Temperature(R)
>>   until (T >=25)

>while 1:
>   R=Themistor.ReadResistance()
>   T=R_to_Temperature(R)
>   if T >=25: break

    Or, for those who don't like while 1:

while T < 25:
  R=Thermistor.ReadResistance()
  T=R_to_Temerature()


-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From samschul at pacbell.net  Thu Dec 14 14:59:03 2000
From: samschul at pacbell.net (Sam Schulenburg)
Date: Thu, 14 Dec 2000 19:59:03 GMT
Subject: Capturing stdout from a console process?
Message-ID: <91b8q3$aqu$1@nnrp1.deja.com>

I am using <os.system("xcopy c:*.* d: /v")> within a python program to
copy files from one directory to another under windows NT. When the
program executes, a new console window is created with the results of
the xcopy command being displayed in the console window. Is their a way
to capture this output by python so I can log error messages?

Sam Schulenburg


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


From michael at stroeder.com  Sat Dec 30 08:54:59 2000
From: michael at stroeder.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Sat, 30 Dec 2000 14:54:59 +0100
Subject: Python legalities
References: <3A2FC63D.652101A@accessone.com> <3A308881.7FBEBAD4@engcorp.com> <3A3120DF.9A9F3725@san.rr.com> <92kk5a$c65$2@srv38.cas.org>
Message-ID: <3A4DE933.FC0E0E7D@stroeder.com>

lvirden at cas.org wrote:
> 
> Another path would be for free software to come with the same kind of
> assurances as commercial software - that the software is judged free and
> clear of such future legal issues and that, if not, the software identify
> some agency who will 'take the heat'...

Are you willing to pay for that?

Ciao, Michael.


From anderson at boi.hp.com  Tue Dec 12 14:03:32 2000
From: anderson at boi.hp.com (Bill Anderson)
Date: Tue, 12 Dec 2000 12:03:32 -0700
Subject: Issues with rebuilding the src rpms for 2.0
Message-ID: <3A367684.22B0E12A@boi.hp.com>


When trying to rebuild the src.rpms provided on python.org for 2.0,
particularly expat, as non-root, I've discovered some issues.

Right off the bat this happens:

"""
...
  inflating: /home/anderson/rpmdir/BUILD/expat/xmlwf/xmlwf.dsp  
+ echo 'Patch #0 (expat-Makefile.patch):'
Patch #0 (expat-Makefile.patch):
+ patch -p1 -s
missing header for context diff at line 3 of patch
The text leading up to this was:
--------------------------
|*** Makefile   2000/07/06 18:08:45     1.1
|--- Makefile   2000/07/06 18:10:34
--------------------------
File to patch: 
"""

Typing in Makefile seems to work.

An then we come to issue #2:

Building an rpm should _not_ install anything. When bulding as root, you
won't notice it, but when building the BeOpen-Python-2.0-1.src.rpm as
non-root, without the permissions to install anything to /usr/*, you'll
see:

"""
...
Executing(%install): /bin/sh -e /var/tmp/rpm-tmp.92343
+ umask 022
+ cd /home/anderson/rpmdir/BUILD
+ cd expat
+ install -D xmlwf/xmlwf /usr/local/bin/xmlwf
install: cannot create regular file `/usr/local/bin/xmlwf': Permission
denied
Bad exit status from /var/tmp/rpm-tmp.92343 (%install)
"""

Under NO circumstance should rebuilding a source rpm go dropping files
into anywhere outside the build tree and RPM Build Root. Fortunately,
modifying the spec file to do this shouldn't be that difficult. If I get
time to do so, I'll post it if the maintainer hasn't fixed that by then.



-- 
Bill Anderson               Linux Specialist
Modular Network Storage     R&D           
Random Quote:
    Portable: survives system reboot.


From colaveaj at sirinet.net  Fri Dec  8 05:01:39 2000
From: colaveaj at sirinet.net (Tony)
Date: Fri, 8 Dec 2000 10:01:39 -0000
Subject: Tkinter frame resizeing
Message-ID: <t30nhn9kq4a5a3@corp.supernews.com>

1.  How do you turn on the resize feature off for frames?

2.  In Java you can modal windows?  How is this achieved with python using
Tkinter?




From rjschwei at mindspring.com  Sat Dec  9 06:10:24 2000
From: rjschwei at mindspring.com (Robert Schweikert)
Date: Sat, 09 Dec 2000 06:10:24 -0500
Subject: compiler switches; getopt behavior
Message-ID: <3A321320.37230B3@mindspring.com>

Is there a compile time switch used when building Python on Linux to
make getopt behave as if the POSIXLY_CORRECT environment variable was
set? Or is the interpreter setting this environment variable somehow in
its environment?

It seems that if one builds Python a bit different one can get the
following undesired behavior.

"""
script try.py, test the system getopt function.
"""
import sys

print sys.argv



Now running this on Linux (not the interpreter that comes with the
distribution) I can produce the following

python try.py -help

Python: invalid option -- h
usage: Python [-d] [-i] [-s] [-u ] [-v] [-c cmd | file | -] [arg] ...

Options and arguments (and corresponding environment variables):
-d     : debug output from parser (also PYTHONDEBUG=x)
-i     : inspect interactively after running script (also
PYTHONINSPECT=x)
-s     : suppress printing of top level expressions (also
PYTHONSUPPRESS=x)
-u     : unbuffered stdout and stderr (also PYTHONUNBUFFERED=x)
-v     : verbose (trace import statements) (also PYTHONVERBOSE=x)
-c cmd : program passed in as string (terminates option list)
file   : program read from script file
-      : program read from stdin (default; interactive mode if a tty)
arg ...: arguments passed to program in sys.argv[1:]

Other environment variables:
PYTHONSTARTUP: file executed on interactive startup (no default)
PYTHONPATH   : colon-separated list of directories prefixed to the
               default module search path.  The result is sys.path.


However, when I leave the dash of I get what I expect:

python try.py help
['try.py', 'help']


Further after setting POSIXLY_CORRECT I also get what I expect:

setenv POSIXLY_CORRECT
python try.py -help
['try.py', '-help']


Thus the question what did I mess up when custom compiling the
interpreter?

Thanks,
Robert

--
Robert Schweikert                      MAY THE SOURCE BE WITH YOU
rjschwei at mindspring.com                         LINUX





From jcm at bigskytel.com  Sat Dec 23 00:04:14 2000
From: jcm at bigskytel.com (David Porter)
Date: Fri, 22 Dec 2000 22:04:14 -0700
Subject: coding convention conversion
In-Reply-To: <920v4s$oqj$2@taliesin.netcom.net.uk>; from garryknight@bigfoot.com on Sat, Dec 23, 2000 at 01:29:00AM +0000
References: <mailman.977516469.1157.python-list@python.org> <920v4s$oqj$2@taliesin.netcom.net.uk>
Message-ID: <20001222220414.A23735@bigskytel.com>

* Garry Knight <garryknight at bigfoot.com>:
> In article <mailman.977516469.1157.python-list at python.org> "Issac
> Trotts" <trotts at llnl.gov> wrote:
> 
> > Does anyone know of a utility to change between the coding conventions
> >  of
> > 
> > def fooBarBaz(): pass
> > 
> > and 
> > 
> > def foo_bar_baz(): pass
> > 
> > It is a pain to do this by hand.
> 
> Much though I hate to do this in a Python newsgroup, here's a Perl
> script that should do the job. It reads lines from STDIN and copies
> them to STDOUT. If a line starts with optional whitespace followed by
> the word "def" followed by more whitespace followed by a name, then if
> that name contains capital letters, each capital is lower-cased and has
> a "_" prepended.
> 
> Copy and paste the text between the cut marks and save it as, say,
> /usr/local/bin/redef (or the equivalent location on your system) and
> use it as follows:
>   redef <infile >outfile
> It's about as simple as it can be and hasn't been tested exhaustively.
> If there are errors, or if you need it enhanced, please let me know.

It only changes the def lines. It should also alter the lines that call the
function (but not imported functions):

from x import importedFunc

def someFunc(x):
    pass

someFunc("what about me?")    

importedFunc(DontAlter)

> --------8X-cut here-----------------------------------------------
> #!/usr/bin/perl
> 
> while (<>) {
>   if (/^\s*def\s+\w+/) {
>     s/([A-Z])/_\l\1/g;
>   }
>   print;
> }
> --------8X-cut here-----------------------------------------------





From jajvirta at cc.helsinki.fi  Wed Dec 13 10:43:09 2000
From: jajvirta at cc.helsinki.fi (Jarno J Virtanen)
Date: 13 Dec 2000 15:43:09 GMT
Subject: Traceback and source code
Message-ID: <slrn93f68c.m2r.jajvirta@sirppi.helsinki.fi>

I'd be interested to know how the traceback module is syncronized
with the actual source code (file). The trouble I keep having is
that traceback reports the corresponding line number correctly,
but the actual source code line is not always correct. I'm not
saying here that traceback module is incorrect or anything, I'd
just be happy to know if I'm doing something wrong or what could
be done to fix the problem. [I'm using python 1.5.2 on a Linux box,
if that's of any significance.]

The problem appears like this:

File "spam.py", line 22, in do_spam
    def do_eggs(self, args):
NameError: BUG

The longer explanation how I got there follows (it's a bit long though):

Suppose a programmer named Terry is assigned to design and implement
a command line interface called Spam. It is to be designed so that the
Spam source code can be reloaded while the program is running and that
it catches errors, prints them and lets the user continue running the
program. Well, Terry decides to split the code to a module that catches
the errors and reloads the other module and to a module that has the
actual features. The Error Catcher code he places in a file named 
main.py that has following source code:

import os, sys, string, traceback

while 1:
    try:
        import spam
        rspam = spam.Spam()
        rspam.cmdloop()
    except spam.reload:
        try:
            reload(spam)
        except:
	    sys.exc_traceback = sys.last_traceback = None
            traceback.print_exc(file=sys.stdout)
    except spam.exit:
        break
    except:
        sys.exc_traceback = sys.last_traceback = None
        traceback.print_exc(file=sys.stdout)

(that sys.exc_traceback = etc. expression I fished from the python
FAQ but it doesn't seem to help and didn't it seem to be the
issue, but one never knows :-)

The actual command line interface goes to file named spam.py:

import readline, cmd, sys

exit = 'exit spam'
reload = 'reload spam'

class Spam(cmd.Cmd):

    prompt = '>>> '

    def do_quit(self, args):
        raise exit

    def do_reload(self, args):
        raise reload

    def do_spam(self, args):
        print 'spam'
        
    def do_eggs(self, args):
        print 'eggs'


Now Terry wants to add more features in the command line interface.
He modifies the spam method to make it more useful:

    def do_spam(self, args):
        print 'spam'
        BUG

Excited as he is of the quick evolution of the Spam he issues reload
and spam commands in the CLI:

>>> reload
>>> spam
spam
Traceback (innermost last):
  File "main.py", line 7, in ?
    rspam.cmdloop()
  File "/usr/lib/python1.5/cmd.py", line 77, in cmdloop
    stop = self.onecmd(line)
  File "/usr/lib/python1.5/cmd.py", line 115, in onecmd
    return func(arg)
  File "spam.py", line 18, in do_spam
    BUG
NameError: BUG

Damn it, Terry says, I've made a BUG. Frustrated from the setback
he decides to go for a coffee to get things sorted in his head. Before
that he makes a reminder to the source code that there's a BUG 
somewhere. Furthermore he notices that the BUG is in the do_spam
-method:

    # There's a BUG here somewhere, fix it!

    def do_spam(self, args):
        print 'spam'
        BUG


As soon as he gets to the coffee machine he remebers and shouts: "Of
course, I made a BUG!". He runs back to his office and as he is 
running he rethinks the solution over and over again and begins
to think that he has already fixed it and issues reload and spam
again:

>>> reload
>>> spam
spam
Traceback (innermost last):
  File "main.py", line 7, in ?
    rspam.cmdloop()
  File "/usr/lib/python1.5/cmd.py", line 77, in cmdloop
    stop = self.onecmd(line)
  File "/usr/lib/python1.5/cmd.py", line 115, in onecmd
    return func(arg)
  File "spam.py", line 20, in do_spam
    def do_eggs(self, args):
NameError: BUG

(Terry gets even more frustrated and the rest is history ;-)



From pehr at linfoo.eink.com  Wed Dec  6 19:41:17 2000
From: pehr at linfoo.eink.com (Pehr Anderson)
Date: Wed, 6 Dec 2000 19:41:17 -0500
Subject: XML-RPC was [pygtk] Python and Orbit Examples
In-Reply-To: <v04220802b654744d4fc6@[192.168.11.33]>; from Blair.Lowe@compeng.net on Wed, Dec 06, 2000 at 04:05:03PM -0700
References: <Pine.LNX.4.21.0012050828550.24722-100000@tupelo.daa.com.au> <200012060632.eB66WIK27614@ZOOstation.cc> <20001206165355.B12869@linfoo.eink.com> <v04220802b654744d4fc6@[192.168.11.33]>
Message-ID: <20001206194117.D12869@linfoo.eink.com>

Dear Blair,

I'd recommend doing XML over a simple socket, without relying on
HTTP or SSL as a transport protocol, but that would require you
to include a security model within your XML specification.

It's not a bad idea to handle security separately from transport
unless you are set on the SSL security model.  It would be
perfectly reasonable to do a HTTP POST with your XML querry
and wait for the reply as an XML document, then all you need
is the SSL interface for python.

Python 2.0 includes support for SSL in its web libraries,
or you could try to do something with SSL by calling
lynx using os.popen("lynx --source https://...").read()

        -pehr

On Wed, Dec 06, 2000 at 04:05:03PM -0700, Blair Lowe wrote:
> Sounds cool.
> 
> I found pyXML. at http://download.sourceforge.net/pyxml/.
> 
> Can one send XML to an SSL site with pyXML as well? If so, where is 
> some good docs for this. The version I got (PyXML-0.6.2-2.0.i386.rpm) 
> has not docs, and the overview at 4suite say I need to download 
> something else.
> 
> Are there some better tools for doing this. I started to look at 
> Zope, but it was rather cumbersome and knowing where to start with 
> the documentation was not easy to find either.
> 
> TTYL,
> Blair.
> 
> At 16:53 -0500 2000/12/06, Pehr Anderson wrote:
> >Dear Folks,
> >
> >I would highly recommend using XML-RPC as a minimally complicated
> >way to communicate data between your apps.
> >
> >This might sound complicated but it really only means
> >	1. specify XML encodings for your data, use these throught your code
> >	2. open a socket on the target
> >	3. send an XML query and wait for an XML response
> >	4. parse the response
> >
> >The beautiful thing about XML-RPC (remote proceedure call) is
> >that it is just dirt simple. You open a socket, dump in
> >human readable text, and wait for a human-readable response.
> >Everything gets completely defined somewhere and servers and
> >clients could be coded in *anything*!
> >No complicated tools are required to implement any parts of
> >this system though you might want to look at how others
> >parse XML rather than doing it all by hand.
> >
> >And the best part is you'll have an app that anyone can understand.
> >No secret church of CORBA, COM or other "special tools".
> >XML is directly comprehensible by HUMAN BEINGS and that makes it
> >a sustainable encoding with simple tcp sockets for transport.
> >
> >	-pehr
> >
> >
> >On Tue, Dec 05, 2000 at 06:32:17PM -1200, Rob Brown-Bayliss wrote:
> >>  >or getting it to work with gnome-python, that is a bit more difficult.
> >>
> >>  Sounds a little ominous...
> >>
> >>  What I am doing (or planning at this stage) is to write a POS and Stock
> >>  management system for a small retail chain.
> >>
> >>  Would I be better of giving Pyhton a miss, or ditching the Corba idea?
> >>
> >>  I have very little knowledge of linux programming, though I do have
> >>  years of basic programming on the Amiga (like 18 I think if 
> >>include vic 20 and c64)
> >>  and a few months VC++ on windows.  I have used Linux for 18 months 
> >>now, and played
> >>  a bit with C and a very very little with Python.
> >>
> >>  I am confident I can tackle the task in either language, but I 
> >>know python will be
> >>  quicker to develop with.
> >>
> >>  I had looked to corba as a means of communicating between the 
> >>stores, I could always
> >>  use some form of file transefer, but it's not as elegant or nearly 
> >>as real time.
> >>
> >>
> >>
> >>
> >>  --
> >>
> >>    Rob Brown-Bayliss
> >>   ---======o======---
> >>
> >>    www.ZOOstation.cc
> >>
> >>  _______________________________________________
> >>  pygtk mailing list   pygtk at daa.com.au
> >>  http://www.daa.com.au/mailman/listinfo/pygtk
> >
> >--
> >http://www.python.org/mailman/listinfo/python-list
> 
> Computer Engineering Inc. http://www.compeng.net
> Phone: 780 499 5687 (9 - 5 MST) Fax:   780 435 0693 (24 Hours)



From emile at fenx.com  Fri Dec 15 10:57:49 2000
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 15 Dec 2000 07:57:49 -0800
Subject: Python advocacy
References: <91bln3$lmr$1@nnrp1.deja.com> <91d63k0msl@news1.newsguy.com> <8iak3t4a3lrqcqp9iqftmekdoahuico2b0@4ax.com>
Message-ID: <3A3A3F7D.6030009@fenx.com>


Steve Horne wrote:

> On Fri, 15 Dec 2000 14:24:27 +0100, "Alex Martelli"
> <aleaxit at yahoo.com> wrote:
> 
>> Makes me wonder why Larry Wall didn't think to include at least
>> ONE strangeness, wart, or inconsistency *somewhere* in Perl...
>> thus making it so boringly perfect and un-idiosyncratic as to
>> leave its advocates without a suitable 'scapegoat feature'!
> 
> 
> Erm - either you haven't used Perl, or your an advocate yourself!
> Personally, if I was selling Python, Perl would be involved in my
> scapegoat manouvre.

I expect Alex had his tongue firmly planted in cheek for that part of 
his post.

But-I-could-be-wrong-ly y'rs

Emile van Sebille
emile at fenx.com



From barry at digicool.com  Sun Dec 17 18:30:14 2000
From: barry at digicool.com (Barry A. Warsaw)
Date: Sun, 17 Dec 2000 18:30:14 -0500
Subject: Lists have moved
Message-ID: <14909.19590.980876.614980@anthem.concentric.net>

The last of the python.org mailing lists have moved to the new
machine.  This should be transparent to you except for the re-homing
of the news gateway to Digital Creation's ISP.

I never really figured out what the state of the news/mail gateway was
at CNRI so I don't know if you're about to be flooded with duplicates,
suddenly deprived of usenet gated messages, or somewhere in between.
Ain't life fun?  In either case, I apologize in advance for whatever
is screwed up. ;}

Please keep an eye on things, as will I, and email me directly if you
notice anything weird and/or bad going on with any of the python.org
mailing lists.

Cheers,
-Barry



From freydaddy at kricket.net  Tue Dec 12 22:05:12 2000
From: freydaddy at kricket.net (Matt Frey)
Date: Wed, 13 Dec 2000 03:05:12 GMT
Subject: compiled or interpreted?
Message-ID: <IHBZ5.126133$DG3.2517319@news2.giganews.com>

Is Python a compiled language that can be run on a platform like c++, or is
it interpreted like java, or is it in a class of its own?




From ullrich at math.okstate.edu  Sun Dec 17 11:21:12 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Sun, 17 Dec 2000 16:21:12 GMT
Subject: Python for REAL Projects ?
References: <3a380d1e.409238@bendigo.news.origin.net.au> <3A3948CC.50939621@spam-no-no.mail.ru>
Message-ID: <3a3ce771.409101476@nntp.sprynet.com>

On Fri, 15 Dec 2000 03:25:16 +0500, Igor Poteryaev
<jah at spam-no-no.mail.ru> wrote:

>Hello DELPHI dragon !
>
>You definitely can get the best from both Delphi and Python worlds.
>Have you ever heard about PythonForDelphi located at
>http://www.multimania.com/marat/  ?
>It is a nice example of both extending and embedding capabilities of 
>python. You get all comfort of using Delphi forms for UI and flexibility 
>of python. 

	Absolutely. (I've tried to say what you just said about
3 times - keep getting strange errors (What does "Spam domain error"
mean???)

	Not to waste a whole post just to say "absolutely" - I'm
curious whether this one will go through.

DU



From spahievi at vega.bg  Mon Dec 11 15:40:24 2000
From: spahievi at vega.bg (Niki Spahiev)
Date: Mon, 11 Dec 2000 22:40:24 +0200
Subject: global __dict__
In-Reply-To: <Pine.LNX.4.10.10012111601001.487-100000@erde.kosmos.all>
References: <Pine.LNX.4.10.10012111601001.487-100000@erde.kosmos.all>
Message-ID: <04354016.20001211224024@vega.bg>

11.12.2000, 17:05:26, Carsten Geckeler wrote:

CG> Hi.

CG> I was wondering if there is a __dict__ dictionary for the global namespace
CG> as for classes and modules.  I was trying the following

>>>> var = 1
>>>> __dict__

CG> and expected that __dict__ would contain something like {..., 'var': 1},
CG> but there is no global __dict__.  Where can i find the global dictionary?

>From globals()

-- 
Best regards,
 Niki Spahiev





From nas at arctrix.com  Sat Dec 30 08:42:19 2000
From: nas at arctrix.com (Neil Schemenauer)
Date: Sat, 30 Dec 2000 05:42:19 -0800
Subject: Python vs PHP (& Java?)
In-Reply-To: <20001228162203.7163.qmail@stimpy.scso.com>; from moshez@zadka.site.co.il on Thu, Dec 28, 2000 at 04:22:03PM -0000
References: <92f3kp01cqe@news1.newsguy.com>, <mailman.977897649.4877.python-list@python.org> <slrn94l86m.a3h.tim@degree.ath.cx> <92f3kp01cqe@news1.newsguy.com> <20001228162203.7163.qmail@stimpy.scso.com>
Message-ID: <20001230054219.A21277@glacier.fnational.com>

On Thu, Dec 28, 2000 at 04:22:03PM -0000, Moshe Zadka wrote:
> (BTW: Are there plans for quixote/mod_snake integration? That'd
> be ultra-cool: one of the benefits of HTML::Mason is that it
> integrates really well with mod_perl)

What would be the advantage of that over Quixote/FCGI?
 
  Neil



From qrczak at knm.org.pl  Fri Dec 29 05:15:36 2000
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 29 Dec 2000 10:15:36 GMT
Subject: Kanji
References: <92dl3u$to0$1@news.idiom.com> <p6qlmt1e3yh.fsf@informatik.hu-berlin.de> <92f3n7$hsf$1@news1.xs4all.nl>
Message-ID: <slrn94op28.srq.qrczak@qrnik.zagroda>

28 Dec 2000 10:12:55 GMT, Boudewijn Rempt <boud at rempt.xs4all.nl> pisze:

> And Kura is a language database written entirely in Python and Qt
> which reached feature-complete stage yesterday...

"Kura" means "hen" in Polish BTW.

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


From neelk at alum.mit.edu  Thu Dec 14 23:17:45 2000
From: neelk at alum.mit.edu (Neelakantan Krishnaswami)
Date: 15 Dec 2000 04:17:45 GMT
Subject: xreadlines (was Re: while true: !!!)
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <slrn93g37o.9tj.jepler@potty.housenet> <f3%Z5.146409$U46.4768601@news1.sttls1.wa.home.com> <91a6it02flt@news2.newsguy.com>
Message-ID: <slrn93jagg.fl.neelk@alum.mit.edu>

On Thu, 14 Dec 2000 11:14:38 +0100, Alex Martelli <aleaxit at yahoo.com> wrote:
>> stdin. I've used fileinput to go through big lists of files (10,000+ email
>> messages) and it works great. It doesn't appear to do any buffering
>> itself--it uses file.readline() to read the files.
>
> If this was a performance problem, it could of course also be fixed
> in a future fileinput version without changing code that uses it (again,
> in-place-rewriting would probably have to inhibit the optimization,
> although that isn't entirely clear).

While it's true that fileinput is somewhat slow, but it can easily be
made faster than the usual while 1: loop everyone uses.(Relatively
speaking -- if you have the RAM, everything loses vis-a-vis
readlines().) Concretely, here's a wrapper class that's *faster* than
the traditional while 1: loop, because it does uses readlines sizehint
trick that everyone on c.l.p describes but no one ever seems to use in
practice.

class Pita:
    """Pita(file)

    This class wraps a file-like object around an iterator so you can
    loop over a file's lines with a for loop. Eg:

    >>> for line in Pita(open('foo.txt')):
    ...     print line[:-1]
    ...
    """
    def __init__(self, file, hint=16384):
        self.file = file
        self.current = 0                           # Current line number
        self.chunk = hint                          # Chunk size to read
        self.buf = self.file.readlines(self.chunk) # A buffer of lines
        self.i = 0                                 # index into buffer
    def __getitem__(self, n):
        if n == self.current:
            self.current = self.current + 1
        else:
            raise KeyError, "Attempt to read stream out of order"
        #
        # If we've reached the end of the buffer, grab a new chunk
        #
        if self.i == len(self.buf):
            self.buf = self.file.readlines(self.chunk)
            self.i = 0
        #
        # An empty buffer -> No more lines in the file
        #
        if self.buf: 
            line = self.buf[self.i]
            self.i = self.i + 1
            return line
        else: 
            raise IndexError

Some timing results on a roughly 5.5 MB text file (Python 1.5.2;
I haven't upgraded yet.)

>>> test('/home/neelk/INBOX')
fileinput = 3.565142
while 1: = 2.653460
Pita = 1.128891
readlines() = 0.337859

The Pita class is almost three times as fast as FileInput and twice as
fast as while 1:, and readlines() is almost four times as fast as
that. FileInput is the slowest of the lot -- but it looks like using
the sizehint trick could bring it up to Pita speed.

Here's the test driver:

def test(filename):
    t1 = time.time()
    for i in fileinput.FileInput(filename):
        pass
    t2 = time.time()
    print "fileinput = %f" % (t2 - t1)
    #
    f = open(filename)
    t1 = time.time()
    while 1:
        line = f.readline()
        if not line: break
    t2 = time.time()
    f.close()
    print "while 1: = %f" % (t2 - t1)
    #
    t1 = time.time()
    for i in Pita(open(filename)):
        pass
    t2 = time.time()
    print "Pita = %f" % (t2 - t1)
    #
    t1 = time.time()
    for i in open(filename).readlines():
        pass
    t2 = time.time()
    print "readlines() = %f" % (t2 - t1)


Neel


From moshez at zadka.site.co.il  Mon Dec 18 23:01:09 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Tue, 19 Dec 2000 06:01:09 +0200 (IST)
Subject: search for pattern in StringList
In-Reply-To: <31575A892FF6D1118F5800600846864D5B1614@intrepid>
References: <31575A892FF6D1118F5800600846864D5B1614@intrepid>
Message-ID: <20001219040109.22E70A84F@darjeeling.zadka.site.co.il>

On Mon, 18 Dec 2000, Simon Brunning <SBrunning at trisystems.co.uk> wrote:
>  
> myList = ['an', 'been', 'call', 'door']
> for i in range(0, len(myList)):
>     if myList[1].find('a') <> -1:
>         print i, myList[i]

Why not
for (i, element) in zip(range(myList), myList):
	if 'a' in element:
		print i, element
(Warning: zip was added in Python 2.0)
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From katz at Glue.umd.edu  Tue Dec 26 13:12:33 2000
From: katz at Glue.umd.edu (Roy Katz)
Date: Tue, 26 Dec 2000 13:12:33 -0500
Subject: Formatting for hh:mm:ss
Message-ID: <Pine.GSO.4.21.0012261309370.13920-100000@z.glue.umd.edu>

Hello,


I've searched the 'time' and 'locale' modules  for any function which
would take a difference of two time.time()'s and format the result as
hour:minutes:seconds, but I can't find any.  All I am doing is this:

  t1 = time.time()
  # some processing here
  t2 = time.time()

  total = t2 - t1

  print_time( total )  

where print_time() would format the time as above. Does anyone know if
this function already exists?


Thanks,
Roey



From moshez at zadka.site.co.il  Sun Dec 17 06:58:03 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Sun, 17 Dec 2000 13:58:03 +0200 (IST)
Subject: Singleton vs Proxies DP (was Re: Solution: Direct access to Printer I/O lines)
In-Reply-To: <InW_5.37707$x6.18119119@news2.rdc2.tx.home.com>
References: <InW_5.37707$x6.18119119@news2.rdc2.tx.home.com>, <917k9b$9i9$1@nnrp1.deja.com> <slrn93hrjn.2vp.ssthapa@localhost.localdomain> <91cnp3$g59$1@nnrp1.deja.com> <91d23f0jk7@news1.newsguy.com> <91gfnp$7r4$1@animus.fel.iae.nl> <91gkh801e49@news2.newsguy.com>
Message-ID: <20001217115803.C1551A82E@darjeeling.zadka.site.co.il>

On Sun, 17 Dec 2000 03:26:32 GMT, "Rainer Deyke" <root at rainerdeyke.com> wrote:

> I consider the singleton pattern as stated in the GOF to be overly
> cumbersome in the context of Python.  (It is necessary in C++ because of
> object creation order.)

Not only that -- laziness is also important.
What I mean is that I *do* like the fact that it initializes when *I*
tell it to. (I don't like module initialization to do more then define
functions and classes...)

The idiom I use is 

class _Foo:

	...write a regular class here...

foo = None
def Foo():
	global foo
	if foo is not None:
		return foo
	foo = _Foo()
	return foo

This has the added *feature* that it is not possible to inherit from Foo.
I'm against inheritance as a rule, and inheritance from singletons positively
shocks me, given that it is so easy in Python to use delegation. Delegation
to singletons is good, and what's more, the semantics are clear-cut.
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From paulb at infercor.no  Mon Dec 11 07:50:01 2000
From: paulb at infercor.no (Paul Boddie)
Date: Mon, 11 Dec 2000 13:50:01 +0100
Subject: creating an odbc link with python
References: <gSVX5.55717$yR4.1401853@news1.rdc1.tx.home.com>
Message-ID: <3A34CD79.F13410CC@infercor.no>

Scott Hathaway wrote:
> 
> I am using python to create an installation program for a web app that
> we have at our company.  The web app uses an access database that is
> hooked up through odbc.  How can I programatically create the dsn in
> python (I guess I need to edit the windows registry)?

I think that what you are asking is how one can call various functions in the
win32 API so that an ODBC data source gets added to the list maintained by ODBC
driver manager. Whilst I have no experience in this area, it would be
interesting to know how one might achieve this. Perhaps it is a case of calling
methods in various COM interfaces.

Regards,

Paul


From fredrik at effbot.org  Tue Dec  5 16:20:33 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 05 Dec 2000 21:20:33 GMT
Subject: Tkinter frame items?
References: <90jlgp$okn$1@nnrp1.deja.com>
Message-ID: <B_cX5.1931$O%.339282@newsc.telia.net>

Sam Schulenburg wrote:
> I have a Tkinter program under windows2000 where I have added a quit
> menu item to exit the program, with a prompt to the user to verify the
> desire to exit. My problem is that the user can still click the upper
> right exit square and close the program. I would like to capture the
> event triggered by clicking on this item.

you can use a WM_DELETE_WINDOW handler.  see example 7-2
under "protocols" on this page:

http://www.pythonware.com/library/tkinter/introduction/events-and-bindings.htm

cheers /F




From cjc26 at nospam.cornell.edu  Wed Dec 20 18:43:40 2000
From: cjc26 at nospam.cornell.edu (Cliff Crawford)
Date: Wed, 20 Dec 2000 23:43:40 GMT
Subject: Database solution for Mac?
References: <91kjtq$ns3$1@news1-2.kornet.net> <91kul3029b5@news1.newsguy.com>
Message-ID: <slrn942hen.gjv.cjc26@synecdoche.sowrong.org>

* Alex Martelli <aleaxit at yahoo.com> menulis:
| "Ahn, Jae-wook" <mail at codex.pe.kr> wrote in message
| news:91kjtq$ns3$1 at news1-2.kornet.net...
| > Hello,
| >
| > I have an iBook. I usually do works on Linux or Windows, but sometimes I
| > need some database support for my small Mac machine.
| >
| > I visited some pages, and found out a lot of database support for Python
| > including ODBC, but could not see anything that supports Macintosh. Please
| > recommend any solution which support database for Mac. Networked or local.
| 
| I think Gadfly should build and run happily on a Mac; it's an
| SQL relational DBMS that is implemented in Python (with optional
| C in you need to speed up certain things), see:
| 
| http://www.chordate.com/kwParsing/gadfly.html

There's also MetaKit, at http://www.equi4.com/metakit/ .  I don't think
this one needs to load the entire database into memory, unlike Gadfly.
(Unfortunately their site seems to be a little slow at the moment, so I
couldn't check that)


-- 
cliff crawford   http://www.people.cornell.edu/pages/cjc26/
                "Blink your eyelids periodically to lubricate your eyes."


From tim.one at home.com  Sat Dec 23 12:35:50 2000
From: tim.one at home.com (Tim Peters)
Date: Sat, 23 Dec 2000 12:35:50 -0500
Subject: A Q concerning map (a comparison to Maple V)
In-Reply-To: <3A448735.4DF3C95C@schneider-kamp.de>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEEDIFAA.tim.one@home.com>

[Franz GEIGER]
> > map(replace, ["abc", "cde"], "c", "C");
> > instead of having to create a lambda function.
> > Wouldn't it make sense to have that in Python too? Do there come
> > more such opportunities into your mind?

[Peter Schneider-Kamp]
> If I get you right you intend this to mean:
>
>   map(replace, ["abc", "cde"], ["c"]*2, ["C"]*2)
>
> So basically your proposal is to reuse map parameters
> which are no sequences in every iteration over the
> parameters who are,

It's hard to tell for sure, but that was my guess too.  In array languages,
this kind of thing is ubiquitous and is often called "scalar broadcast"
(where a scalar is any atomic value (as opposed to an array), and is
"broadcast" to every position of the arrays in which it's combined via some
operation).

There was a long debate about this a few years ago on c.l.py, where I
championed scalar broadcast in map specifically.  It collapsed under its own
hideous weight when-- as such things always do <wink> --that originally
modest goal got hijacked by people seeking to extend the semantics to every
corner of the language.  That won't happen.

> ...
> But how would you decide if "abc" is a sequence parameter
> or not?

That is a problem!  Kinda.  Mostly people want to broadcast numbers (as in
Franz's original example), and there's no problem there.  A cheap workaround
for sequences that are desired to be treated as scalars would have been to
introduce a new builtin scalar() function, that simply hid the
"sequenceness" of its argument from map.

Note that the listcomps in 2.0 make scalar broadcast much easier to spell;
e.g.

    [replace(x, "c", "C") for x in ("abc", "cde")]
    [x**2 + 5 for x in L]

So the better solution is not to extend map, but to forget it <0.9 wink>.

a-life-strategy-of-universal-applicability-ly y'rs  - tim




From dgorisek at iface.ch  Wed Dec 13 09:51:21 2000
From: dgorisek at iface.ch (David Gorisek)
Date: Wed, 13 Dec 2000 15:51:21 +0100
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <kuzoi15d60.fsf@lasipalatsi.fi> <Pine.BSF.4.10.10012122310510.48551-100000@tcfreenet.org> <kupuiw698s.fsf@lasipalatsi.fi>
Message-ID: <3A378CE9.3642901D@iface.ch>

Erno Kuusela wrote:

> In article <Pine.BSF.4.10.10012122310510.48551-100000 at tcfreenet.org>,
> Aaron Jon Reichow <reichowa at tcfreenet.org> writes:
>
> | On 13 Dec 2000, Erno Kuusela wrote:
> || the thing that struck me as a little foreign was the apparent
> || isolation of the smalltalk image from the rest of the operating system
> || and the filesystem. i want to be able to grep code!
>
> | While I'm not a proponent of Smalltalk isolationism, grep is a useless
> | tool (more or less), simply an artifact, ye, a symptom of the
> | sickness known as file-based development.
>
> point was that it's valuable to me to be able to manipulate stuff via
> the filesystem. version control, editors, access control, locking,
> backups, and all the rest of the universe works with files (same
> critique goes for zope).

I don't know what universe you are talking about, but VA Java too uses a
repository and so does Oracle. Also, every business application I know uses a
database instead of the file system. Usualy locking in a repository or in a
database is also more granular than what you would get in a file system.

Anyway, folks I guess this discussion belongs to comp.lang.smalltalk.advocacy,
don't you?

David



From jnix at medi.physik.uni-oldenburg.de  Fri Dec 22 10:48:04 2000
From: jnix at medi.physik.uni-oldenburg.de (jnix at medi.physik.uni-oldenburg.de)
Date: 22 Dec 2000 16:48:04 +0100
Subject: Building Python with Linux/ccc on the Alpha Platform
Message-ID: <91vt3l$l97@news.Informatik.Uni-Oldenburg.DE>

Hi,

I am interested in building Python with Numeric extension using the
Compaq CCC compiler on a Alpha21264 machine under Linux. I have a good
reason for this: ccc binaries usually run three to five times faster
than the corresponding gcc binaries, especially if they are
floating-point intensive - at this point in time, gcc just not
optimizes well on the Alpha.

Anybody did this before?

What do I have to observe?

Johannes





From cjensen at bioeng.ucsd.edu  Wed Dec  6 18:31:36 2000
From: cjensen at bioeng.ucsd.edu (Curtis Jensen)
Date: Wed, 06 Dec 2000 15:31:36 -0800
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <3A2DD2CA.14A5D7E2@engcorp.com> <90l3j9$4bn$1@kapsel.intranet> <976139270.161152@clam-ext>
Message-ID: <3A2ECC58.DE426E9F@bioeng.ucsd.edu>

Mark Hadfield wrote:
> 
> "Piotr Kuchta" <piotr at itam.zabrze.pl> wrote in message
> news:90l3j9$4bn$1 at kapsel.intranet...
> > ...Dutchmen are very talented: common Dutchman speaks well two foreign
> > languages, English and German.
> 
> Whereas common Australian speaks poorly one foreign language: English.
> 
> ---
> Mark Hadfield
> m.hadfield at niwa.cri.nz  http://katipo.niwa.cri.nz/~hadfield/
> National Institute for Water and Atmospheric Research
> PO Box 14-901, Wellington, New Zealand

No dramas mate.  I's like me english just the way it is.  
Though, it's not as if the Kiwi are heaps better at speakin either. 
Good luck understanding the Maori people all together, "eh bro."

-- 
Curtis Jensen


From aahz at panix.com  Wed Dec 20 14:56:03 2000
From: aahz at panix.com (Aahz Maruch)
Date: 20 Dec 2000 11:56:03 -0800
Subject: Threads - Newbie Questions
References: <3A410BB1.AAB6BE3C@NONOraqiaMAPS.com>
Message-ID: <91r2sj$e7b$1@panix3.panix.com>

In article <3A410BB1.AAB6BE3C at NONOraqiaMAPS.com>,
David Lees  <DavidL at NONOraqiaMAPS.com> wrote:
>
>I am now starting to play with Python threads.  Could someone refer me
>to documentation other than the Library reference for some simple
>examples?

http://starship.python.net/crew/aahz/

>My specific questions are:
>1. I am passing the lock as an argument to each thread.  Is this the
>proper way to do it?

More or less, though you may need more than one lock.  Take a look at
threading.py for some source code examples of constructing higher-level
locking objects.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

I don't really mind a person having the last whine, but I do mind
someone else having the last self-righteous whine.  --Aahz


From emile at fenx.com  Thu Dec 28 11:03:54 2000
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 28 Dec 2000 08:03:54 -0800
Subject: Interesting behaviour of the assignment
References: <92f4gk$qk6$1@news.nuri.net>
Message-ID: <92fo9g$6vams$1@ID-11957.news.dfncis.de>

IIRC, the integers through 99 are interned, and all
references have the same ID.

--

Emile van Sebille
emile at fenx.com
-------------------


"June Kim" <junaftnoon at nospamplzyahoo.com> wrote in message
news:92f4gk$qk6$1 at news.nuri.net...
> >>> a=100
> >>> b=100
> >>> a is b
> 0
> >>> a=99
> >>> b=99
> >>> a is b
> 1
>
> Is this an intended/expected result? What is the lying
structure beneath
> this?
>




From dsh8290 at rit.edu  Fri Dec  1 21:25:07 2000
From: dsh8290 at rit.edu (D-Man)
Date: Fri, 1 Dec 2000 21:25:07 -0500
Subject: [OT] Apple IIe help please
Message-ID: <20001201212507.A14226@westsidecnc.rh.rit.edu>

Hi all.  I apologize for the off topic-ness of the message and the cross-post. 
I have a friend who has a program she likes (written in BASIC) on an Apple IIe. 
I have a way to get into the code and list it on the screen.  What I am looking
for is someone who has had some experience using an Apple IIe who can tell me
how I can get the listing to be redirected to the printer port.  Also if it is
possible to get a copy of the program file onto a disk for an IBM compatible
comptuer (Windows or Linux) that would be great.

Thanks in advance for all the gurus out there who can help me.
-D




From claird at starbase.neosoft.com  Tue Dec 26 14:17:48 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 26 Dec 2000 13:17:48 -0600
Subject: python telnet script for openvms host
References: <92ainp$jab$1@nnrp1.deja.com> <kuito7rrg7.fsf@lasipalatsi.fi> <92apa7$oac$1@nnrp1.deja.com>
Message-ID: <D8D70FCFA8CC11C0.4BEE4AB5F2A3BCF7.EAD0DA106483FAE8@lp.airnews.net>

In article <92apa7$oac$1 at nnrp1.deja.com>,
Ennis Rosamond  <ennisr at my-deja.com> wrote:
>In article <kuito7rrg7.fsf at lasipalatsi.fi>,
>  Erno Kuusela <erno-news at erno.iki.fi> wrote:
>> Ennis Rosamond <ennisr at my-deja.com> writes:
>>
>> | tn.read_until('[c')    #last char of session hearder
>> | tn.write('$ dir\r')
>>
>> | #following this, I receive back the sequence:
>> | #recv '\377\373\001'
>> | #IAC WILL 251
>> | #recv '\377\373\003'
>> | #IAC WILL 251
>> | #recv '\003\\\033Z'
>> | #
>> | #which I don't know how to interpret.
>>
>> that looks like telnet option negotiation, telnetlib should
>> take care of all that automaticalyl for you. i suspect
>> the above is just some debug output from telnetlib.
>>
>> are you sure you are seeing it after your last .write call,
>> and not at the start of the telnet connection?
			.
			.
			.
>You are correct.  The sequence I listed is the debug output.  I am
>seeing it after the "tn.write('$ dir\r')" call.  It may not be relevant
>if the host simply isn't ready to receive the write call.
>
>I tried using "tn.read_until('$ ')" in the previous read instead of
>"tn.read_until('[c')", but that sequence is never found.  That is why I
>think the host is not sending back a prompt.  But then I am pretty naive
>about this whole topic.
			.
			.
			.
1.  OpenVMS term negotiation is hairy--or at least
    I always forget the magic after about a year,
    and also where I put the answer the last time.
    Based on my experience, I'd guess it *very*
    likely that all your difficulties have to do
    with that initial negotiation.
2.  This kind of automation is a great idea.  You'll
    be plenty happy once you get it working.
-- 

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


From z3penguin at penguinpowered.com  Wed Dec 20 15:05:52 2000
From: z3penguin at penguinpowered.com (Z 3 Penguin)
Date: Wed, 20 Dec 2000 15:05:52 -0500
Subject: finding where an object is referenced?
Message-ID: <91r3fn$jpr$1@bob.news.rcn.net>

is there any way for me to find out where a given object is being referenced?  it does not seem to be destructing correctly, like it has an extra reference somewhere.
-z\



From devlin at hot.ee  Mon Dec 18 21:25:54 2000
From: devlin at hot.ee (Teet Tarno)
Date: Tue, 19 Dec 2000 02:25:54 GMT
Subject: Helpme "Corrupt installation detected"
Message-ID: <3a3ec297.51392347@news.neti.ee>

Hi !

I have a dumb problem for all you nice people out there.
One day I downloaded the Beopen-Python-2.0.exe and tried to install it
on W95. Well, I ran the installer and it ran out of diskspace. No, not
the 18M install, but FAT16 and 1800 (a lot) files. I used
Quarterdeck's Cleansweep, uninstalled the thing, freed some more space
and now when I try to run it, it does nothing and simply says : 
"Corrupt installation detected"
and exits.
 To the best of my knowledge, Cleansweep should have removed all
traces of the install (HD, *.ini, registry) . I did search in the
registry and took a quick look around the disk. Nothing seemingly
left.

Where does the BeOpen installer save the info about the failed
installation ?
I could download the installer again, but 5M over slow modem is not
THAT much fun. But if it has to Be...

If anybody has any idea - I would greatly appreciate it !
Thanks in advance !

Teet Tarno



From aahz at panix.com  Fri Dec 29 11:05:29 2000
From: aahz at panix.com (Aahz Maruch)
Date: 29 Dec 2000 08:05:29 -0800
Subject: Great New Book
References: <t4o1vr8hktk531@corp.supernews.com>
Message-ID: <92ico9$jde$1@panix2.panix.com>

In article <t4o1vr8hktk531 at corp.supernews.com>,
Jason Cunliffe <jasonic at nomadicsltd.com> wrote:
>
>The only shame is not the author's fault - it is the still crazy high price
>$45 and thick paper which bulks the 760pages to 2.5" thick.

Hmmm....  I don't know how thick the pages actually are, but I sure
wouldn't want a book to have tissue-paper pages.  It's a tricky balance.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

Ninth Virtual Anniversary: 2 days and counting


From dale at out-think.NOSPAMco.uk  Mon Dec  4 05:12:14 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Mon, 04 Dec 2000 10:12:14 +0000
Subject: Python programs (or scripts) as drop targets
References: <hv5k2tg9cdf33j8v333qpkk7jo45ht596n@4ax.com> <90d9hi029hm@news1.newsguy.com> <3A2ACF21.4060406@ActiveState.com>
Message-ID: <k6rm2tk2fav7tlj8pea4a6ep686b0b27ce@4ax.com>

Mark Hammond <MarkH at ActiveState.com> wrote:


>To be quite honest, while I see that the "batch file" wrapper solution 
>isnt as "clean" as it could be, I don't believe a whole bunch of C++ 
>code or registry additions are worth saving this one trivial file.
>
>Mark.

Yes, it is only a trivial file but it's the messing about to set it up and you need one for each
program you want to work this way.

It's not just a matter of neatness. It's completeness, too.

--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants




From hs at paradise.nirvananet  Sun Dec 24 00:18:59 2000
From: hs at paradise.nirvananet (Hartmann Schaffer)
Date: 24 Dec 2000 00:18:59 -0500
Subject: string split
References: <slrn94b2go.srm.jp@localhost.localdomain>
Message-ID: <924103$n5c$1@paradise.nirvananet>

In article <slrn94b2go.srm.jp at localhost.localdomain>,
Jacek =?iso-8859-2?Q?Pop=B3awski?= <jp at ulgo.koti.com.pl> wrote:
>I have string:
>
>s="one two <br> three"
>
>splitting it gives:
>
>['one', 'two', '<br>', 'three']
>
>but if string looks like that:
>
>s="one two<br> three"
>
>string.split(s) gives me:
>
>['one', 'two<br>', 'three']
>
>is it possible to split not by space (" ") but by "<" and ">" ?
>
>I need:
>
>['one two','<br>','three']
>
>How to do it in simple way?

did you check the manual?  split takes a second argument.  however,
you can split only on either '<' or '>', and the delimiter won't be
part of any of the arguments.  you would need to roll your own
function:

def mysplit(string):
    list1= split(string, '<')
    list1= [list1[0]] + map(lambda x: '<' + x, list1[1:])
    list2= map(lambda x: string.split(c, '>'), list1)
    list3= [ ]
    for s in list3:
        list3.append(map(lambda x: x+'>', s[:-1])+[s[-1]])
    return reduce(lambda x,y: x+y, list3, [ ]

should do the trick.  you could easily generalize that.

hs


From m.faassen at vet.uu.nl  Fri Dec 15 11:02:17 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 15 Dec 2000 16:02:17 GMT
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu> <slrn92tu7t.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com> <tcmay-8BCFAB.23025408122000@news.got.net> <90uddk$31f@gap.cco.caltech.edu> <tcmay-4AACBE.16294509122000@news.got.net> <3A388DF4.836440B4@caltech.edu> <XM5_5.7660$4y4.196807@e420r-atl1.usenetserver.com>
Message-ID: <91dfa9$cfa$1@newshost.accu.uu.nl>

Steve Holden <sholden at holdenweb.com> wrote:
> Nathan Gray <nospam at caltech.edu> wrote:
>> Tim May wrote:
>>
>> > In article <90uddk$31f at gap.cco.caltech.edu>, Nathaniel Gray
>> > <nospam at caltech.edu> wrote:
>> >
> [snip, snippety, snip, snip; snip, snip]
>>
>> How about we skip it?
>>
> Oh, sorry, I thought you said "snip".  Yes, skipping it would be nice :-)

> Here endeth the flamewar (unless I've now offended you both and become the
> common enemy, in which case I apologize to the whole group and promise not
> to respond to this thread again).

Excellent work! You haven't offended me; best post in the thread yet!

Regards,

Martijn



-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From gogo at trust-me.com  Mon Dec 11 07:49:21 2000
From: gogo at trust-me.com (gogo at trust-me.com)
Date: Mon, 11 Dec 2000 12:49:21 GMT
Subject: mxODBC + MSSQL = code -9 error (again)
References: <90m7qv$rj2$1@nnrp1.deja.com>
Message-ID: <3a34cc98.10510959@192.168.0.1>

Hello.

I encountered the same problem and compiled a version of mxODBC.dll
that does not throw an error if a type is unknown. Instead it tries to
return a string representation of the value by default.

So the problem starts at line 2039 of mxODBC.cpp that comes with the
package. I added the default behaviour there.

Regards, Gogo.




On Wed, 06 Dec 2000 20:33:35 GMT, Matt <mksql at my-deja.com> wrote:

>I am just starting to use the mxODBC package, and have come across an
>issue that has been posted in the newsgroups before, but I have not yet
>found a solution.
>
>I am using ActiveState Python 2.0 "ActivePython 2.0, build 202
>(ActiveState Tool Corp.) based on Python 2.0 (#8, Oct 19 2000,
>11:30:05) [MSC 32 bit (Intel)] on win32", on a Windows 2000
>Professional PC, and attempting to connect to a Microsoft SQL Server
>7.0 database, using a simple test script adapted from your
>documentation. Thes test script, included below, works fine when
>connecting to an Access database, so I am assuming my installation is
>valid.
>
>But, when I connect to a SQL Server database, and have a result set
>that includes a Unicode character data  (SQLDMO_DTypeUVarchar -9
>Variable length) type, I get the "InterfaceError: SQL type (code -9)
>not implemented" error. This is avoidable since my tables to not use
>Unicode fields, but apparently some of the SQL Server system tables do.
>
>The script below, when run against a MS Access DSN, properly returns a
>list of tables. But it appears the same call against the MS SQL DSN
>returns Unicode data in the result set (in c.tables).
>
>***
>
>import ODBC
>db = ODBC.Windows.Connect('SQLTest','sa')
>c = db.cursor()
>c.execute('select count(*) from testdata')
>c.fetchone()
>c.tables(None,None,None,None)
>ODBC.print_resultset(c)
>c.close()
>db.close()
>
>Traceback (most recent call last):
>  File "d:\python20\pythonwin\pywin\framework\scriptutils.py", line
>301, in RunScript
>    exec codeObject in __main__.__dict__
>  File "D:\Python20\Lib\ODBC\Misc\sqltest.py", line 7, in ?
>    ODBC.print_resultset(c)
>  File "d:\python20\lib\ODBC\ODBC.py", line 75, in print_resultset
>    lines = apply(format_resultset,(cursor,headers),kws)
>  File "d:\python20\lib\ODBC\ODBC.py", line 33, in format_resultset
>    rows = cursor.fetchall()
>InterfaceError: SQL type (code -9) not implemented
>
>
>Sent via Deja.com http://www.deja.com/
>Before you buy.



From ullrich at math.okstate.edu  Wed Dec 20 10:15:13 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Wed, 20 Dec 2000 15:15:13 GMT
Subject: Python in "math through programming" curriculum
References: <2g2t3t06r2agrrjmfnr41eu4lc2u39a8qi@4ax.com> <3A3EFF47.67DE69CF@gte.net> <513v3toslkjkggh9oumlf20qc9bi1gdtif@4ax.com> <3A3FC677.FDB19937@gte.net> <CeP%5.521$AH6.45315@newsc.telia.net>
Message-ID: <91qidp$38j$1@nnrp1.deja.com>

In article <CeP%5.521$AH6.45315 at newsc.telia.net>,
  "Fredrik Lundh" <fredrik at effbot.org> wrote:
> Clifford J. Nelson wrote:
> > I'm sure I don't attach ANY weight to any of your opinions.
> > I've seen your web sites about Synergetics and programming!
>
> clifford, please keep your personal problems off
> comp.lang.python.

If you're gonna referee these things you should try to be
impartial. Looks to me like the other guy was the one
who started the discussion of persons as opposed to issues:

>> I'm not sure I need to attach much weight to any of your
>> opinions.  ...
>>
>> Kirby

> I'm sure I don't attach ANY weight to any of your opinions.
> I've seen your web sites about Synergetics and programming!
>
>   Cliff Nelson

> </F>
>
>

--
Oh, dejanews lets you add a sig - that's useful...


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


From dante at oz.net  Fri Dec 29 17:54:25 2000
From: dante at oz.net (Michael Esveldt)
Date: 29 Dec 2000 22:54:25 GMT
Subject: Interesting behaviour of the assignment
References: <92f4gk$qk6$2@news.nuri.net> <92fhsr01u88@news1.newsguy.com> <mailman.978059501.14500.python-list@python.org>
Message-ID: <92j4n1$mhu$0@216.39.162.232>

[Thomas Wouters]
> In other words, 'x == y' first does 'x is y', and if it's true, the 
> result
> of the operation is true. Only if 'x is y' is false will it start the
> (possibly very expensive) comparison-by-contents operation.

This is a somewhat ignorant question, but if 'x == y' checks for 'x is 
y' first, why use 'is' at all? If you only get a tiny bit of speed out 
of using 'is' is it worth it to have this in the language at all? (Not 
that it would ever change, just curious.)

Michael Esveldt
-- 
dante at oz.net - Michael Esveldt - #FightThePower on Openprojects


From DavidA at ActiveState.com  Fri Dec  8 12:37:24 2000
From: DavidA at ActiveState.com (David Ascher)
Date: Fri, 8 Dec 2000 09:37:24 -0800 (Pacific Standard Time)
Subject: Looking for compile.py
In-Reply-To: <3a30c245.95478818@news.wanadoo.fr>
Message-ID: <Pine.WNT.4.21.0012080936480.4108-200000@loom>

> I am looking for the David Ascher's compile.py script to create
> Microsoft Visual C++ project files. 
> The link I have found on
> http://www.python.org/doc/current/ext/win-cookbook.html is :
> 
>  http://starship.python.net/crew/da/compile/
> 
> But it doesn't work ((111) Connection refused) .

Sorry, that site is a mess.  The file is attached.  I'll look for an
alternative hosting solution.

--david ascher
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: compile.py
URL: <http://mail.python.org/pipermail/python-list/attachments/20001208/cf179b7b/attachment.ksh>

From rhamorim at my-deja.com  Thu Dec  7 07:05:49 2000
From: rhamorim at my-deja.com (rhamorim at my-deja.com)
Date: Thu, 07 Dec 2000 12:05:49 GMT
Subject: PIL binary for Python 2.0 for Windows ?
References: <3a2f472f$0$235@hades.is.co.za>
Message-ID: <90nueq$6jv$1@nnrp1.deja.com>

I have the DLLs compiled statically with jpeg and zlib, although I
couldn't link imagingtk.dll statically to tcl/tk - thus you need
tcl83.dll and tk83.dll on your path.

I need a place though to make them available. If you have a place (or
want me to email it to you) just answer this message (on the newsgroup -
 I rarely check my deja account).

Best regards,

Roberto Amorim

In article <3a2f472f$0$235 at hades.is.co.za>,
  "Leon Booyens" <leonb at pqafrica.co.za> wrote:
> Where can I find a binary distribution for Windows of the Python Image
> Library 1.1.1. (It is not as yet available from Pythonware as a
binary and I
> unfortunatly have no means of compiling the code on a windows
platform)
>
> Any help would be appreciated.
>
> Thanks
> Leon
> (leonb at pqafrica.co.za)
>
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From aleaxit at yahoo.com  Fri Dec 15 08:40:46 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 15 Dec 2000 14:40:46 +0100
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <3a36c85e.4409136@News.CIS.DFN.DE> <owen_nospam-1312001131590001@cs204-90.fsmodem.washington.edu> <3A37DD31.C9A5F18C@pacbell.net> <91bhlh$c64u$1@nntp6.u.washington.edu>
Message-ID: <91d7270niu@news1.newsguy.com>

"Russell E. Owen" <owen at astroNOJNK.washington.edu.invalid> wrote in message
news:91bhlh$c64u$1 at nntp6.u.washington.edu...
    [snip]
> Smalltalk is my favorite language. However, I am the only one at the
> observatory who knows it and feels it is worthwhile. As I fight Python's
> disgusting GUI I keep wavering, and I may yet give up and switch to

Which one of Python's many available GUIs do you find disgusting?  Me,
I like wxPython, and HTML-based interfaces, but I realize Tkinter's
more widespread -- and there are just SO many others that, if I was
GUI-oriented at all, I'd be in a tizzy about trying them all (but I'm
more of a command line oriented dinosaur, so...).  Or, do you mean
IDE?  As in, IDLE vs BoaConstrictor vs etc etc?  For the latter,
ActiveState is in the early stages of developing IDE's (for Python,
Perl, and XSLT) based on VisualStudio .NET, and others (same
langugaes) based on Kommodo (open-source, cross-platform).  The
Python Refactoring Browser, aka 'bicycle repair-man', may also be
worth exploring, assuming you share the 'refactoring culture' which
seems to be part of Smalltalk (and works so well with Python too);
it currently lives on sourceforge.

None of this will change the _language_, so, if Smalltalk is what
you truly hanker for, I would suggest you seriously consider using
it instead!  It *is* a great language.  But I'm somewhat surprised
of seeing 'the' (singular?) 'GUI' of Python rubbished and considered a
key reason to give up on the language, when you get so much choice
of GUI's (and IDE's) to go with it.


Alex





From akrobat at my-deja.com  Thu Dec 14 08:38:22 2000
From: akrobat at my-deja.com (akrobat at my-deja.com)
Date: Thu, 14 Dec 2000 13:38:22 GMT
Subject: mouse input output
Message-ID: <91aigd$mc1$1@nnrp1.deja.com>

is there a module which manage da mouse io in python

thx by advance


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


From dman at res137a-039.rh.rit.edu  Sat Dec 16 12:32:41 2000
From: dman at res137a-039.rh.rit.edu (D-Man)
Date: Sat, 16 Dec 2000 12:32:41 -0500
Subject: while true: !!!
In-Reply-To: <slrn93n70i.fuu.grey@teleute.rpglink.com>; from grey@despair.rpglink.com on Sat, Dec 16, 2000 at 04:45:07PM +0000
References: <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <915jcf0c6o@news2.newsguy.com> <9162p3$7t2$1@animus.fel.iae.nl> <MBWZ5.334$h67.37748@newsread1.prod.itd.earthlink.net> <3A397C74.6C8906E1@cosc.canterbury.ac.nz> <91d5520m6j@news1.newsguy.com> <wvw_5.10652$h67.670217@newsread1.prod.itd.earthlink.net> <slrn93mh41.1nk.scarblac@pino.selwerd.nl> <slrn93n70i.fuu.grey@teleute.rpglink.com>
Message-ID: <20001216123241.A32055@westsidecnc.rh.rit.edu>

On Sat, Dec 16, 2000 at 04:45:07PM +0000, Steve Lamb wrote:
> 
> while 1:
>   if cond:
>     break
>   do something
> 

why not

while ( !cond ) :
	do something

It involves less typing and more importantly shows the next reader
when the loop should end.  I strongly dislike the fake infinite loops
like "while 1 :".

-D


PS.  I haven't been following most of this thread, so forgive me if
this has already come up.




From emile at fenx.com  Mon Dec 25 13:54:54 2000
From: emile at fenx.com (Emile van Sebille)
Date: Mon, 25 Dec 2000 10:54:54 -0800
Subject: classes (was Re: Same again please for OOP)
References: <92708d0g4g@news1.newsguy.com>, <mailman.977623149.8184.python-list@python.org> <9257mf0sl1@news1.newsguy.com> <roy-E5B925.13324224122000@news.panix.com> <925khi019hs@news1.newsguy.com> <kubsu1t81x.fsf@lasipalatsi.fi> <mailman.977739488.28224.python-list@python.org>
Message-ID: <928569$6ckoi$1@ID-11957.news.dfncis.de>

Just to throw in my $.02.

If not for __getattr__ and __setattr__, I probably wouldn't
have continued with Python.  I'm not sure I would have found
anything better, so I may have come back anyway, but these
two automagic functions allowed me to write a proof of
concept model in about two days, when in other environments
I had nothing worthwhile after two weeks.

--

Emile van Sebille
emile at fenx.com
-------------------


"Moshe Zadka" <moshez at zadka.site.co.il> wrote in message
news:mailman.977739488.28224.python-list at python.org...
> On Mon, 25 Dec 2000, "Alex Martelli" <aleaxit at yahoo.com>
wrote:
>
> > Other languages may well-nigh force one to frame
> > things up in terms of accessors/mutators, but that's an
> > issue with *those* languages -- not one we should copy
> > in Python, which lets us be more direct and simple.
>
> You'd be right, if Python's __getattr__ would be better.
There are a couple
> of PEPs on the issue (Barry's, which AFAIK was rejected,
and Paul's), just
> to show many people in the Python's development team are
dissatisfied. If
> you've ever written a __getattr__ method, you probably
know what a pain it
> is. In contrast to the rest of Python, where I can code up
200-line application
> with something on the order of 2 (easily found) bugs,
__getattr__ always
> seems to lead to twisted recursion bugs, and all kinds of
headaches.
> *Particularily* if you also mix in __setattr__.
>
> I've learned the hard way not to write __getattr__
methods. Currently, none
> of my code contains them. True gurus, like Gordon
McMillan, can use __getattr__
> (e.g., in filecmp.py), to get lazy evaluation of
attributes, but I tend to
> fear those things.
>
> The "computed attributes" interface needs to be improved
befroe Python can
> be said to be not one of *those* languages.
>
> --
> Moshe Zadka <sig at zadka.site.co.il>
> This is a signature anti-virus.
> Please stop the spread of signature viruses!
>




From jcollins at endtech.com  Fri Dec 22 14:56:37 2000
From: jcollins at endtech.com (Jeff Collins)
Date: Fri, 22 Dec 2000 12:56:37 -0700 (MST)
Subject: Embedding Python, PyImport_* etc etc...
In-Reply-To: <977506990.851496@lave>
Message-ID: <Pine.LNX.4.21.0012221240530.18882-100000@euclid.endtech.com>

On Fri, 22 Dec 2000, t o b e wrote:
<snip>

Have you looked at Tools/freeze in the Python distribution?  This tool
provides the ability to create a single executable from set of python
modules.  Since you are embedding Python in another application, you won't
need all of the contents of the generated frozenmain.c file, but you can
extract the compiled bytecodes for each module (they appear as C arrays)
and the module lookup table (PyImport_FrozenModules) for use in you code.
You'll also have to set the flag Py_FrozenFlag to indicate to Python where
to find the modules.  Try an example and see what happens - it's use is
straightforward.


-- 
Jeffery D. Collins, Ph.D.
Sr. Software Developer
Endeavors Technology, Inc.





From meowbot at meowing.net  Wed Dec 13 05:55:02 2000
From: meowbot at meowing.net (Meowbot from Outer Space)
Date: 13 Dec 2000 05:55:02 -0500
Subject: Plan9 port of python
In-Reply-To: D-Man's message of "Tue, 12 Dec 2000 19:23:02 -0500"
References: <914hfv$pi1$1@nnrp1.deja.com>
	<87ofyhood3.fsf@litterbox.meowing.net>
	<20001212192302.G16853@westsidecnc.rh.rit.edu>
Message-ID: <873dfsoak9.fsf@litterbox.meowing.net>

> I had heard of Plan9 briefly a couple of years ago.  I'm just
> wondering who has/uses Plan9.

Most Plan 9 installations are currently found in academentia, this
being an artifact of Lucent's licensing terms up to this year.  The
no-charge source distribution from Bell Labs and reasonably priced
shrinkwrap packages only appeared *very* recently, so the OS is still
somewhat obscure.  It also emphasizes cluster-ish distributed
environments more than standalone systems, though it will run without
the dedicated file and compute swervers mentioned so prominently in
the press releases.

RMS hates the license, so the deeply religious may wish to avert their
eyes.


Plan 9 FAQ: <URL:http://www.fywss.com/plan9/>

Newsgroup: <URL:news:comp.os.plan9>

List version of above: <URL:http://cse.psu.edu/mailing_lists/listinfo/9fans>
  [ObPython: the list runs on Mailman]

Bell Labs distribution and documentation: <URL:http://plan9.bell-labs.com>

Commercial distribution: <URL:http://www.vitanuova.com/>


From ivdkleyn at xs4all.nl  Mon Dec 11 10:08:18 2000
From: ivdkleyn at xs4all.nl (Iwan van der  Kleyn)
Date: Mon, 11 Dec 2000 16:08:18 +0100
Subject: Microsoft SOAP not compatible with PythonWare's SOAP?
Message-ID: <912q8s$c4i$1@news1.xs4all.nl>

Hi there,

I'm developing an Intranet-site with Python. Part of the needed
business-logic is already implemented as Visual Basic (COM) components
(let's call this a "legacy-sytem" :-). I've investigated several means of
integrating them into my intranet-application and a low-cost, conventient
way would be to build a SOAP-interface which I would access with the
SOAPlib.py from Pyhtonware (http://www.pythonware.com/products/soap/) .

However, during testing I've not been able to retrieve any data from the
test MS SOAP components. The soaplib seems to working fine, though. Is this
the result of "competing" SOAP implementations? Is there any info availbale
about this topic? (? haven't found any on the internet)

Regards,

Iwan




From slhath at home.com  Thu Dec 28 15:42:51 2000
From: slhath at home.com (Scott Hathaway)
Date: Thu, 28 Dec 2000 20:42:51 GMT
Subject: how do i check if a file exists with python?
References: <RlM26.98766$yR4.2639444@news1.rdc1.tx.home.com> <92g5v8$r5k$1@nnrp1.deja.com>
Message-ID: <fBN26.99045$yR4.2640173@news1.rdc1.tx.home.com>

Thanks for both responses!

It helps to know where to look.

:)

Scott


"David C. Ullrich" <ullrich at math.okstate.edu> wrote in message
news:92g5v8$r5k$1 at nnrp1.deja.com...
> In article <RlM26.98766$yR4.2639444 at news1.rdc1.tx.home.com>,
>   "Scott Hathaway" <slhath at DownWithSpam.home.com> wrote:
> > I have searched Deja, looked in two pyton books, and searched the
> python.org
> > site trying to find out this simple question.
>
> Have you tried the _documentation_?
>
> > Can someone please give me the call and the module it lives in?
>
> I don't recall what it is exactly...[five minutes later]:
>
> One way is with glob:
>
> import glob
>
> def FileExists(filename):
> return len(glob.glob(filename)) > 0
>
> if FileExists(r'c:/windows/notepad.exe'):
> print 'yup'
> if not FileExists(r'c:/windows/notepad.exk'):
> print "I didn't think so."
>
> May or may not be the Right way.
>
> > Thanks,
> > Scott
> >
> >
>
> --
> Oh, dejanews lets you add a sig - that's useful...
>
>
> Sent via Deja.com
> http://www.deja.com/




From fluxent at yahoo.com  Wed Dec  6 13:27:38 2000
From: fluxent at yahoo.com (Bill Seitz)
Date: Wed, 06 Dec 2000 18:27:38 GMT
Subject: CGI File Upload
References: <p8ce0tcgtt2bksb3adrf9cfr2i8egr4ff5@4ax.com> <8u7fn7$jpf$1@la-mail4.digilink.net> <90j10o$5p5$1@nnrp1.deja.com>
Message-ID: <90m0eg$kl7$1@nnrp1.deja.com>

In article <90j10o$5p5$1 at nnrp1.deja.com>,
  Bill Seitz <fluxent at yahoo.com> wrote:
> In article <8u7fn7$jpf$1 at la-mail4.digilink.net>,
>   "Pete Shinners" <pete at visionart.com> wrote:
> > "Simon Faulkner" <news at titanic.co.uk> wrote
> > > I would like users to be able to upload .jpeg files via a web page
> > > <input type=file>
> > >
> > > What python code do I need to be able to save the file to a
> directory
> > > on the hard disk?
> >
> > i have some code that does this.. you'll probably want to clean
> > it to suit your own purposes, but here it is. at some point i
> > wanted to add PIL support to resize the images to a certain
> > resolution and convert them to be JPG, but that day hasn't come
> > yet... heh. note it also allows the user to save a comment as a
> > .TXT file. anyways, you can use it as a start for whatever you want
>
> Hmm, this works for me with text data (though the file length is
> slightly different: maybe line-end changes from Win to *nix?). With
> binary it's totally off. Uploading a JPEG I consistently only get 109
> bytes written to the file system. With GIFs I consistently get ~500
> bytes (but not always exactly the same number). With an MsWord file I
> get a bigger file written, but still not the whole thing (60kb out of
> 75kb).
>
> What do I do next?

Some recommendations were made by Barry Pederson
(bpederson at geocities.com). I didn't need them since setting the mode
was sufficient, but wanted to share them anyway, since others might
find them necessary/helpful...
--------

I was recently working on a CGI that did file uploads under
Win32/Apache, and found these 3 things needed adjusting for it to work
properly:

1. Make sure you use a fixed version of the cgi.py module.  The
the version in 2.0 or earlier reads the entire file upload
into memory, which is impractical in the case of multi-megabyte
uploads.  It's supposed to be fixed in Python 2.1, but
you can fix the 2.0 or earlier version yourself by commenting
out the 3 or so lines that call "self.lines.append()"

2. If on Win32, be sure to either arrange to run python with the
"-u" parameter, or with the PYTHONUNBUFFERED environment variable
set to 1.  Otherwise the uploaded files will be truncated if
characters such as Ctrl-Z appear in them.

This can be done in the Apache config with the SetEnv directive:

     SetEnv PYTHONUNBUFFERED 1

right after the ScriptAlias directive for example.

3. In addition to the fix listed above, you also will need a tweaked
version of rfc822.py (which is called by cgi.py), that puts a
try/except wrapper around the one call to tell() that doesn't
already have one.  It's been submitted as a bug to the Python guys,
and Guido says he's made my suggested fix, so it should also be in
Python 2.1.


Sent via Deja.com http://www.deja.com/
Before you buy.


From glandrum at my-deja.com  Fri Dec 29 18:03:50 2000
From: glandrum at my-deja.com (Greg Landrum)
Date: Fri, 29 Dec 2000 23:03:50 GMT
Subject: Is 100,000 entries big for a dictionary?
References: <92iug4$vqb$1@nnrp1.deja.com>
Message-ID: <92j58k$55c$1@nnrp1.deja.com>

In article <92iug4$vqb$1 at nnrp1.deja.com>,
  rturpin at my-deja.com wrote:
> I'm in the process of architecting an application with medium sized
data
> sets. Python dictionaries are tempting as part of the representation
> mechanism. How do they perform when they have 100,000 entries? Can you
> offer any other experience with this size data set in Python?
>

One of the nice things about python is that it's easy to test things
like this.  The following simple test:
#--------------
import time

def populate(n):
  d = {}
  for i in xrange(n):
    d[i] = i
  return d

def access(n,d):
  for i in xrange(n):
    foo = d[i]


n = 1e5

print 'populating'
t1 = time.clock()
d = populate(n)
t2 = time.clock()
print '\tthat took %f seconds'%(t2-t1)

print 'accessing'
t1 = time.clock()
access(n,d)
t2 = time.clock()
print '\tthat took %f seconds'%(t2-t1)
#------------
runs in about less than a second on my crappy desktop machine.
(350MHz PIII running Win2K)

-greg


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


From johngrayson at home.com  Thu Dec 28 07:44:33 2000
From: johngrayson at home.com (John Grayson)
Date: Thu, 28 Dec 2000 12:44:33 GMT
Subject: Is this a BUG?
References: <3A4B0C5E.1E7976D2@wielun.dhs.org> <mailman.978004690.11112.python-list@python.org>
Message-ID: <92fcjh$611$1@nnrp1.deja.com>

In article <mailman.978004690.11112.python-list at python.org>,
  Thomas Wouters <thomas at xs4all.net> wrote:
> On Thu, Dec 28, 2000 at 10:48:14AM +0100, QdlatY wrote:

>
> It's the space inherit in the system. (See the documentation for
'print' --
> it's supposed to do that :) You have to forcibly oppress it to avoid
the
> space:
>
> >>> print "!%s!"%1
> !1!
>
> --


We must form a party to ban the forcible oppression of spaces!

Who do they think they are? They won't get away with it!


can't-help-being-a-pedant-ly y'rs - john


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


From avv at quasar.ipa.nw.ru  Mon Dec  4 20:09:08 2000
From: avv at quasar.ipa.nw.ru (Alexander V. Voinov)
Date: Mon, 04 Dec 2000 17:09:08 -0800
Subject: mp3 module
References: <3A2BB89B.D4C30300@ira.uka.de> <3A2C3F39.BEB46122@quasar.ipa.nw.ru>
Message-ID: <3A2C4034.62BAFFBC@quasar.ipa.nw.ru>

Hi

"Alexander V. Voinov" wrote:
> write(os.popen("mpg123 -", "w"), my.brand.big.blob)?

Oops, 'write' is a method, but it's just an idea anyway.

Alexander


From jkndeja at my-deja.com  Fri Dec 15 11:58:32 2000
From: jkndeja at my-deja.com (jkndeja at my-deja.com)
Date: Fri, 15 Dec 2000 16:58:32 GMT
Subject: Solution: Direct access to Printer I/O lines
References: <917k9b$9i9$1@nnrp1.deja.com> <slrn93hrjn.2vp.ssthapa@localhost.localdomain> <91cnp3$g59$1@nnrp1.deja.com> <91d23f0jk7@news1.newsguy.com>
Message-ID: <91dijj$5hn$1@nnrp1.deja.com>

Hi Alex
    Thanks very much for your reply!

My suggestion about having singleton objects was because I wanted to
avoid clashes over what is, after all, a contrained resource.


# Facetious example-parta: module controlpowerstation.py:

import directio
mylpt = directlpt.directlpt(1)
mylpt.out(directlpt.DATA, SHUTDOWN)

...

# Facetious example-partb: module ballisticmissile.py:

import directio
mylpt = directlpt.directlpt(1)
mylpt.out(directlpt.DATA, LAUNCH)


I'll consider your suggestions further though. Thanks again.

    Jon N


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


From pyth at relax.dk  Thu Dec  7 07:55:22 2000
From: pyth at relax.dk (pyth at relax.dk)
Date: Thu, 7 Dec 2000 13:55:22 +0100 (CET)
Subject: list -> tuple
Message-ID: <Pine.LNX.4.21.0012071349420.22360-100000@relax.dk>

Hi all, I'm new... so don't hate me, if it is a stupid question  :o)

I am trying to do something like this:
how = '%2s%2s.%1s';
vals = ('ab','cd','e');
print how % vals;

Well this works fine, but I've tried to use a list... that doesn't
work... not for me anyways ;o)

My problem is, that I'm getting a list, so I have to convert it to a
tuple, before I can use it... any easy way to do that
list2tuble(list,tuple) ? :o)

Or is it possible to do the "print how % vals", with a list, somehow? :o)

Thanx :o)

Dan Larsen




From sholden at holdenweb.com  Thu Dec 14 10:32:32 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 14 Dec 2000 10:32:32 -0500
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu> <slrn92tu7t.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com> <tcmay-8BCFAB.23025408122000@news.got.net> <90uddk$31f@gap.cco.caltech.edu> <tcmay-4AACBE.16294509122000@news.got.net> <3A388DF4.836440B4@caltech.edu>
Message-ID: <XM5_5.7660$4y4.196807@e420r-atl1.usenetserver.com>

Nathan Gray <nospam at caltech.edu> wrote:
> Tim May wrote:
>
> > In article <90uddk$31f at gap.cco.caltech.edu>, Nathaniel Gray
> > <nospam at caltech.edu> wrote:
> >
[snip, snippety, snip, snip; snip, snip]
>
> How about we skip it?
>
Oh, sorry, I thought you said "snip".  Yes, skipping it would be nice :-)

Here endeth the flamewar (unless I've now offended you both and become the
common enemy, in which case I apologize to the whole group and promise not
to respond to this thread again).

regards
 Steve





From tim.one at home.com  Sun Dec 24 13:56:46 2000
From: tim.one at home.com (Tim Peters)
Date: Sun, 24 Dec 2000 13:56:46 -0500
Subject: Some Python 2.1 ideas
In-Reply-To: <005f01c06d01$d8114820$74eb0b18@C322162A>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEGEIFAA.tim.one@home.com>

Note that the deadline for 2.1 PEPs has already passed.  Any other proposed
core changes will need a PEP for, at earliest, 2.2.

    http://python.sourceforge.net/peps/




From donn at u.washington.edu  Fri Dec 22 18:27:51 2000
From: donn at u.washington.edu (Donn Cave)
Date: 22 Dec 2000 23:27:51 GMT
Subject: A Suggestion for Python Dictionary/Class
References: <920fe9$c20$1@hecate.umd.edu>
Message-ID: <920o1n$a7fi$1@nntp6.u.washington.edu>

Quoth William Djaja Tjokroaminata <billtj at y.glue.umd.edu>:
...
| Therefore, is there any way for me to define a class, with some definite
| members just like in C/C++/Java, so that no new members can be defined?  I
| am trying to do it, but using some conventions, not using the Python
| language features.
|
| Because Python provides a read-only list in terms of tuple, I think it
| will be nice if Python provides another class type with fixed members,
| just like a dictionary with read-only keys.  Any opinion, anyone?

When you say "using some conventions, not using the Python language
features", do you mean to reject a solution like this?

    class Fixed:
        def fixdict(self):
            for a in self.fixed_attributes:
                self.__dict__[a] = None
        def __setattr__(self, a, v):
            if self.__dict__.has_key(a):
                self.__dict__[a] = v
            else:
                raise AttributeError, a

    class Plant(Fixed):
        fixed_attributes = ('flower', 'stem', 'root', 'leaf')
        def __init__(self, n):
            self.fixdict()
            self.root = n

    p = Plant(5)
    print p.root
    print p.stem

I hope that will work for you, because it seems to provide the
behavior you describe and it works already, with no changes to
Python.

	Donn Cave, donn at u.washington.edu


From freydaddy at kricket.net  Tue Dec 12 00:58:53 2000
From: freydaddy at kricket.net (Matt Frey)
Date: Tue, 12 Dec 2000 05:58:53 GMT
Subject: compiled or interpreted?
Message-ID: <w8jZ5.74702$IP1.2487256@news1.giganews.com>

Is Python a compiled language that can be run on a platform like c++, or is
it interpreted like java, or is it in a class of its own?




From jiva at no-spam.com  Sat Dec  9 19:08:31 2000
From: jiva at no-spam.com (Jiva DeVoe)
Date: Sun, 10 Dec 2000 00:08:31 GMT
Subject: corba, soap, com, etc...
References: <pKsY5.58762$yR4.1496678@news1.rdc1.tx.home.com>
Message-ID: <3QzY5.9825$dg6.637046@pouncer.easynews.com>

Scott Hathaway wrote:

> Hello Everyone,
> 
> I need some advice.  I want to write an api engine in python that will be
> used as the back end to a gui.  The gui will not be written in python.  I
> need a way to expose the api I create that is cross platform.  On Windows,
> I
> would use COM.  Since I need cross platform, what are my options and what
> are the advantages and disadvantages.  I know about CORBA, but I do not
> know
> if CORBA can be used from python.  I have also heard about SOAP, but is it
> only web based?
> 
> If there is a corba or soap interface to python, can you please point me
> to it?
> 
> Your input is greatly appreciated!
> 
> Thanks!
> Scott
> 
> 

Check out fnorb.  http://www.fnorb.org  It's an excellent open source CORBA 
implementation for python.




From alvaro at linuxfan.com  Thu Dec 14 17:31:18 2000
From: alvaro at linuxfan.com (Alvaro Lopez Ortega)
Date: Thu, 14 Dec 2000 23:31:18 +0100 (CET)
Subject: Capturing stdout from a console process?
In-Reply-To: <91b8q3$aqu$1@nnrp1.deja.com>
Message-ID: <Pine.LNX.4.21.0012142330080.694-100000@servidor.home.alo>

On Thu, 14 Dec 2000, Sam Schulenburg wrote:

> I am using <os.system("xcopy c:*.* d: /v")> within a python program to
> copy files from one directory to another under windows NT. When the
> program executes, a new console window is created with the results of
> the xcopy command being displayed in the console window. Is their a way
> to capture this output by python so I can log error messages?

	Look the popen() function. :-)

-- 
Greetings, alo.




From lpointal at planete.net  Fri Dec  8 14:48:06 2000
From: lpointal at planete.net (Laurent Pointal)
Date: 8 Dec 2000 19:48:06 GMT
Subject: Effect of Ctrl-C on threaded apps
References: <90pe6a$fo8$1@nnrp1.deja.com>
Message-ID: <9004D7F60lpointalplanetenet@195.134.198.12>

[posted and mailed]

gtalvola at NameConnector.com wrote in <90pe6a$fo8$1 at nnrp1.deja.com>:

...
>Basically, is there any way to get more control over this behavior?
>I'm using 1.5.2 on NT in case there's a difference.  Ideally, I'd be
>able to register a handler for Ctrl-C that would call a function that
>I specify, then return control to the application at the same place
>it left off without an exception being thrown.
>
>- Geoff

If you have Microsoft online help installed anywhere, search for the C signal 
function. You will see that Ctrl-C support under NT is not nice for processing.
Typically, NT will start a new thread to process the Ctrl-C. And I have seen no 
way to stop it and to be able to stay in Python interpreter when Ctrl-C is 
typed (using Python signal handling module).

If you find a solution... thanks to indicate what it is.

A+

Laurent.



From danunder at columbus.rr.com  Sun Dec 24 17:45:24 2000
From: danunder at columbus.rr.com (Dan Underwood)
Date: Sun, 24 Dec 2000 22:45:24 GMT
Subject: Tkinter: title oddity
Message-ID: <3A467C2F.548F5D42@columbus.rr.com>

root.title('my nice title')
and
root.title(string='my nice title')

produce the same result... but

someOtherToplevel.title('my nice title')
works ok
but

someOtherToplevel.title(string='my nice title')
gets the message: unexpected keyword 'string'

Did I do it wrong, or is this a known inconsistency?

Thanks.
Dan.


From aahz at panix.com  Fri Dec 29 18:40:20 2000
From: aahz at panix.com (Aahz Maruch)
Date: 29 Dec 2000 15:40:20 -0800
Subject: Is 100,000 entries big for a dictionary?
References: <92iug4$vqb$1@nnrp1.deja.com>
Message-ID: <92j7d4$6qd$1@panix2.panix.com>

In article <92iug4$vqb$1 at nnrp1.deja.com>,  <rturpin at my-deja.com> wrote:
>
>I'm in the process of architecting an application with medium sized data
>sets. Python dictionaries are tempting as part of the representation
>mechanism. How do they perform when they have 100,000 entries? Can you
>offer any other experience with this size data set in Python?

I asked this question a couple of years ago.  The Timbot says it
shouldn't be a problem -- provided you have enough RAM.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

Ninth Virtual Anniversary: 2 days and counting


From syver.enstad at sensewave.com  Tue Dec  5 08:49:15 2000
From: syver.enstad at sensewave.com (Syver Enstad)
Date: Tue, 5 Dec 2000 14:49:15 +0100
Subject: activex scripting and debugging with Python win32com
Message-ID: <90irub$24b$1@troll.powertech.no>

As far as I've experienced neither the MS Script Debugger nor Visual
Interdev is able to debug Python ASP code. (Using Active State Python 2.0
and IIS 5.0/Win 2000).

Anybody knows if there are any plans to rectify this, and in what time
perspective. I would be interested in doing some work on making this feature
work as it would help me work more productively with Python at work.

Relevant skills in regard to this is: Some Python experience (last few
months), fairly strong C++ skills, fair COM skills. My handicap is primarily
that I have very limited knowledge of the plumbing between COM and Python.




From cg at schlund.de  Mon Dec  4 09:26:43 2000
From: cg at schlund.de (Carsten Gaebler)
Date: Mon, 04 Dec 2000 15:26:43 +0100
Subject: Printable time
References: <7m9n2tcp5ff10o97sr1d57qokakjqsa2m9@4ax.com>
Message-ID: <3A2BA9A3.63214210@schlund.de>

Dale Strickland-Clark wrote:

> Is there a more succinct way of getting the current local time as text than:
> time.strftime('<format>', time.localtime(time.time()))


time.ctime(time.time())


cg.


From alan.gauld at gssec.bt.co.uk  Tue Dec 12 11:05:22 2000
From: alan.gauld at gssec.bt.co.uk (Alan Gauld)
Date: Tue, 12 Dec 2000 16:05:22 +0000
Subject: Python-like compiled language
References: <913jr3$dmv$1@sylvester.vcn.bc.ca> <3A35835B.AAA92005@bioeng.ucsd.edu>
Message-ID: <3A364CC2.263C0958@gssec.bt.co.uk>

Curtis Jensen wrote:
> get LaPack lite for free.  Odepack is free and it's from llnl.  Nag is
> NOT free, it is very expensive.  

Are you sure about that? I thought I red recently that NAG was 
now open source? Their gift for the millenium or something...

I may be wrong, I don't use NAG just noticed a magazine newsbyte.

Alan g.


-- 
=================================================
This post represents the views of the author 
and does not necessarily accurately represent 
the views of BT.


From jim at interet.com  Fri Dec 15 09:11:32 2000
From: jim at interet.com (James C. Ahlstrom)
Date: Fri, 15 Dec 2000 09:11:32 -0500
Subject: Deleting a stored file in a ZipFile
References: <91aoe1$1cno$1@news.rchland.ibm.com>
Message-ID: <3A3A2694.3B305D62@interet.com>

Larry Whitley wrote:
> 
> Looking over ZipFile, it looks like it's not possible to delete a file from
> the archive.  True?

True, but you can create a new file without certain files
easily.  That is what would be required anyway to avoid
leaving a "hole" in the file to waste space.

Jim Ahlstrom


From g_m at dds.nl  Mon Dec 11 11:13:01 2000
From: g_m at dds.nl (gm)
Date: Mon, 11 Dec 2000 17:13:01 +0100
Subject: python and Applets
Message-ID: <3A34FD0D.5CA08FC1@dds.nl>

Hello,

I use python for cgi-scripting. I use a applet to send a file from a
client-browser to a server.
If the file is send, python do operations with that file.
How can I check when the file is arrived?



From mwh21 at cam.ac.uk  Wed Dec 13 05:45:25 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 13 Dec 2000 10:45:25 +0000
Subject: unpacking Python tuple in C
References: <917e64$3bp@micky.ibh-dd.de>
Message-ID: <m31yvc38hm.fsf@atrus.jesus.cam.ac.uk>

"Mike M?ller" <mmueller at dgfz.de> writes:

> Hi,
> 
> Calling Python from C I get a PythonTupleObject

C from Python, do you mean?

> PyObject *result = PyObject_CallMethod(...)
> 
> PyTuple_Check(result)  is 1 so its really a tuple!
> 
> The tuple contains only Python floats.
> 
> Using
> 
> PyArg_ParseTuple(result, "dd",  &double1, &double2)
> 
> does not change double1, or double2.

What does PyArg_ParseTuple return?  If, as you say later, result is
quite a long tuple then it will have raised an exception.

>  I can get  an item of the tuple:
> 
> PyObject* floatFromPy0 = PyTuple_GetItem(result, 0);
>   *double1= PyFloat_AsDouble(floatFromPy0)
> 
> since the tuple is pretty long and I need all of the tuple items
> PyTuple_GetItem seems to not the best solution.

Why not?  Why not use a loop and stuff the values into an array of
doubles?

> Any hints how I can get all the tuple items into C doubles?

Supplying the right number of d's to PyArg_ParseTuple, maybe?

Stepping back from your immediate problem, what are you actually
trying to do?  I can infer it from what you've said here, I'm afraid.

Cheers,
M.

-- 
  Any form of evilness that can be detected without *too* much effort
  is worth it...  I have no idea what kind of evil we're looking for
  here or how to detect is, so I can't answer yes or no.
                                       -- Guido Van Rossum, python-dev


From grante at visi.com  Thu Dec  7 12:18:06 2000
From: grante at visi.com (Grant Edwards)
Date: Thu, 07 Dec 2000 17:18:06 GMT
Subject: parallel port access?
References: <3A268378.4DA1B831@sympatico.ca> <3A272881.6F444750@engcorp.com> <3A2D6761.814ABE8E@bioeng.ucsd.edu> <slrn92su1q.12g.ssthapa@localhost.localdomain> <udhf4gk079.fsf@box.home.de>
Message-ID: <iDPX5.3227$TU6.96385@ptah.visi.com>

In article <udhf4gk079.fsf at box.home.de>, Jochen K?pper wrote:
>>>>>> Suchandra Thapa wrote on Thu, 07 Dec 2000 08:26:19 GMT:
>
>Suchandra> The parallel port and serial port controls are really
>Suchandra> platform specific.  Windows has a standard serial interface
>Suchandra> but for the parallel interface, windows 98 allows direct
>Suchandra> access but nt requires a kernel driver.  Unix systems
>Suchandra> probably have there own oddities ranging from non-pc
>Suchandra> machines which may not have a parallel or serial port to
>Suchandra> whatever access mechanisms linux/*BSD/HURD have.  The task
>Suchandra> would probably come fairly close to writing a new driver
>Suchandra> for each platform.
>
>Well, on many unices there is the common idea of a printer beeing a
>file:-) 

It could be that the questioner doesnt care about printers, and
just wants to mess with the parallel port in order to
manipulate some non-printer HW connected to it.  But I just
guessing.

-- 
Grant Edwards                   grante             Yow!  I'm EMOTIONAL
                                  at               now because I have
                               visi.com            MERCHANDISING CLOUT!!


From jvanar1 at my-deja.com  Fri Dec  8 12:26:48 2000
From: jvanar1 at my-deja.com (jvanar1 at my-deja.com)
Date: Fri, 08 Dec 2000 17:26:48 GMT
Subject: Link problems embedding python in Linux. HELP!
Message-ID: <90r5kf$q9j$1@nnrp1.deja.com>

Hello all,

I am having serious link errors trying to embed python within a C
prog.  I cant even gete something as simple as this to run:

#include <stdio.h>
#include <python1.5/Python.h>

int main(int argc, char *argv[])
{
  Py_Initialize();
  printf("Done.\n");
  Py_Finalize();
  return 1;
}

I've tryed this on gcc 2.96 AND 2.95 AND 2.92 with any luck.  Here is
the error I get:

/tmp/ccC8gDMG.o: In function `main':
/tmp/ccC8gDMG.o(.text+0x7): undefined reference to `Py_Initialize'
/tmp/ccC8gDMG.o(.text+0x1c): undefined reference to `Py_Finalize'
collect2: ld returned 1 exit status

I've read every resource available, but nothing tells me what to link
with. here is my compilation line:

gcc -o go -lpython1.5 -L/usr/lib/python1.5 go.c

I am beating my head against the wall.. Oh yeah, per the python.org
pages I've tried obtaining the link option with ">>>import
distutils.sysconfig" but it doesn't exist!  I need serious help here!

I've tryed this on Linux 6.2, 6.1 and 7.0 to no avail.

Thanks for any and all help...
-J


Sent via Deja.com http://www.deja.com/
Before you buy.


From wware at world.std.com  Wed Dec 20 13:18:55 2000
From: wware at world.std.com (Will Ware)
Date: Wed, 20 Dec 2000 18:18:55 GMT
Subject: Greetings from the Interpreter!!!
References: <mailman.977314029.15733.python-list@python.org>
Message-ID: <G5vpJK.Lpn@world.std.com>

gunter (gb at a3design.de) wrote:
> import string,time,os...

That's just too cute. I'm afraid I'm going to have to steal that,
as my .sig has been getting a little old.



From rjroy at takingcontrol.com  Mon Dec  4 17:35:37 2000
From: rjroy at takingcontrol.com (Robert Roy)
Date: Mon, 04 Dec 2000 22:35:37 GMT
Subject: Converting Perl Regex to Python!
References: <90h0a8$jgm$1@nnrp1.deja.com>
Message-ID: <3a2c1b94.417616000@news1.on.sympatico.ca>

On Mon, 04 Dec 2000 20:54:37 GMT, thelocust at my-deja.com wrote:

>I'm working on porting a program from Perl to Python, and i'm to a
>point where I need to convert their Perl regex stuff over to python.
>Now, I understand what is going on in the Perl if statement below, but
>i haven't seen to many good examples of regex stuff for Python, so i'm
>hurting for an translation
>
>The string i need to grab the info from looks like this:
>
>MSG notify timecode [00:00:00:00]
>
>if ($msg =~ /^MSG notify timecode \[(\d\d:\d\d:\d\d):\d\d\]/) {
>    $cajun->{state_clock}=$1;
>    }
>
>So, how exactly do i convert this to python?
>
>
>Sent via Deja.com http://www.deja.com/
>Before you buy.

Python regular expressions (re module) use Perl syntax so the
translation in this case is very straightforward. 

#
import re
str = 'MSG notify timecode [10:02:34:00]'

m =  re.match('^MSG notify timecode \[(\d\d:\d\d:\d\d):\d\d\]', str)
if m:
    print 'foundit', m.group(1)
else:
    print 'no such luck'





From michael at stroeder.com  Sun Dec 31 08:57:25 2000
From: michael at stroeder.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Sun, 31 Dec 2000 14:57:25 +0100
Subject: SMB Module
References: <92n1lh02evv@drn.newsguy.com>
Message-ID: <3A4F3B45.CB459045@stroeder.com>

Armin Steinhoff wrote:
> 
> In article <slrn94trot.b73.mballen at angus.foo.net>, mballen at erols.com 
> says...
> >
> >smbclient.so [..] Perl module using it.
> >I just thought one of you guys might want to do the
> >same for Python.
> 
> Makes no sense, because a SMB client mapps normaly a M$ volume to a 
> directory tree (UNIX) or a virtual drive (M$) ...

SMB means Server Message Block. SMB is a application level protocol
which is not only used for file and printer sharing.

> I don't see in that case any tasks for Python.

How about password stuff, naming service, locating shares etc.? Just
primitive examples because I'm not a SMB expert. But I think it
would make sense to provide a Python binding for this smbclient.so.

Ciao, Michael.


From jochen at pc1.uni-duesseldorf.de  Fri Dec  8 18:26:21 2000
From: jochen at pc1.uni-duesseldorf.de (Jochen =?iso-8859-1?q?K=FCpper?=)
Date: 09 Dec 2000 00:26:21 +0100
Subject: Documentation - underscores in \section
References: <Pine.LNX.4.21.0012081707130.29380-100000@saturn.ispol.com>
Message-ID: <udvgsuv6jm.fsf@box.home.de>

>>>>> Gregory Trubetskoy wrote on Fri, 08 Dec 2000 22:15:55 GMT:

Gregory> [Sorry this is perhaps somewhat off-topic]

Yes, there is comp.text.tex as well. I keep it here, as there is no
LaTeX bug/misbehaviour to see right now. In case you see a
LaTeX/TeX-problem, please crosspost and set f'up there!

Gregory> Has anyone noticed that underscores do not work in \section,
Gregory> even if preceeded by a \ (backslash).

Do you have a minimal example? The following works without any
error/warning for me; actually it gives the expected output as well:)
,----
| \documentclass{article}
| \begin{document}
| \section{Mod\_python}
| \end{document}
`----

What packages are loaded by these docs? Have you compiled them
yourself? What LaTeX/TeX versions?

Greetings,
Jochen
-- 
Einigkeit und Recht und Freiheit                http://www.Jochen-Kuepper.de
    Libert?, ?galit?, Fraternit?                GnuPG key: 44BCCD8E
        Sex, drugs and rock-n-roll


From db3l at fitlinxx.com  Fri Dec 22 20:30:20 2000
From: db3l at fitlinxx.com (David Bolen)
Date: 22 Dec 2000 20:30:20 -0500
Subject: A Suggestion for Python Dictionary/Class
References: <920fe9$c20$1@hecate.umd.edu>
Message-ID: <uhf3vsz4j.fsf@ctwd0143.fitlinxx.com>

William Djaja Tjokroaminata <billtj at y.glue.umd.edu> writes:

> In C/C++/Java, if a member does not exist and we try to access or assign
> it (because of some typo, perhaps), it will be caught by the compiler
> simply as syntax error.  Not so with Python class.
> 
> Therefore, is there any way for me to define a class, with some definite
> members just like in C/C++/Java, so that no new members can be defined?  I
> am trying to do it, but using some conventions, not using the Python
> language features.

One approach to handling this is to define a __setattr__ function in
your class, which will get called every time you try to set any
attribute of that object.  You can impose any sort of constraints
there that you want - for example, verifying that the attribute
already exists before permitting the assignment.

I believe there have been more elaborate posts in this vein to this
group in the past - try the "How often to problems arise ... ?" thread at:

http://x72.deja.com/threadmsg_ct.xp?AN=697282174.1&mhitnum=0&CONTEXT=977534927.1986658345

or more specifically the post within it from Alex Martelli that
implements a Strict class that only permits changing existing attributes:

http://x61.deja.com/threadmsg_ct.xp?AN=697340791.1&mhitnum=5&CONTEXT=977534789.910295077

> Because Python provides a read-only list in terms of tuple, I think it
> will be nice if Python provides another class type with fixed members,
> just like a dictionary with read-only keys.  Any opinion, anyone?

Since you can implement this policy your own (and if at the class
level, permit that policy to be inherited), I'm not sure it's
something that needs to be added to the core language itself.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From chris_barker at my-deja.com  Wed Dec  6 14:53:28 2000
From: chris_barker at my-deja.com (chris_barker at my-deja.com)
Date: Wed, 06 Dec 2000 19:53:28 GMT
Subject: Python: what all the hype is about?
References: <90locf$d80$1@nnrp1.deja.com>
Message-ID: <90m5fk$pdj$1@nnrp1.deja.com>

In article <90locf$d80$1 at nnrp1.deja.com>,
  aivkov at my-deja.com wrote:
> Excuse my ingnorance in the subject area. Can someone explaing me very
> briefly, why one should choose Python over Java or is it better to
> compare it with Javascript/ECMA?

Perhaps these links will be useful to you:

http://www.python.org/doc/essays/comparisons.html
(this is slightly out of date)

http://www.python.org/doc/Comparisons.html

-Chris


Sent via Deja.com http://www.deja.com/
Before you buy.


From sweisel at drew.edu  Sat Dec  9 13:42:09 2000
From: sweisel at drew.edu (Scott Weisel)
Date: Sat, 09 Dec 2000 13:42:09 -0500
Subject: Dumb Newbie Question
Message-ID: <3A327D01.9841EF98@drew.edu>

I'm trying to set up Python so I can import the programs that I wrote.  I
want to modify $PYTHONPATH so the directory with my programs is in the path
but I'm not sure how.

I thought I might be able to modify sys.path like this
sys.path = sys.path + "c:\\python16\\programs 
but I'm getting a TypeError stating "illegal argument type for built-in
operation."

Any suggestions would be appreciated
-Scott


From joconnor at mail.cybermesa.com  Wed Dec  6 20:04:03 2000
From: joconnor at mail.cybermesa.com (Jay O'Connor)
Date: Thu, 07 Dec 2000 1:04:03 GMT
Subject: Mahagony EMail Client
Message-ID: <200012070104.SAA14537@mail.cybermesa.com>

An embedded and charset-unspecified text was scrubbed...
Name: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20001207/da10479b/attachment.ksh>

From rodsenra at correionet.com.br  Mon Dec 11 19:43:44 2000
From: rodsenra at correionet.com.br (Rodrigo Senra)
Date: Mon, 11 Dec 2000 22:43:44 -0200
Subject: HPUX install
References: <913ett$u0b$1@nnrp1.deja.com>
Message-ID: <3A3574C0.DA89DCD9@correionet.com.br>

Brad wrote:
> 
> We are running HPUX 10.20 and I was wondering why I cant get python to
> install.
> 
> make does not...it gives the following error
> 
>   cd Parser ; make OPT="-O" VERSION="2.0" \  prefix="/usr/local"
> exec_prefix="/usr/local" all  cc -Ae -O -I./../Include -I.. -DHAVE_CONFIG_H
> -c pgenmain.c (Bundled) cc: warning 480: The -A option is available only with
> the C/ANSI C pro duct; ignored. (Bundled) cc: warning 480: The -O option is
> available only with the C/ANSI C pro duct; ignored. cpp: "pymem.h", line 7:
> error 4036: Can't open include file 'pyport.h'. *** Error exit code 1

I do not know of a single soul that succeded in using cc to compile
Python in HP-UX. Download gcc for HP-UX and use it instead. ;o)
This is not a solution the problem you raised, but may be a solution
to the problem you have.
HTH
Rod
-- 
Rodrigo Senra         
Computer Engineer   (GPr Sistemas Ltda)  rodsenra at correionet.com.br 
MSc Student of Reflection (IC- UNICAMP) Rodrigo.Senra at ic.unicamp.br
http://www.ic.unicamp.br/~921234 (see also   http://www.gpr.com.br)





From michael at stroeder.com  Fri Dec 29 07:49:53 2000
From: michael at stroeder.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Fri, 29 Dec 2000 13:49:53 +0100
Subject: Python vs PHP (& Java?)
References: <92f3kp01cqe@news1.newsguy.com>, <mailman.977897649.4877.python-list@python.org> <slrn94l86m.a3h.tim@degree.ath.cx> <mailman.978021012.21874.python-list@python.org>
Message-ID: <3A4C8871.DB734376@stroeder.com>

Moshe Zadka wrote:
> 
> Apache modules. Yes, now that we have mod_snake, you
> can do these things in Python, in theory: but mod_snake is still
> young, while mod_perl is quite mature.

mod_python?

Ciao, Michael.


From fgranger at teleprosoft.com  Tue Dec 26 05:05:06 2000
From: fgranger at teleprosoft.com (Fran=?ISO-8859-1?B?5w==?=ois Granger)
Date: Tue, 26 Dec 2000 11:05:06 +0100
Subject: Duplicated List...
References: <929lq7$ds1$1@hiline.shinbiro.com>
Message-ID: <B66E2BE2.20C88%fgranger@teleprosoft.com>

in article 929lq7$ds1$1 at hiline.shinbiro.com, dyo at dyo at shinbiro.com wrote
on 26/12/00 9:54:

> Hi everyone.
> 
>>>> a=[1,2,3]
>>>> b=[a,10,20,30]
>>>> b[0].append('x')
>>>> b
> [[1, 2, 3, 1, 2, 3, 'x'], 10, 20, 30]
>>>> a
> [1, 2, 3, 1, 2, 3, 'x']
>>>> a.append('xxx')
>>>> a
> [1, 2, 3, 1, 2, 3, 'x', 'xxx']
>>>> b
> [[1, 2, 3, 1, 2, 3, 'x', 'xxx'], 10, 20, 30]
> 
> Duplicated List work above..
> 
> 
> But in this case..
> 
>>>> a=[1,1000]
>>>> a
> [1, 1000]
>>>> b
> [[1, 2, 3, 1, 2, 3, 'x', 'xxx'], 10, 20, 30]
>>>> 
> 
> Why do not change List 'b' ?
> What happen List 'b' ?

a is a reference to the list. You then make a e reference to another list.

I guess that if you do
a[:] = [1,1000]
it may work ?

Salutations,
Francois Granger
-- 
fgranger at teleprosoft.com - <http://www.teleprosoft.com>
tel: +33 1 41 88 48 00 - Fax: + 33 1 41 88 04 90



From boud at rempt.xs4all.nl  Thu Dec 28 05:12:55 2000
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 28 Dec 2000 10:12:55 GMT
Subject: Kanji
References: <92dl3u$to0$1@news.idiom.com> <p6qlmt1e3yh.fsf@informatik.hu-berlin.de>
Message-ID: <92f3n7$hsf$1@news1.xs4all.nl>

Martin von Loewis <loewis at informatik.hu-berlin.de> wrote:
> "James T. Dennis" <jadestar at idiom.com> writes:

>>  I have the necessary fonts installed.  But how can I access 
>>  them from Python?

> Depends on your system, somewhat. I would recommend to use Tkinter;
> make sure you use at least Tk 8.3 and Python 2.0. Then, when you pass
> Unicode objects to the Tk functions, Tk should automatically select
> the right font (or else you can always explicitly select one). If you
> need to read data in other but UTF-8/16 encodings, you'll also need
> the JapaneseCodecs package. Check www.python.org/sigs/i18n-sig for
> details.

Likewise, PyQt supports Unicode. All you have to do is make sure you
feed the widgets Unicode strings, and they'll display anything you
want. I've just finished a small Unicode editor/character picker in
PyQt for which I'm going to write a Chinese IME today. And Kura is
a language database written entirely in Python and Qt which reached
feature-complete stage yesterday... The editor is part of the Kura 
package:

http://www.valdyas.org/linguistics

-- 

Boudewijn Rempt  | http://www.valdyas.org


From dale at out-think.NOSPAMco.uk  Wed Dec  6 07:59:38 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Wed, 06 Dec 2000 12:59:38 +0000
Subject: General default exception handler
References: <lj0n2to9vs8bmv3oo49mqgoc6h7esh8t37@4ax.com> <NZMW5.10915$lF4.165334@e420r-atl1.usenetserver.com> <138n2tojuquau637pm7obq8c4gvtl7hutd@4ax.com> <uhf4jkiuy.fsf@ctwd0143.fitlinxx.com>
Message-ID: <f1es2t0022kik40uabl79gl366k4s2ll6r@4ax.com>

David Bolen <db3l at fitlinxx.com> wrote:

>Dale Strickland-Clark <dale at out-think.NOSPAMco.uk> writes:
>
>> Interesting idea but that won't work in this case. I have to trap
>> the error inside the COM object or much of the detail is lost.
>
>Then how about just structuring your COM object so that the various
>entry points just pass control through some sort of dispatch function
>that can provide the single general purpose wrapper for unhandled
>exceptions.  Python's ability to pass in a dynamic function reference
>and arbitrary parameters through such a dispatch should make this
>fairly straight-forward.
>
>--
>-- David

That's an interesting idea.

Thanks
--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants




From sparrowhawk9909 at my-deja.com  Wed Dec 20 23:49:16 2000
From: sparrowhawk9909 at my-deja.com (sparrowhawk9909 at my-deja.com)
Date: Thu, 21 Dec 2000 04:49:16 GMT
Subject: file upload CGI error
Message-ID: <91s24b$cf8$1@nnrp1.deja.com>

I was using python 1.5.1 as CGI script with IIS
on windows NT. That was
working fine. But when I upgrade to python 1.5.2,
every time I upload
binary file such as GIF files cause CGI error.
I've tried python 1.6
and python 2.0 also, but the result was same.
Followings are the
error message.

CGI Error
The specified CGI application misbehaved by not
returning a complete set
of HTTP headers. The headers it did return are:


Traceback (most recent call last):
  File "C:\InetPub\wwwroot\python\sample2
\upload.py", line 5, in ?
    form = cgi.FieldStorage()
  File "c:\inetpub\wwwroot\cgi.py", line 869, in
__init__
    self.read_multi(environ, keep_blank_values,
strict_parsing)
  File "c:\inetpub\wwwroot\cgi.py", line 954, in
read_multi
    headers = rfc822.Message(self.fp)
  File "c:\python\lib\rfc822.py", line 92, in
__init__
    self.readheaders()
  File "c:\python\lib\rfc822.py", line 135, in
readheaders
    startofline = tell()
IOError: (0, 'Error')

Can anyone help me out ?
Thanks.




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


From agroten at htwm.de  Fri Dec  8 10:05:20 2000
From: agroten at htwm.de (Alexander Groten)
Date: Fri, 8 Dec 2000 16:05:20 +0100
Subject: reading directory structure
Message-ID: <90qtar$eve$1@narses.hrz.tu-chemnitz.de>

Hi!

I'm new in Python.
For a script I need a algorithm, that read a directory to a defined depth.
Similar FIND.

I have no plan (in the moment). Any suggestions?

Thanks!
Alex




From amk at mira.erols.com  Wed Dec  6 21:28:59 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 7 Dec 2000 02:28:59 GMT
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu>
Message-ID: <slrn92tu7t.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com>

On Wed, 06 Dec 2000 17:06:55 -0800, Nathan Gray <nospam at caltech.edu> wrote:
>It's really upsetting how often I've seen this sort of response to a polite
>question on this newsgroup lately.  Tell me, what's so terrible with asking a
>group of experts for pointers when researching a subject?  It's a perfectly
>legitimate and highly effective way to learn about something.  There's plenty

I've noticed this trend, too, though T.C. May, the person you're
replying to, has always had a low tolerance for cluelessness, as I
noted when lurking on the cypherpunks mailing list in the early 90s;
May's posts were often informative and entertaining, but not often
placid.  (I was pleased to turn up here, though I'm not sure why he's
taken up Python.)

Generally I don't bother to flame do-my-homework-for-me posters and
simply ignore them, because I don't want c.l.py to approach
comp.lang.perl.misc's reputation for unfriendliness.  Private mail to
me with such questions also gets unceremoniously deleted, though
usually in private mail people are asking for essays on _Fifth
Business_, not Python help.  (_Fifth Business_ is commonly studied in
Canadian high schools, so I often get that sort of request.  On one
occasion I was flabbergasted to get e-mail from a *parent* asking if I
had an essay on FB their child could use.)

--amk


From jschmitt at vmlabs.com  Fri Dec 15 03:16:15 2000
From: jschmitt at vmlabs.com (jschmitt at vmlabs.com)
Date: Fri, 15 Dec 2000 08:16:15 GMT
Subject: Python advocacy
References: <91bln3$lmr$1@nnrp1.deja.com> <3A39AC3E.FA75C4FF@engcorp.com>
Message-ID: <91ck0d$cmf$1@nnrp1.deja.com>

Have you read Gerald Weinerg's book called "The Psychology of Computer
Programming"?  I really enjoyed it and it's where I first encountered
the notion of "ego-less programming".  The idea is that in a code
review situation, I'm allowed to comment on your code without you
taking it personally.  That's the sort of thing I got out of the
article about advocacy.  You have permission to critique the language
without people taking personally.

http://www.geraldmweinberg.com/Bookstuff/Each_Book/Psychology.html

Another problem I have with advocacy relates to progress.  Someone here
at work made the observation that there are two types of language
designers.  There is the type that invent a language, and preach it far
and wide and not make progress on the language.  Forth was the example
used.  Then there is the type that invent a language, observe that it
has flaws, and then design a significantly better language, even if
it's backwards incompatible.  The example used for this was Wirth with
Pascal, Modula-2, Oberon, etc.

I don't know all that much about the status of Forth or Modula-2's
successors.  But I do like the idea of recognizing flaws and designing
something better, even if it means backwards incompatibility.

(Sadly, that hasn't resulted in commercial success for M-2, M-3, or
Oberon.)

If, for some reason, Python3K didn't run Python 2.0 code, I wouldn't be
too bothered.  Mostly it's because my investment in Python isn't that
great but also because I still have Python 2.0 for any legacy code.

Anyway, I like Python much more than Perl and I tell people as often as
the occasion arises.  The only thing I'm advocating here is that people
acknowledge their shortcomings (programming language shortcomings, that
is :-) and deal with them.

In article <3A39AC3E.FA75C4FF at engcorp.com>,
  Peter Hansen <peter at engcorp.com> wrote:
> jschmitt at vmlabs.com wrote:
> >
> > I just read this good article about Perl advocacy.  I think it
applies
> > to Python equally.  I found the link mentioned from /. this morning.
> >
> > http://www.perl.com/pub/2000/12/advocacy.html?wwwrrr_20001213.txt
[snip]
>
> I didn't bother reading the whole thing because I noticed early on
(and
> confirmed throughout) that the author is oblivious to the huge
> investment people put into developing systems in a particular language
> (not to mention the investment in *learning* that language) and the
fact
> that as long as the language stays popular, improves, and continues to
> serve their needs well, they get a payoff from that investment.
>
> If the language "dies", that investment is largely lost.
>
> The theory is that by advocating a language one can increase the
> likelihood of its survival and improvement.  If true, then advocating
a
> language directly serves ones own needs.
>
> There is no reason to focus on the fact that people appear to take
this
> sort of thing personally.  That's probably just human nature, and
likely
> not even a Bad Thing.  The point is there are good and wholly
rationale
> reasons for advocacy even in its current style, and you could say that
> anyone who does *not* advocate a particular language is contributing
to
> its demise (the "if you're not part of the solution, you're part of
the
> problem" theory of life :-).
>
> I think I'll just ignore whatever other, possibly valid points
> Mark-Jason Dominus may have made and carry on advocating as I always
> have.  He hasn't convinced me.
>


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


From nospam at caltech.edu  Thu Dec 14 02:39:20 2000
From: nospam at caltech.edu (Nathan Gray)
Date: Wed, 13 Dec 2000 23:39:20 -0800
Subject: Moving from Perl to Python
References: <3a3a9552.76673185@news.tcn.net>
Message-ID: <3A387928.D36F7902@caltech.edu>

Richard,

Here's something kinda neat:
http://starship.python.net/~da/jak/cookbook.html

It's a Python/Perl Phrasebook with side-by-side listings of common
idioms.  The 'last modified' on the page says 1995 so not everything will
be current, but it may be a good way to bootstrap yourself into the basics
of Python.

-n8gray
-at
-caltech
-dot
-edu

Richard Zilavec wrote:

> I just got back from two book stores, nothing on Python, I was
> shocked!  So I went online and found a pile but I hate ordering
> online, I like to browse through them first.  Anyways, I really like
> O'Reilly books, is the Programming Python really advanced or is more
> along the lines of the Programming Perl where it covers the basics to
> advanced?
>
> man python on my machine doesn't work, I'm assuming something is wrong
> or maybe its a perldoc equivalent?  Where can I find the FAQs?
>
> Thanks in advanced for any help or suggestions.
>
> --
>  Richard Zilavec
>  rzilavec at tcn.net



From jasonic at nomadicsltd.com  Tue Dec 19 22:32:07 2000
From: jasonic at nomadicsltd.com (Jason Cunliffe)
Date: Tue, 19 Dec 2000 22:32:07 -0500
Subject: MIDI extensions (WIN32, Mac)?
References: <3A3FACEF.DAF77229@cableinet.co.uk> <3A3FE5C3.76B4@loop.com> <3A3FF97B.BBF116C3@cableinet.co.uk>
Message-ID: <t409smaivhaof6@corp.supernews.com>

Not Python exactly,

..but maybe what you need ++and++ __lots__ of Fun.
Perhaps can be nicely connected to Python since there is much golden work
done already to explore:

KeyKit
http://nosuch.com/keykit/

PD
http://iem.mhsg.ac.at/pd/

enjoy!
- Jason
___________________________________________________________
Jason CUNLIFFE = NOMADICS['Interactive Art and Technology']





From jerry_spicklemire at my-deja.com  Thu Dec 21 16:10:04 2000
From: jerry_spicklemire at my-deja.com (jerry_spicklemire at my-deja.com)
Date: Thu, 21 Dec 2000 21:10:04 GMT
Subject: 2 question from a python newbie!
References: <91tl43$rko$1@neptunium.btinternet.com>
Message-ID: <91trj7$qc8$1@nnrp1.deja.com>

  G. Willoughby asked

> 1). Can python scripts be compiled to 'stand alone' .exe files?

Yes they can, see:

 http://www.mcmillan-inc.com/builder.html

> 2). What's the best book/s available for learning python from a total
> beginers point of view?

There are lots of materials on line, see:

 http://www.python.org/doc/

and you should also check out:

 http://www.python.org/psa/bookstore/

Up until just a few days ago, there was no hard copy readily available
for a newcomer to programming, that emphasized Python as the
introductory language, but now an updated version of Alan Gauld's on
line tutorial

 http://www.crosswinds.net/~agauld/

is available as 'Learn to Program Using Python', ISBN# 0201709384.

If you have any interest in XML, a bonus that comes with Sean McGrath's
book, 'XML Processing with Python', ISBN# 0130211192, is an excellent
introduction to Python.

Good Luck!
Jerry S.


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


From magnus.heino at rivermen.se  Fri Dec  8 03:23:27 2000
From: magnus.heino at rivermen.se (Magnus Heino)
Date: Fri, 8 Dec 2000 08:23:27 +0000
Subject: xml.dom
Message-ID: <90q292$k9k$1@merlin.rembrandtstr.bocholt.de>

Hi.

How do I build an xmldocument using something in xml.dom?

All examples that I have found, including those in the latest 
PyXML-package, seems to be outdated.

Thanks,

/Magnus Heino


From fredrik at effbot.org  Wed Dec 20 14:45:20 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 20 Dec 2000 19:45:20 GMT
Subject: Running pdflatex from Python
References: <3A4112DD.E8B59F3F@et.dtu.dk>
Message-ID: <k%706.701$AH6.70148@newsc.telia.net>

Brian Elmegaard wrote:
> This is posted to comp.lang.python and miktex at dsts.dk
>
> When I try to run pdflatex (miktex 1.20e) from a Python script with
> popen("pdflatex "+filename) under win98 SE I get a fault in kernel32.dll.
>
> This does not happen for pure latex.

perhaps you're using an evil version of Norton Antivirus 2000?

    http://www.python.org/2.0/

    "Incompatibility warning: Norton Antivirus 2000 can cause blue
    screen crashes on Windows 98 when a function in the os.popen*()
    family is invoked. To prevent this problem, disable Norton Antivirus
    when using Python. (Confirmed on Windows 98 Second Edition with
    Norton Antivirus version 6.10.20. The same Norton Antivirus version
    doesn't have this problem on Windows 2000. Norton Antivirus
    version 5 on Windows 98SE doesn't have this problem either.)"

</F>




From bobalex at home.com  Mon Dec 25 11:14:10 2000
From: bobalex at home.com (Bob Alexander)
Date: Mon, 25 Dec 2000 08:14:10 -0800
Subject: Some Python 2.1 ideas
References: <925tl301ub0@news1.newsguy.com>, <mailman.977590932.19335.python-list@python.org> <Cw616.23054$qN6.3308205@typhoon.we.rr.com> <3A452881.7686E717@uswest.net> <%vb16.187233$U46.5801214@news1.sttls1.wa.home.com> <3A464D5D.B541C04E@uswest.net> <mailman.977688488.25438.python-list@python.org> <20001225150250.6FA06A84F@darjeeling.zadka.site.co.il>
Message-ID: <011f01c06e8d$b70a6610$74eb0b18@C322162A>

> I think in that case a function which does the "readline_chomp" is better:
> you don't need to change many object's declerations, mix them in, not
> to mention that it doesn't help built-in types at all. Frankly, I fail
> to see why people are so afraid of functions: generic functions *enhance*
> polymorphism.

As the person who started this "chomp" thread, I'd like to chime in that I
agree that having a family of "readline-without-the-newline" functions would
be good (e.g. counterparts for both readline and readlines). But having a
stand-alone chomper might still be good since Python program landscapes are
often littered with newline-terminated strings stored away in various
variables that need to be decontaminated later. It's a frequently-occurring
need to convert between newline-terminated and non-newline-terminated
strings.

readline_chomp would have to return something like None for end-of-file,
since empty lines would be valid data.

Finally, I'd be disappointed if the name "chomp" actually survived! I just
used it in my initial post as a placeholder until a better name came along,
since some folks would likely recall it from Perl. (If this were Java, we
could just call it removeTrailingNewLineCharacter(), but that's not very
Python-esque  :-)

Bob

P.S. Merry Christmas!




From jerry_spicklemire at my-deja.com  Thu Dec 21 14:45:32 2000
From: jerry_spicklemire at my-deja.com (jerry_spicklemire at my-deja.com)
Date: Thu, 21 Dec 2000 19:45:32 GMT
Subject: Wanted: Python White Papers
References: <91oj8v$i0r$1@nnrp1.deja.com>
Message-ID: <91tmkq$lru$1@nnrp1.deja.com>

  Noah Spurrier wrote:
> Is anyone aware of any white papers to explain the use of Python
> at any big, high-profile companies?
> I'm trying to use a "look-how-these-cool-companies-use-python"
> argument to defend the use of Python for a project. That won't be my
> only argument, of course :-) But currently Python is not winning many
> management converts here based solely on its technical merits.
> Many still say, "What's Python? Never heard of it...". It helps
> a lot when I say "Google uses it!", but it would help more if I
> knew HOW it is used and WHY.
>
> Actually, the only high-profile company that uses Python that I
> can think of is Google. Is anyone familiar with their architecture?
> How does Python fit into their design?

Actually InfoSeek used Python for years, before Google existed, and
some other high profile users include -

HP, especially in the recent E-Speak project :

 http://pysol.sourceforge.net/

 http://www-uk.hpl.hp.com/people/ak/doc/webe98.html

and even though their promotional materials are very careful to stress
a "Java centric" nature of E-Speak, this page clearly includes Python
as an "e-services" development tool :

 http://devresource.hp.com/devresource/Tools/ToolLibrary.html

and don't forget IBM, apparently favoring Python over their own
products, in at least one case:

 http://www.raleigh.ibm.com/cgi-bin/bookmgr/BOOKS/EZ304100/2.3.9

 http://opendx.watson.ibm.com/dx/mailArchives/mails.html/opendx-
dev.0006/msg00211.html

 http://ncc.hursley.ibm.com/majordomo/IBM-NETREXX/archives/ibm-
netrexx.archive.0003/Author/article-9.html

 http://ncc.hursley.ibm.com/majordomo/IBM-NETREXX/archives/ibm-
netrexx.archive.0003/date/article-11.html

, and also at JavaSoft, where Python facilitates development of Java.
It seems that Java is too awkward to be optimally effective for use in
the QA role of testing itself. Go figure.

Here's a starting point that confirms Python use at a place called NASA:

 http://lheawww.gsfc.nasa.gov/~bridgman/AstroPy/matrix.html

 http://lheawww.gsfc.nasa.gov/~bridgman/AstroPy/index.html

and there's that Disney place in California, which uses Python:

 http://squeak.cs.uiuc.edu/mail/squeak/msg07429.html

plus George Lucas' Industrial Light and Magic :

 http://www.ilm.com/cgi-bin/getjob.pl?182-99

 http://www.ilm.com/cgi-bin/getjob.pl?61-01


and assorted research labs, such as -


Lawrence Berkeley :

http://smb.slac.stanford.edu/summer_school/software_workshop/WorkshopAge
nda.html


Lawrence Livermore:

 http://www.aps.org/meet/DPP00/baps/abs/S2260050.html

 http://www.foretec.com/python/workshops/1998-
11/proceedings/papers/dubois/dubois.html

 http://www-pcmdi.llnl.gov/software/cdat/


Los Alamos (including an actual White Paper!):

 http://bifrost.lanl.gov/MD/MD.html

 http://www.python.org/workshops/1997-10/proceedings/beazley.html

 http://starship.python.net/pipermail/mmtk/2000/000227.html


Oak Ridge :

 http://www.epm.ornl.gov/pvm/pvm_home.html


Sandia (et. al.):

 http://www-unix.mcs.anl.gov/~curfman/cca/web/cca_paper.html


BTW, Metacrawler is handy for this sort of thing!

Later,
Jerry S.


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


From m.j.fisher at ex.ac.uk  Wed Dec 13 12:09:18 2000
From: m.j.fisher at ex.ac.uk (M.J.Fisher)
Date: Wed, 13 Dec 2000 17:09:18 +0000
Subject: zipfile.py
Message-ID: <3A37AD3E.7233E7B7@ex.ac.uk>

Hi,

I have been having some trouble with using zipfile.py, I need to unzip
.zip files some created in windows and
some in unix.

I have the following piece of code:

import zipfile
if zipfile.is_zipfile(self.path):
   zf = zipfile.ZipFile(self.path, 'r')
   for file in zf.namelist():
        newFile = open(self.directory+'/'+ file, 'wb')
        newFile.write(zf.read(file))
        newFile.close()
   zf.close()
else:
   print 'We cannot process this file'

But when I try and use it I get the following error:

  File "fileAgent.py", line 251, in GetFileTypes
    item.ProcessZip()
  File "fileAgent.py", line 160, in ProcessZip
    newFile.write(zf.read(file))
  File "/home/sheerkahn3/mjfisher/lib/python2.0/zipfile.py", line 242,
in read
    bytes = dc.decompress(bytes)
zlib.error: Error -3 while decompressing: incomplete dynamic bit lengths
tree

I think it may be a problem with trying to unzip zip files created in
unix but I am not sure, I would be grateful for
any help you could give me.

Thanks in advance,

Shelley



From dale at out-think.NOSPAMco.uk  Sat Dec  9 09:12:04 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Sat, 09 Dec 2000 14:12:04 +0000
Subject: Unicode coercion not the same as ASCII
Message-ID: <dde43tgd9hikfq5sd6eau5tgn91akk244b@4ax.com>

Is there a particularly good reason for this apparently inconsistent behaviour?

>>> str(123)
'123'
>>> unicode(123)
Traceback (innermost last):
  File "<interactive input>", line 1, in ?
TypeError: coercing to Unicode: need string or buffer, int found

To get a number represented as a Unicode string, do I really need to do this?

>>> unicode(str(123))
u'123'

When working in ASCII it is handy to be able to stuff anything into str() to get a readable
representation of it. 

When working in Unicode, you have to test datatypes and treat them differently which isn't very
convenient.

--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants


From arturo_perez at wanadoo.es  Wed Dec  6 04:09:22 2000
From: arturo_perez at wanadoo.es (Arturo Pérez Mulas)
Date: Wed, 06 Dec 2000 09:09:22 GMT
Subject: [DAO Database in win32] What is more efficient
References: <0g8X5.152$7U6.3922@m2newsread.uni2.es> <PXlX5.117656$U46.3885672@news1.sttls1.wa.home.com>
Message-ID: <6nnX5.37$rZ6.2159@m2newsread.uni2.es>

> If you aren't using ADO you should be. DAO was replaced by ADO some time
> ago.
>
> The recordset's EOF property will be true right after the OpenRecordset()
if
> the recordset is empty.

Thank you! I will check ADO

> I don't know what you mean by "register" in this context. Your examples
are
> simply checking a table to see if any row has the desired value in a
column.
> If you don't know the database structure you will have trouble writing
> queries; SQL statements can't refer to columns that aren't defined. You'll
> generate an error, not an empty recordset.

Yeah, sorry... by register I meant "row"... I know the database structure,
but
I did not design it, and cannot change it. With that I meant that I cannot
set
a column to have "unique" data as you can within Access, so I need to
"manually"
check for the existence of a value before inserting it.

> The simplest and fastest way to see if any row has a desired value in a
> column is:
>
>   rs = db.OpenRecordset("select count(*) as 'nrows' from table where field
=
> value")
>
> The resulting recordset will have exactly one row, and one column named
> 'nrows.' The value of that column is how many rows were found where field
=
> value. If the number is 0 there were no matches.
>
> Using the database engine to scan the table will be a lot faster than
> looping over a cursor as your second example does.

That is great! Thank you so much!!!! That hits the nail.

Arturo.




From dbrueck at edgix.com  Tue Dec 12 09:29:59 2000
From: dbrueck at edgix.com (Dave Brueck)
Date: Tue, 12 Dec 2000 07:29:59 -0700
Subject: Q: Print whithout \n
In-Reply-To: <Pine.GSO.4.21.0012121444320.9661-100000@luna.dsv.su.se>
Message-ID: <014001c06448$032911e0$760214ac@PRODUT2KDAVE>

> As others have said, print with a , works, but it gives you a space
> instead.  If this disturbs you as it does me, you can use the
> write method

You can also set sys.stdout.softspace to 0, but that's annoying if you have
to do it more than once or twice:

>>> def msg(s):
... 	print 'Msg is:',
... 	sys.stdout.softspace = 0
... 	print s
...
>>> msg('Blah')
Msg is:Blah

-Dave




From mikeb at mitre.org  Mon Dec 18 06:00:35 2000
From: mikeb at mitre.org (Mike Brenner)
Date: Mon, 18 Dec 2000 06:00:35 -0500
Subject: Error evaluating numbers starting in zero
Message-ID: <3A3DEE53.621D0D94@mitre.org>

>> I have a python checkbook program that evaluates date strings, such
>> as "01/31/99". Most dates are fine, but if I have a date such as
>> "01/08/99", python runs into an error when it tries to eval("08").

> Use int() instead of eval().


So, that may also explain the failure with 

	i = 08

and with reading in numbers,
the python interpretor must be using EVAL instead of INT?




From red5sf at my-deja.com  Sat Dec  9 18:36:17 2000
From: red5sf at my-deja.com (red5sf at my-deja.com)
Date: Sat, 09 Dec 2000 23:36:17 GMT
Subject: editing python path in pythonwin
Message-ID: <90ufli$8f3$1@nnrp1.deja.com>

for some reason, my edits to the pythonpath using the pythonwin
(release 2.0) "EditPythonPath" feature aren't being made (on both win98
and 2000).

Is this a known issue?  The only workaround i've been able to figure is
creating a PYTHONPATH environment variable in my autoexec.bat(on 98) or
in my user defined varable list (on 2000)

do any of you know how to fix pythonwin to be able to set this?

thanks,
christian


Sent via Deja.com http://www.deja.com/
Before you buy.


From shindich at my-deja.com  Tue Dec  5 02:01:23 2000
From: shindich at my-deja.com (shindich at my-deja.com)
Date: Tue, 05 Dec 2000 07:01:23 GMT
Subject: ANN.: Weak Reference Extension Module has been converted to "C"; Beta 2 is out
Message-ID: <90i3s2$fkv$1@nnrp1.deja.com>

I am happy to announce that Weak Reference Extension project has been
fully converted to "C". Weak Reference extension is
 now completely CPython compatible.
 Check out Beta 2 at Source Forge
 (http://pyweakreference.sourceforge.net/ ;
 http://sourceforge.net/projects/pyweakreference/)

 Enjoy!

 Alex Shindich
 Visit http://www.shindich.com/



Sent via Deja.com http://www.deja.com/
Before you buy.


From jkndeja at my-deja.com  Wed Dec  6 10:29:19 2000
From: jkndeja at my-deja.com (jkndeja at my-deja.com)
Date: Wed, 06 Dec 2000 15:29:19 GMT
Subject: parallel port access?
References: <3A268378.4DA1B831@sympatico.ca> <3A272881.6F444750@engcorp.com>
Message-ID: <90lm04$b3r$1@nnrp1.deja.com>

Hello there
    I have only been glancing at c.l.python occasionally, and haven't
seen any of the previous discussions on this topic. I have also been
thinking about accessing the printer port (and I/O ports in general).
Alhough none of this is tested (or even tried out in reality!), the
original poster might find it useful. Or not. I solicit further
details...

I'm assuming that you are using some flavour of Windows. In order to
access the LPT port _directly_, you want to be able to do a form of I/O
port access. This is a rough equivalent to the BASIC code

PEEK(inport)
and
POKE(outport, value)

in 80X86 assembly language these take the form

in reg, (port)   ; and
out (port), val

It is not advisable to be able to do this on a platform such as the PC,
since you can crash your machine etc. very easily like this. Depending
on the Operating System you are running, you might not even be allowed
to do this even in Assembly language - it involves a change in
operating system privileges which you may not be allowed to do.

The way that you _are_ allowed to write directly to ports is via some
form of 'driver'. Again, depending of the OS you are using, this might
be a VxD, or a DLL. A Search on Altavista etc. for the 'Parallel Port
FAQ' will give you lots of information on this.

FWIW, my intention is to use python to access a simple ISA I/O card
containing 2x 8255 PIOs and a 8254 counter/timer. I'm probably going to
write, or modify, a VxD written by Karen(?) Hazzah, and downloadable
from the Windows Developer's Journal, which is designed to 'grant'
access to the ports in a range (eg the printer port, exactly what the
original poster wants). I'll then use either Mark Hammond's Win32
extensions, or Sam Rushing's 'CallDLL' methods, to access the VxD from
Python via the CreateFile(...) etc. functions.

This might be rather more detail than you require, and I may be making
things too complicated, but at least I've exposed my plans to the
world...


    Regards
    Jon Nicoll


In article <3A272881.6F444750 at engcorp.com>,
  Peter Hansen <peter at engcorp.com> wrote:
> DOC wrote:
> >
> > I would like to be able to control the individual
> > lines of the parallel port using Python.
> >
> > Anyone know of any Class packages to do this or
> > do I need to write one myself?
>
> What platform?
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From kno at jtan.com  Mon Dec 25 15:17:35 2000
From: kno at jtan.com (Matt Dunford)
Date: Mon, 25 Dec 2000 20:17:35 GMT
Subject: ready for C or C++?
References: <20001225031852.04296.00003766@ng-fp1.aol.com>
Message-ID: <G654DB.BGx.0.sinope@news.jtan.com>

xmnemonic at aol.com (Xmnemonic) writes:

>If I've learned Python and become proficient with it, am I ready to learn C or
>C++?


Yes, you are.  You'll really (and I mean really) appreciate Python
after working with C for a whie.  Once you learn the basics, I'd suggest
trying your hand at writing a Python extension in C.  That way you an
practice your C without ever leaving the comfortable world of Python.
--


From peter at engcorp.com  Sun Dec 10 12:58:01 2000
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 10 Dec 2000 12:58:01 -0500
Subject: How can I use printer port in python?...
References: <90vnol$l5p$1@news1-2.kornet.net>
Message-ID: <3A33C429.FE4EF79C@engcorp.com>

"??????" wrote:
> 
> As the title says, I was wondering if there's anyway I can
> 
> use printer port in python...

You don't mention which platform nor what you want to do with the
"printer port", nor even whether the "printer port" you refer to is a
parallel port, serial port, or some other port.  

We could assume you mean a parallel port, and that you are on Windows,
and that you don't actually plan to talk to a printer here, but then
again we could be wrong.  Help us out!

On top of which, may I suggest you check the archives for the last week
and read the discussion which just took place with respect to (what I
assume is) exactly this topic, under the subject "parallel port access?"


From tim.hochberg at ieee.org  Mon Dec 11 11:21:22 2000
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Mon, 11 Dec 2000 16:21:22 GMT
Subject: A Newbie Question: How To Call Other Python Programs From A Main Python Program.
References: <912tpb$efl$1@nnrp1.deja.com>
Message-ID: <6a7Z5.65025$15.13712960@news1.rdc1.az.home.com>

<jbranthoover at my-deja.com> wrote:
> Hello All,
> I have written several small Python programs.  These programs
> are not just functions but complete programs with their own menuing
> systems.
>
>         My question is,  how do I (and is it a good idea) call these
> independent Python programs from one main program that selects which
> program to run from a main menu?  I tried to use the import command.
> This seems to work OK except that the programs only run once.  If the
> same program is selected from the menu,  nothing happens.  It seems to
> return immediately without running the program.

I suspect that your subprograms are organized like:

# subprogram_1
#....
fire_up_menus_and_get_going()

Modules that you import only get executed on the first import and are then
cached for speed reasons. This is why you see the behaviour you do. You
could get around this by playing tricks with reload, but that would not be
"the right thing to do"(TM). Instead, reorganize your subprograms as
follows:

# subprogram_1
#....
def run():
    fire_up_menus_and_get_going()
if __name__ == "__main__"
    run()

And then access them from your main program using:

import subprogram_1
subprogram_1.run()

This should run your program afresh everytime it is called. The 'if __name__
== "__main__"' line runs the subprogram only if invoked from the command
line.

-tim







From jcm at bigskytel.com  Tue Dec 26 03:09:26 2000
From: jcm at bigskytel.com (David Porter)
Date: Tue, 26 Dec 2000 01:09:26 -0700
Subject: ~Python ?
In-Reply-To: <3a4833f8.346848@news.t-online.de>; from gerson.kurz@t-online.de on Tue, Dec 26, 2000 at 06:03:40AM +0000
References: <3a4833f8.346848@news.t-online.de>
Message-ID: <20001226010926.A29800@bigskytel.com>

* Gerson Kurz <gerson.kurz at t-online.de>:

> On Linux, ~ usually refers to the users homedir, but not so in Python
> (at least, in ActivePython 2.0). I found out the hard way when using a
> script that issued os.makedirs() relative to ~ - It created an actual
> directory named ~. You can have real fun if you try to deleting an
> actual directory named ~ on Linux, boy. 

It's easy. Just escape it with a backslash: 'rmdir \~'.

> I understand that ~ is probably handled by the shell, but if Python
> doesn't deal with it, its hard to use Python in some kinds of scripts.

os.path.expanduser() will interpret a tilde as the user's home directory:

>>> import os.path
>>> print os.path.expanduser('~')
'/home/dap'

you can also specify the user:

>>> print os.path.expanduser('~root')
'/root'


David



From cpsoct at my-deja.com  Sun Dec 31 09:01:19 2000
From: cpsoct at my-deja.com (cpsoct at my-deja.com)
Date: Sun, 31 Dec 2000 14:01:19 GMT
Subject: how to flatten & lump lists
Message-ID: <92ne7e$5fj$1@nnrp1.deja.com>

I want to be able to flatten a list and i could have sworn that i saw
that there was something like flatten() built in someplace but i can't
find it now anywhere in the docs.

if i have the following code:

# mirror.py -- [1,2,3] returns [1,2,3,2,1]
# how to make it return [1,2,3,2,1,2,3,2,1]
# this should really be a tuple so that it is preserved.

l=[1, 2, 3, 4, 5, 6, 7]
foo=l[:-1]      #copy list, excluding last element
foo.reverse()   #flip it
l.append(foo)   #stick it on the end of the other list
#we need to flatten it some how in to a single list
print l


the flipped part (foo) tacked on to the end of the list as a single
element instead of a single long list.

[1, 2, 3, 4, 5, 6, 7, [6, 5, 4, 3, 2, 1]]

how can i get: [1, 2, 3, 4, 5, 6, 7, 6, 5, 4, 3, 2, 1]?

additionally i would like to do the opposite, take a list like
[1, 2, 3, 4, 5, 6, 7, 8, 9, 10]

and have it lump up into subgroups (randomly) and return:
[[1],[2],[3, 4, 5], [6, 7], [8, 9, 10]]

Any help would be appreciated. I have lots of list processing ideas that
i want to impliment, but i need to know this first. Additionally, if
anyone knows of such list processing and pattern making funcs, please
let me know. i found a few on "vaults" but not of the type i was hoping
for.


happy new year,
kevin parks
seoul, korea
cpsoct at lycos.com



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


From SBrunning at trisystems.co.uk  Thu Dec 14 04:00:41 2000
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Thu, 14 Dec 2000 09:00:41 -0000
Subject: A question.
Message-ID: <31575A892FF6D1118F5800600846864D5B15EC@intrepid>

> From:	Alex Martelli [SMTP:aleaxit at yahoo.com]
> Glad I landed a nicely-sized fish with this one.  I was
> rather betting on 'shew', though -- doesn't it just _look_
> like a typo?-)
 
Not if you've read Shakespeare, it doesn't...

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From erno-news at erno.iki.fi  Sat Dec 16 11:53:15 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 16 Dec 2000 18:53:15 +0200
Subject: Is Stackless and uthread.py SMP friendly?
References: <91br9v$qa4$1@nnrp1.deja.com> <ku1yva5ocq.fsf@lasipalatsi.fi> <91dp4j$5fj$1@news.gte.com> <91dt39$fbk$1@nnrp1.deja.com>
Message-ID: <kusnno48as.fsf@lasipalatsi.fi>

In article <91dt39$fbk$1 at nnrp1.deja.com>, jbalding at hotmail.com writes:

| Okay. My next question: Is there a simple way of taking advantage of
| multiple processors using Python, or is this something that is going to
| be like pulling teeth? In other words, are there existing packages that
| allow relatively high-level, transparent access to multiple processors?

using several python processes is pretty much here. for communication
some of the rpc modules might be relevant.

  -- erno


From support at internetdiscovery.com  Sat Dec  9 18:40:13 2000
From: support at internetdiscovery.com (Mike Clarkson)
Date: Sat, 09 Dec 2000 23:40:13 GMT
Subject: WWW_URLEcho with dde
Message-ID: <3a32b7e6.8685100@24.0.228.33>

Has anyone used the dde module in Python to register a URLEcho from
netscape or Internet Explorer? Looking at the source code to dde I see
comments about Netscape's use of poke, which indicates the author
may have been playing with URLEcho.

I'd like to start a dde server and register it with netscape to send
me an echo when it loads a web page. Has anyone done this?

Many thanks in advance,

Mike.


From mwh21 at cam.ac.uk  Thu Dec  7 05:02:54 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 07 Dec 2000 10:02:54 +0000
Subject: can i make sys.stdin.read() not wait?
References: <90ncct$p58$1@nnrp1.deja.com>
Message-ID: <m3ofyo4khd.fsf@atrus.jesus.cam.ac.uk>

thelocust at my-deja.com writes:

> I'm working on a program that requires me to read in input (usually one
> character) if it is input.  For example, i have a while loop that looks
> like this:
> 
> while 1:
>       x = sys.stdin.read(1)
>       if x == 'q':
>            break
>       test.handlestuff()
> 
> Well, my problem is is that it seems that read MUST have input to move
> on.  Is there any way that it just attempts to get whatever is on the
> TTY?  or maybe the last thing that was put to it?

You can use a mixture of select.select, termios.tcsetattr and os.read
to do what you want; something like this:

import termios
import TERMIOS
import os
import select,sys

def read1():
    oldattr = termios.tcgetattr(0)
    try:
        attr = termios.tcgetattr(0)
        attr[2] = (attr[2] & ~(TERMIOS.NLDLY)) | TERMIOS.NL0
        attr[3] = attr[3] & ~(TERMIOS.ICANON|TERMIOS.ECHO)
        termios.tcsetattr(0,TERMIOS.TCSANOW,attr)
        while 1:
            r,_,_ = select.select([0],[],[],0)
            if r:
                return os.read(0,1)
            else:
                return ''
    finally:
        termios.tcsetattr(0,TERMIOS.TCSANOW,oldattr)

season-to-taste-ly y'rs
M.

-- 
  please realize that the Common  Lisp community is more than 40 
  years old.  collectively, the community has already been where 
  every clueless newbie  will be going for the next three years.  
  so relax, please.                     -- Erik Naggum, comp.lang.lisp


From MarkH at ActiveState.com  Tue Dec 19 20:11:29 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Wed, 20 Dec 2000 01:11:29 GMT
Subject: Events in win32com
References: <Y1g=Op6IfWSoJn1ISFtOyuvqwBb3@4ax.com> <FY0=OlZFLy71DtnK62u5TEPlhOts@4ax.com>
Message-ID: <3A40060B.10504@ActiveState.com>

Paul Moore wrote:

> After scanning Deja, I found an article describing DispatchWithEvents.
> The example in the documentation (OnVisible) works fine. But the
> OnQuit event doesn't seem to work :-( Basically, the event routine is
> never fired. Is this because the event routine has a ByRef "Cancel"
> parameter? Or is it something else?

OnQuit works for me.  test\testMSOfficeEvents.py traps this.

Mark.



From robin at alldunn.com  Thu Dec 14 23:40:00 2000
From: robin at alldunn.com (Robin Dunn)
Date: Thu, 14 Dec 2000 20:40:00 -0800
Subject: ANNOUNCE:  New Berkeley DB 3.1.x wrappers
Message-ID: <008001c06651$1638f6f0$0b01a8c0@ARES>

I'm pleased to announce a beta release of yet another BerkeleyDB wrapper
module for Python.  This one is different from the ones that Gregory P.
Smith and I have done in the past in that it doesn't use SWIG.  This one is
completly hand-crafted, and there are no "shadow classes" or other SWIG
related overhead.

This module provides a nearly complete wrapping of the Sleepycat C API for
the DatabaseEnvironment, Database, Cursor, and Transaction objects, and each
of these is exposed as a Python Type in the bsddb3.db module.  The databse
objects can use different access methods, btree, hash, recno, and queue.
For the first time all of
these are fully supported in the Python wrappers.  Please see the documents
in the docs directory of the distribution, or on the website, for more
details on the types and methods provided.

http://pybsddb.sourceforge.net/

--
Robin Dunn
Software Craftsman
robin at AllDunn.com
http://wxPython.org     Java give you jitters?
http://wxPROs.com        Relax with wxPython!





From DavidL at NOraqiaSPAMMY.com  Tue Dec 19 13:39:54 2000
From: DavidL at NOraqiaSPAMMY.com (David Lees)
Date: Tue, 19 Dec 2000 13:39:54 -0500
Subject: FIFO in python?
Message-ID: <3A3FAB7A.DC63610C@NOraqiaSPAMMY.com>

I am trying to learn how to use FIFO and am unable to open a fifo under
Linux RedHat 6.2.  I hang on the open (open(fn,'rw'), even though I have
done a successful check for file existence (os.path.exists(fn)).  The
same code works under NT.  Any thoughts on what is going on?

Thanks in advance,

david lees


here is my code:
-------------------------
import string, os, sys

def OpenFileRW(n):
    if os.path.exists(n):
        fn=open(n,'rw')
        return fn
    else:
        print 'Filename '+name+' does not exist'
        sys.exit(-1)



fifoName='junkfifo'
fifo=OpenFileRW(fifoName)
fifo.close()
print 'done'


From aleaxit at yahoo.com  Wed Dec 13 05:24:43 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 13 Dec 2000 11:24:43 +0100
Subject: Python-like compiled language
References: <913jr3$dmv$1@sylvester.vcn.bc.ca> <3A35835B.AAA92005@bioeng.ucsd.edu> <kuzoi178kg.fsf@lasipalatsi.fi> <9163mr02v2t@news1.newsguy.com> <kulmtl6xqx.fsf@lasipalatsi.fi>
Message-ID: <917ipt023tk@news2.newsguy.com>

"Erno Kuusela" <erno-news at erno.iki.fi> wrote in message
news:kulmtl6xqx.fsf at lasipalatsi.fi...
    [snip]
> some say that the latest virtual machines should be quite fast, but

I've heard that repeated continuously over the last 5+ years.  It
will no doubt become true one day, "in the long run", but you know
what Keynes had to say about that.


> [good scheme suggestion snipped]
>
> it is also rumored that ocaml has a good optimizing compiler.

It has an _excellent_ compiler, but it has no similarity to
Python whatsoever.  The totally static typesystem is only the
start of the differences; such issues as needing to use `+.`
when you sum floats (you can only use `+` to sum _integers_)
take it away from "the average programming language" in the
exactly _opposite_ direction.  My point re Scheme/stalin was
predicated on at least 'some' semantic similarities between
Scheme and Python.

What about Modula-3, by the way?  I gather that one has
similarities, too (having provided some of our BDFL's
inspiration), but don't know much about its compilers'
effectiveness.


Alex





From urner at alumni.princeton.edu  Mon Dec 11 12:13:45 2000
From: urner at alumni.princeton.edu (Kirby Urner)
Date: Mon, 11 Dec 2000 09:13:45 -0800
Subject: important programming languages
References: <912jps$5vr$1@nnrp1.deja.com>
Message-ID: <kh0a3tc6lk2v2l1a4a38fep3047k4amrjq@4ax.com>

Tuomas Pellonpera <tuope at my-deja.com> wrote:

>4.) Can Python be used in encryption? If so, IS it being used, too? :)

I've read the comments to date and found much to agree with.

Python is extended by writing methods in C in a particular way, 
and then importing the API -- meaning a lot of what you do in 
Python ends up running as native C, and so is relatively fast.

So as a Python programmer, you can piggyback on what's already 
been written in C to extend Python.  In this connection, you 
already have access to A.M. Kuchling's crypto library:
http://weathership.homeport.org/~adam/crypto/python.phtml

When it comes to learning programming, I agree that Python 
provides a clean, well-designed syntax.  So even though 
pure Python source code would be too slow for most practical
crypto applications, you might usefully learn the guts of 
an algorithm that way -- knowledge you could then transfer 
to another language (even Assembler).  

In this sense, training in Python helps you think the way a 
programmer thinks.  Then you can branch into C/C++ or Java 
or whatever (or pick them up simultaneously, if you're impatient 
to get on with it).

Case in point:  I've just recently implemented the Blowfish 
algorithm in pure Python, in the context of a larger math-
oriented curriculum which uses Python as the core teaching
language.  Yes, it's slow, and minimal (no CBC, no key-from-
passphrase generator) but it works (runs through the test 
vectors properly, encrypts/decrypts binary data files, 
like GIFs).  

I think someone new to crypto, and to programming in general, 
could pick up some valuable experience studying such material
(which is why I write it).  Plus some of the math stuff I do 
on other pages (e.g. using Python's big number capabilities) 
provides useful background (as an earlier poster said, you 
need some theory, not just programming language smarts, when
it comes to crypto).

Here's the URL (posted about it a couple days ago here on 
sci.crypt as well, inviting comments): 

  http://www.inetarena.com/~pdx4d/ocn/clubhouse.html

<SOAP BOX>

I'm a big fan of the computer-at-your-elbow approach to 
studying math concepts, and am frustrated that classroom 
teachers seem stuck on an evolutionary plateau with graphing 
calculators.  Calculators are cool, but a lot of school 
districts have pumped big bucks into computers, and now 
feel ripped off because Johnny only uses them to run CDROM
page-turner software or play games, and Sally just wants 
to download MP3s.  

But that's a failure of the curriculum, not the technology.  
Kids _could_ be learning real programming languages, and 
math class (not necessarily some separate computer science 
elective, which many schools don't offer) would be the 
logical place to get this going.  

What we're talking about here is basic numeracy, and in 
this day and age, that includes having knowledge of computer 
concepts as well as math concepts (a lot of the walls 
between math and computer science are very artificial to
begin with, especially at the elementary level, where 
overspecialization is only destructive -- is enforced 
mostly to protect a woefully obsolete status quo curriculum).

Teachers mostly won't support this "math through programming"
approach unless the text books publishers do, but the text 
book upgrade cycle is way way too slow.  Enter the web, and 
the option to drive the upgrade cycle on "internet time".

More on all this:
http://www.oreillynet.com/pub/a/python/2000/10/04/pythonnews.html

</SOAP BOX>

Kirby




From fredrik at effbot.org  Fri Dec  8 05:25:11 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 08 Dec 2000 10:25:11 GMT
Subject: problems with urllib in Python20?
References: <90pdir$k7$1@bob.news.rcn.net>
Message-ID: <bG2Y5.2165$Kd1.313362@newsb.telia.net>

Z 3 Penguin wrote:
> whenever i try to use urllib.urlopen() in Python20/Win32, i get this error:
> -START-
> Fatal Python error: Interpreter not initialized (version mismatch?)
>
> abnormal program termination
> -END-

your interpreter is picking up an old (1.5.2?)  DLL.  to figure
out which one, and where it is, run Python with the -v option.

fixing the PYTHONPATH should take care of this.

</F>




From peter at engcorp.com  Mon Dec 11 08:29:00 2000
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 11 Dec 2000 08:29:00 -0500
Subject: Backslash escape in regular expressions
References: <jon.976507791@bezek.dstc.monash.edu.au>
Message-ID: <3A34D69C.BAC68E27@engcorp.com>

Jonathan Giddy wrote:
> 
> According to the re module documentation, backslash either escapes special
> characters, or signals a special sequence.  The special sequences are
> then listed.
> 
> However, as this code shows, there are some special sequences (mainly
> the whitespace characters) that are special, but aren't listed.  Is this
> a lapse in the re implementation or the re documentation?  Can I safely
> expect re.compile(r'\(hello\)\n') to always match '(hello)\n' (the current
> behaviour) and not match '(hello)n' (the documented behaviour?)

The documentation I have clearly shows that \\ is a special sequence
which turns into the backslash character itself.  With the raw-string
form with 'r' your "current behaviour" above *is* the documented
behaviour, isn't it?  At least, just using those strings with "print"
shows that you don't get "(hello)n"...


From noahspurrier at my-deja.com  Thu Dec 21 22:54:50 2000
From: noahspurrier at my-deja.com (noahspurrier at my-deja.com)
Date: Fri, 22 Dec 2000 03:54:50 GMT
Subject: Wanted: Python White Papers
References: <91oj8v$i0r$1@nnrp1.deja.com> <91tmkq$lru$1@nnrp1.deja.com>
Message-ID: <91uja8$dhh$1@nnrp1.deja.com>

Thanks! This is great. It makes good ammo
for my defense of Python.

Does infoseek still use it... or did Go
pretty much obsolete everything infoseek had?

Yours,
Noah

In article <91tmkq$lru$1 at nnrp1.deja.com>,
  jerry_spicklemire at my-deja.com wrote:
>   Noah Spurrier wrote:
> > Is anyone aware of any white papers to explain the use of Python
> > at any big, high-profile companies?
> > How does Python fit into their design?
>
> Actually InfoSeek used Python for years, before Google existed, and
> some other high profile users include -
>
> HP, especially in the recent E-Speak project :
... a lot of stuff ...
>  http://www-unix.mcs.anl.gov/~curfman/cca/web/cca_paper.html
>
> BTW, Metacrawler is handy for this sort of thing!
>
> Later,
> Jerry S.
>
> Sent via Deja.com
> http://www.deja.com/
>


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


From News at Titanic.co.uk  Wed Dec 27 09:33:36 2000
From: News at Titanic.co.uk (Simon Faulkner)
Date: Wed, 27 Dec 2000 14:33:36 GMT
Subject: Controlling the Parallel Port
Message-ID: <mvuj4tkmn22gbth1a6t8fag39902qeeu9l@4ax.com>

Is it easy to control (read/write) the Parallel Port in linux using
Python?

I am hoping to use it to drive some simple electronics

(Usual disclaimer understood - If it smokes then you probably got the
wiring wrong!)


Simon
Simon Faulkner


From see at below  Tue Dec  5 20:30:35 2000
From: see at below (Paul Foley)
Date: 06 Dec 2000 14:30:35 +1300
Subject: "inherited" keyword in Python?
References: <3A2B9EFA.3A64C13B@schlund.de> <90gevs02uu1@news2.newsguy.com> <m2elznipbl.fsf@mycroft.actrix.gen.nz> <90ievt025bl@news2.newsguy.com> <m266kzhua6.fsf@mycroft.actrix.gen.nz> <90jlrk012u7@news1.newsguy.com>
Message-ID: <m23dg2ibes.fsf@mycroft.actrix.gen.nz>

On Tue, 5 Dec 2000 22:11:08 +0100, Alex Martelli wrote:

> To clarify: Python's rule (as applied by, e.g., getattr) is anything
> but broken, but it doesn't easily let you return (for a given class X

Python's rule works for Python, sure -- but Python doesn't provide any
way to call a superclass method without knowing exactly what you want
to call (else all this would be unnecessary!)

> that inherits from A, B, C, ...) a single class Z that will ensure that
> getattr(Z,whatever) is the same as getattr(X,whatever) for any
> whatever that is not a key in X.__dict__.

I'm not sure that's all that's needed, though.

>> that won't work.  I thought super() was just supposed to call the
>> appropriate same-named method of a parent class.  Anything else can be
>> done through the usual self.whatever, surely?

> Not really.  To translate literally a Java class which has methods X
> (which calls the base's Y) and Y (which calls the base's X method),
> I'd need a Python's super() to be like Java's super (returning a class

I've never bothered with Java.  How can you write a method X which
calls the base's Y without explicitly calling the base's Y?  And if
you have to do it explicitly, that'll work in Python, too.

> object reference) and not like what I name 'supercall'.  self.X and
> self.Y just won't do for this.

It seems to me, what you said above about Java notwithstanding, that
the only time it's OK to call X where X is a method in a base class,
and where self.__class__ also has a method named X (so that self.X()
will do the wrong thing), is from inside self.X.

>> > But, how often is it going to be satisfactory to call 'just
>> > the one' (of several) inherited-method of a certain name, in
>> > a multiple-inheritance context?  Just as often we may want
>> 
>> It isn't (at least, not to me), which is why I avoided doing it that

> But you did do that -- you call just one of the several 'inherited'
> implementations of the method, in the code below.

Any one call to super() only calls one method, naturally.  But if
everyone calls super() _all_ the methods eventually get called.  Under
"Python rules", only the methods in D, B and A can ever run -- the one
in C never does, unless called explicitly with C.foo(self) from D; and
in that case, it's called out of order -- you either get D, B, A, C or
D, C, B, A, or, more likely D, B, A, C, A or D, C, A, B, A, with A.foo
getting called twice, since both B and C inherit A.  This sucks.  I
get D, B, C, A, which is the right order (B.foo and C.foo are both
called before their common superclass's A.foo, B.foo is called before
C.foo since B is to the left of C in D.__bases__) and A.foo is only
called once.

>> way.  Having figured out how to get the frame info, here's an
>> implementation.  No need to pass in the object and method name:

> I'd still want to pass them, at least optionally, to handle the
> above case where I want to call a _different_ inherited method.

I'm working on the assumption that if you want to do that, you're
doing something wrong.

>> If you just call super(__thisclass) with no args other than the class,
>> it passes on the args the calling method got; or you can specify other
>> args explicitly [yes, I know this prevents you calling with no args if
>> super's caller has args]

> As, once again, convenience gets in the way of simplicity and

Maybe.  This is an attempt to copy the CALL-NEXT-METHOD functionality
of CLOS; in CLOS, all the methods will have similar argument lists,
because they're all part of one generic function, not independent
members of various classes.  Just chop out the 5 lines in the middle
that copy the caller's arglist if you think it's better not to do
this.

> generality.  Oh well:-).  "Tantum convenientia potuit suadere
> malorum", as Lucretius might write if he lived today!-)

[Am I supposed to be one of them?  And shouldn't "malorum" be dative?]

-- 
Cogito ergo I'm right and you're wrong.                 -- Blair Houghton

(setq reply-to
  (concatenate 'string "Paul Foley " "<mycroft" '(#\@) "actrix.gen.nz>"))


From gtalvola at my-deja.com  Sat Dec 16 14:43:42 2000
From: gtalvola at my-deja.com (gtalvola at my-deja.com)
Date: Sat, 16 Dec 2000 19:43:42 GMT
Subject: Source Code Management
References: <fb4f3tsgffnmvme64aol1cn397te7ln349@4ax.com> <91aq96$tba$1@nnrp1.deja.com> <3A3AA3B7.3AC3F4BF@uswest.net>
Message-ID: <91ggld$ble$1@nnrp1.deja.com>

In article <3A3AA3B7.3AC3F4BF at uswest.net>,
  Bjorn Pettersen <pbjorn at uswest.net> wrote:
> Do you have a link to how to get the result in Python
marshalled dictionary
> format? (I looked at the perforce website but couldn't
find anything...)
>

Here's a snippet from the Perforce mailing list about the feature
(I don't have the original author's name):

> You can use the -G option to Perforce to cause Perforce to do input
> and output as Python marshalled dictionaries.  (This is a new but
> unsupported feature in 2000.1 -- see "p4 help undoc".)
>
> Using this feature it's easy to query Perforce and read the results.
>
>      import marshal
>      import win32pipe
>      stream = win32pipe.popen('p4 -G changes', 'r')
>      changes = []
>      try:
>          while 1:
>              changes.append(marshal.load(stream))
>      except EOFError:
>          for c in changes:
>              print c['change'], c['desc']
>
>


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


From aahz at panix.com  Mon Dec 25 15:12:10 2000
From: aahz at panix.com (Aahz Maruch)
Date: 25 Dec 2000 12:12:10 -0800
Subject: One Python 2.1 idea
References: <mailman.977771949.17911.python-list@python.org>
Message-ID: <9289mq$5vq$1@panix6.panix.com>

In article <mailman.977771949.17911.python-list at python.org>,
Tim Peters <tim.one at home.com> wrote:
>
>my-python-code-runs-5x-faster-this-month-thanks-to-dumping-$2K-
>    on-a-new-machine-ly y'rs  - tim

Those of y'all who save Python quotes should save this one.

(I'm highlighting it just in case there's someone stupid out there who
doesn't read every Tim Peters post top-to-bottom.)
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

'"Crisp" is a good quality for crackers; less so for pot roast.'  --pnh


From fredrik at effbot.org  Tue Dec 19 14:55:46 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 19 Dec 2000 19:55:46 GMT
Subject: Converting a list to a string
References: <sRO%5.168$LU6.63588@typhoon.aracnet.com>
Message-ID: <63P%5.516$AH6.44957@newsc.telia.net>

Daniel Klein wrote:
> My application needs to be able to convert a list to a string and vice
> versa. Taking the first problem, the best I could come up with is

I assume "list" means list of strings:

    s = string.join(mylist, "")

(or in 1.6 and later, s = "".join(mylist))

not sure what "vice versa" means, but maybe string.split
(or the split method) is what you need?

check the string module docs for details (it's in the library
reference).

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From kevrjohn at my-deja.com  Sun Dec 24 09:15:50 2000
From: kevrjohn at my-deja.com (kevrjohn at my-deja.com)
Date: Sun, 24 Dec 2000 14:15:50 GMT
Subject: registry tweaking program
Message-ID: <9250ek$ueo$1@nnrp1.deja.com>

   I would like to write a program that will automatically modify
certain parts of the registry in win98.  I have never programed before
(but I am trying to learn) and I thought that this would be a
good "real app" for me because I modify the registry a lot, and I would
like to see if I could automate the process to save some time.  I
really only want it to go and do a few preprogrammed things.  Nothing
spectacular like regedit.exe.  So how do you get a python script to
access and/or modify the registry?  I have looked around and I haven't
found any HOW-TO on the subject.
Thanks in advance,
Kevin Johnson


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


From pbjorn at uswest.net  Fri Dec 22 13:38:29 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Fri, 22 Dec 2000 11:38:29 -0700
Subject: How use urllib.urlopen()?
References: <9012313F4gustaflalgonetse@194.213.69.148>
Message-ID: <3A439FA5.FD2CC51D@uswest.net>

Try:

#! /usr/local/bin/python
print "Content-type: text/html"
print

then your program. If that doesn't work, try:

 - make sure the script is executable
 - change the extension of your script to .cgi

If none of that works try the following:
 - put the following in tt.cgi
 - make tt.cgi executable
 - create yyy.txt and make it world writable (important)
 - run it from your web-browser
 - look at what went wrong (it should be in yyy.txt).

#!/bin/sh
echo "Content-type: text/html"
echo
/home/bjorn/public_html/xxx.py >> yyy.txt 2>&1
#echo "<h1>hello world</h1>"

Hope this gets you on your way.

-- bjorn

Gustaf Liljegren wrote:

> Can't get the urlopen() function to work. I have a script on a UNIX web
> server that should go to another webpage when the user clicks a button. To
> isolate the problem, I made a test script:
>
> #!/usr/local/bin/python
>
> import urllib
> u = urllib.urlopen("http://www.yahoo.com")
> print u.read()
>
> When running this I get HTTP 500 - Internal server error. :-(
>
> What's missing?
>
> Gustaf



From patrick.fichou at galactica.it  Fri Dec  8 06:02:59 2000
From: patrick.fichou at galactica.it (Patrick)
Date: Fri, 08 Dec 2000 11:02:59 GMT
Subject: Newbie : Tk an data input
Message-ID: <Dd3Y5.107318$DG3.2083614@news2.giganews.com>

Hi,

Is someone kind enough to send me an example of a program making a data
input into a file (an MSAccess Data Base would be great). This is what I
would like to do on may Windows system :

* Have a Tk form with menus... (Add, Edit, Delete, Print...Exit)
* Let the user input a Name, a Social Security code (for example)
* Some other fields,
* Click the 'Ok' button,
* Write the DataBase,
* Clear the screen and be ready to gon on...

Thank you very much,
Patrick




From dale at out-think.NOSPAMco.uk  Wed Dec 13 12:57:52 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Wed, 13 Dec 2000 17:57:52 +0000
Subject: Problem with += (augmented assignment)
References: <9188of$q55$1@nnrp1.deja.com>
Message-ID: <t0ef3t073otc6o0ujo04mqelelk4r153np@4ax.com>

mjjacob at my-deja.com wrote:

>Hi folks!
>
>I tried this under Python 2.0 running on Win 2k:
>
>x = 0
>while x <= 42:
>	x += 1
>
>Python always complains a Syntax Error: Invalid
>Syntax on this:
>
>    x += 1
>       ^
>SyntaxError: invalid syntax
>
>Augmented assignment only seems to work when used
>in first indentation level (script level).
>Otherwise, I always get that syntax error, it
>does not matter whether I am using if, while or
>whatever.
>
>Is there something terribly wrong with the +=
>operator or is that all my fault?
>
>Help appreciated.
>
>
>Sent via Deja.com
>http://www.deja.com/

I copied your code and pasted it into a new Python script and it worked fine.

Python 2.0 on Win 2K, here too.

I have to assume that either you're not running Python 2 or your done something odd with your tabs.

--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants


From pbjorn at uswest.net  Sat Dec 30 13:37:32 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Sat, 30 Dec 2000 11:37:32 -0700
Subject: Q: how to sort a dictionary by the values ?
References: <Pine.GSO.4.10.10012292313350.21787-100000@ux11.cso.uiuc.edu>
Message-ID: <3A4E2B6B.AA0E215D@uswest.net>

items = [ (y,x) for x,y in mydict.items() ]
items.sort()

-- bjorn

hwan-jo yu wrote:

> Hi,
> If I sort the items() of dictionary, it is sorted by first the key and
> second the value.
> How can I sort it by the values ?
> Thanks in advance.



From sholden at holdenweb.com  Mon Dec 18 09:52:14 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 18 Dec 2000 09:52:14 -0500
Subject: Alarms in threads - how to get/mimic?
References: <Pine.LNX.4.21.0012161749380.7913-100000@traal.eecs.berkeley.edu> <ltd7erk8xd.fsf@asfast.com>
Message-ID: <tzp%5.39168$n51.703032@e420r-atl1.usenetserver.com>

Lloyd Zusman <ljz at asfast.com> wrote in message
news:ltd7erk8xd.fsf at asfast.com...
> Luca de Alfaro <dealfaro at eecs.berkeley.edu> writes:
>
> > [ ... ]
> >
> > What is not clear to me is the following.  When there is a timeout,
> > the exception Timeout is not caught by httplib.  Hence, if the thread
> > that called httplib just continues its job giving up on the
> > connection, httplib will not have a chance to do any clean-up, such as
> > for example releasing/closing the socket.
> >
> > Hence, it seems to me that in order to abort a web page load cleanly
> > upon timeout, I also have to modify the httplib module to handle the
> > exception Timeout, and do clean-up work if that happens.
> >
> > Is my understanding correct?
>
> Not entirely.  When using timeoutsocket, all you have to do is start
> catching timeout exceptions when you make the various method calls
> within httplib
[snip code]

Basically, if the code you make use of traps the errors raised by
timeoutsocket and re-raises them after it's done whatever cleanup it needs
to then there should be no problem.  However, there MAY be problems if the
timeoutsocket error is not caught by the higher-level protocol code, and
makes it to the outer code which called it.  In that case there is a
possibility that necessary clean-up may not be performed.
>
> It's recommended in Python that exception handling be done whenever
> it's appropriate; therefore, by wrapping your httplib calls in this
> same manner, you'll remain true to the Pythonic Way, and you won't
> have to re-write any httplib code when using timeoutsocket.
>
> Note that the timeoutsocket.setDefaultSocketTimeout() method call
> should be made before making any httplib calls.  If you want to use
> different timeout values when making different httplib calls (or other
> calls socket-based methods), you'll have to re-invoke the
> setDefaultSocketTimeout() method before each call that uses a changed
> timeout value.
>
Note also that you can set the timeout on each socket individually.
setDefaultSocketTimeout is usually used to ensure that connect() attempts in
upper-layer protocols (which happen before you have a socket object in your
application to manipulate) don't hang forever.

regards
 Steve








From aleaxit at yahoo.com  Mon Dec 25 03:52:34 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 25 Dec 2000 09:52:34 +0100
Subject: One Python 2.1 idea
References: <mailman.977590932.19335.python-list@python.org> <9237md$orm$1@nnrp1.deja.com> <924bjt0e3n@news1.newsguy.com> <925baq$5sj$1@nnrp1.deja.com> <925tak01tpu@news1.newsguy.com> <926ici$vi9$1@nnrp1.deja.com>
Message-ID: <92722s0l1f@news1.newsguy.com>

<rturpin at my-deja.com> wrote in message news:926ici$vi9$1 at nnrp1.deja.com...
    [snip]
> C++ does NOT provide the same level of abstraction as
> Python. It does not have sequences, dictionaries, nor
> memory management as built in parts of the language.

What difference does it make that std::vector, std::map,
and std::auto_ptr are in the Standard C++ Library rather
than 'built in parts of the language'?  While there _are_
valid points to what you say (memory mgmt just _can't_
be made quite as smooth in C++ as it's in Python), this
bit about 'sequences and dictionaries' makes little sense
to me (maybe it's the eggnog:-).

C++ is statically typed (like Haskell, ML dialects, Eiffel,
Java, Ada, Object Pascal, Modula-3, etc).  This is a
crucial difference (I believe latent typing gives me a
great boost in productivity) and Python (like Smalltalk,
Scheme, Erlang, etc) places itself in a different
category by choosing latent typing.  But I wouldn't
frame this as an issue of level-of-abstraction.

C++ offers lots of very-low-level constructs AS WELL
AS pretty high-level ones, and, as I said, the low-level
parts keep coming up then and again even when one is
trying to work more abstractly -- also, from the wide
range of abstractions supported comes in part the huge
complexity of C++, its worse practical defect (very few
can master it fully, and anyway, complexity breeds
errors).


But Python's dynamic, latent-typed character itself
militates against your preferences: that it should
expose just one sequence-abstraction, for example,
_without_ letting the programmer express what it
knows about how a certain sequence will get used.

RDBMS enjoy a crucial difference: the low-level
information on how best to optimize (e.g.) a join
is in full possession of the RDBMS; it does not so
much depend on what SQL queries I'm using now
or in the future, but on the physical arrangement of
tables, indices, &c, which the DB 'knows'.  Optimal
arrangements will depend on the set of queries that
are intended, and that's anything BUT a solved
problem -- have you ever seen, e.g., an Oracle DB
running before and after an expert human tuner got
to it, studying the tables and the queries and tweaking
things manually?  Orders of magnitude of difference!
But SQL itself is not where the tuning is expressed,
and definitely not the queries.

I like the general approach of Dylan: let me state
things as abstractly and dynamically as I like, to
get things started -- the compiler's ability to optimize
will of course be seriously hampered.  Then, one can
tune, adding constraints, expressing what the coder
*knows* and the compiler could at best *guess*.

I _don't_ like programs which try to read my mind
and "do what's best for me whether I want it or not".
The LACK of black-magic of this sort in Python is a
good part of its charm for me.  The highly distorted
coding one sees in high-level languages which LACK
such expressive ability, to snuggle into a sweetspot
where the compiler will match a pattern and make
nice code, is *far* to common... let me say what I
mean, and I undertake to mean what I say!

Which is why I'd like to see it move towards letting
me be just as explicit as I want to be, about the
'declarative constraints' my code is designed for,
and either insert code to check the constraints, or
use them as springboards for its optimizations.

Which, I think, is the crux of our disagreement,
and I think we can agree to disagree.


> want implementation directives, they should be
> syntactically distinct annotations, appended to the
> relevant code:
>
>     stopWords = {}     #PYDIR: nodeletes
>
> At least that way, programmers reading and debugging
> the code can (in theory) ignore the annotation. (The

Which is EXACTLY what I don't want!  IF I have a
semantic constraint about there being no deletes
on this object, it's a CRUCIAL part of my abstraction
for it -- I want to state it clearly and readably, and
let compiler and human readers make of my statement
what they wish.

> compiler can also ignore the directive, if it is wrong
> headed.) In my opinion, you get far more bang for the
> buck if the compiler makes these optimizations
> automatically.

The compiler may be able to notice that there are no
deletes NOW, but it's not going to broadcast this key
semantic fact to the readers of the program, and it's
going to be hard put to do so if the code can grow
dynamically at any time.  As the program designer, I
*know* whether I intend deletes to be 'forbidden' on
the object, or if it just happens accidentally that there
are none YET; I want to be able to state what I know
*in the language*, because it's important, first of all,
for people who READ my code and want to understand
what it's up to.  That such constraints -- as well as
helping readers -- can also be checked at runtime for
better debugging, OR used to effect optimization, is
just great... but in any case *they're part of my code's
semantics*, and I want to EXPRESS its semantics, NOT
have compilers or human readers have to guess it!-)


Alex

>
> (Apologetic prologue: This mind dump is occuring after
> Christmas eve celebrations. I hope everyone reads it
> generously, and that I am not too embarrassed when I
> read it tomorrow.)
>
> Russell
>
>
> Sent via Deja.com
> http://www.deja.com/




From wolfson at midway.uchicago.edu  Tue Dec 12 21:01:51 2000
From: wolfson at midway.uchicago.edu (Ben Wolfson)
Date: Wed, 13 Dec 2000 02:01:51 GMT
Subject: A question.
References: <9152a0$acb$1@panorama.wcss.wroc.pl> <915elr0681@news2.newsguy.com> <9166uv$8rs$1@animus.fel.iae.nl> <916cd20b3b@news1.newsguy.com>
Message-ID: <jMAZ5.170$x3.2004@uchinews>

In article <916cd20b3b at news1.newsguy.com>,
Alex Martelli <aleaxit at yahoo.com> wrote:

> [I don't wink, I don't adverbialize, I iconize
>only very occasionally, the polysyllabic eruditeness of my hallmark
                                          ^^^^^^^^^^^
>circumlocutions pushes my Gunning Fog AND my Flesch-Kincaid up
>in the high 20's -- what more could I do to shew hostility...?]

*Surely* you mean "erudition"!  Honestly, I just don't know to what the
world might be said to be coming when such a seemingly intelligent young man
as yourself can make a mistake, not only *while* intending to show off his
ability to use his prolixity to confounding (and astounding!) effect, but
make the damning mistake in the morphology of _the_ _very_ _word_ which best
encapsulates his fearsome mental prowess--all that without the merest hint
of irony (not that irony could truly save him, but it would have some
however slight redeeming effect)!

It's a hard life.



-- 
BTR
It infuriates me that Dracula may have needed my soul more than I do.
 -- Stanley Donwood, "Dracula"



From root at [127.0.0.1]  Mon Dec 25 20:00:56 2000
From: root at [127.0.0.1] (deadmeat)
Date: Tue, 26 Dec 2000 01:00:56 GMT
Subject: variable scope?
Message-ID: <c5S16.8616$KY1.23318@news1.rivrw1.nsw.optushome.com.au>

This doesn't work:

---
import time, string

month = "2"
year = "2000"

def foo():
   if string.atoi(month) <= 3:
      year = `string.atoi(year) - 1`
      print year

foo()
---

reporting:

Traceback (most recent call last):
  File "pfft.py", line 11, in ?
    foo()
  File "pfft.py", line 8, in foo
    year = `string.atoi(year) - 1`
UnboundLocalError: Local variable 'year' referenced before assignment

but this does

---
import time, string

month = "2"
year = "2000"

def foo():
   year = "2000"
   if string.atoi(month) <= 3:
      year = `string.atoi(year) - 1`
      print year

foo()
---

how do I force foo() to use the global 'year' variable?





From monamant at freemail.sx.cn  Wed Dec 20 12:34:39 2000
From: monamant at freemail.sx.cn (monamant at freemail.sx.cn)
Date: Wed, 20 Dec 2000 17:34:39 GMT
Subject: Please help
Message-ID: <91qqjf$au4$1@nnrp1.deja.com>

Hi,
I am just thinking about taking Python as my first programming language
for serious learning, but some distrusts with its applicability prevent
me from fully applying myself to it. Would you kindly help clear up
some of my doubts about Python?
1. I want to use IIS 4.0 for my webserver, can I use scripts compiled
by Python in it?
2. What about the ASP? If I want to use ASP, can I use Python to do
such jobs like user authentication? How well can Python be used in Web
programming on Windows platform?
3. I used VB in the past and you can make a '.exe' file after
compiling. But how can I actually execute the Python program in
Windows? It seems that I can not make a .exe file for it to execute. It
is dependent on the interpreter?
4. If I really sit my self to Python, does it reward equally for the
same effort in learning C? In the aspect of the quality of programs.
Thank you in advance for answering my questions.
                                   An eager learner waiting for reply

(By the way, if I begin my serious learning process, I will translate
the tutorial into Chinese.)


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


From andreas at andreas-jung.com  Mon Dec 18 03:29:29 2000
From: andreas at andreas-jung.com (Andreas Jung)
Date: Mon, 18 Dec 2000 09:29:29 +0100
Subject: www,deskcode.com hot new coding resource python
In-Reply-To: <91h4ca$qat$1@nnrp1.deja.com>; from deskcode@my-deja.com on Sun, Dec 17, 2000 at 01:20:13AM +0000
References: <91h4ca$qat$1@nnrp1.deja.com>
Message-ID: <20001218092929.A16339@yetix.sz-sb.de>

On Sun, Dec 17, 2000 at 01:20:13AM +0000, deskcode at my-deja.com wrote:
> 
> Thank You
> www.deskcode.com

Does not work with Netscape 4.7X. It refers to an URL like

  http://p.moreover.com/cgi-local/page?index_developer ..+js

that seems to be invalid (Squid outputs an error). But it works
fine with MSIE.

Andreas



From howard at eegsoftware.com  Wed Dec 20 19:15:30 2000
From: howard at eegsoftware.com (howard at eegsoftware.com)
Date: Thu, 21 Dec 2000 00:15:30 GMT
Subject: object deletion order question
References: <3a41016d.3397213@news-server.socal.rr.com> <j%706.700$AH6.70137@newsc.telia.net>
Message-ID: <3a41476d.21319691@news-server.socal.rr.com>

On Wed, 20 Dec 2000 19:45:19 GMT, "Fredrik Lundh" <fredrik at effbot.org>
wrote:

>howard at eegsoftware.com wrote:
>> However, that led my to question:
>>         What ordering does Python use for object deletions?
>
>any order it likes.  it may remove an object as soon as
>possible, or never at all.
I neglected to mention that this code worked fine in the Python 1.5.2
version I was using but failed in the 2.0 version.

I guess my question was at termination time.  I had assumed that,
subject to cyclical references, all objects would be 'deleted' upon
interpreter termination.  Otherwise, there may have been a user object
that did file manipulation on deletion/closing which would never be
performed.  Say it aint so.

>> When Python terminated (or at other deletions), it called the
>> delete method of the SoundBuffer  objects after deleting the
>> Sound objects, leaving DirectX calling into unknown data space
>> (I know..it's a Microsoft tradition).
>
>Direct X uses reference counting just like Python.  if you mess
>up the reference count in Python, it'll crash as well.
>
>avoiding this is easy, of course.  just make sure you keep the
>reference counts in sync:
>
>-- use Python wrappers for all reference-counted Direct X
>   objects (e.g. SoundObject, SoundBufferObject)
>
>-- only access Direct X objects via the Python wrappers
>   (e.g. through an LPDIRECTSOUND member in the Sound
>   object class)

Actually, I thought I DID do most of that.  The CDX interface is built
with SWIG.  Cursory examination DOES show what I believe is reasonably
correct reference counting behavior.  All the deletions were
implicitly done when the interpreter terminated.  

>-- add a PyObject (or better, SoundObject) pointer to the
>   SoundBuffer wrapper class, and make it point to the Sound
>   object when you create the buffer (don't forget to call
>   Py_INCREF on it)
>
>-- only destroy Direct X objects in the Python wrapper's
>   dealloc method.  in the SoundBuffer wrapper, make sure
>   you decrement the pointer to the Sound object.

This is what I did in the Python wrapper code.

My intention is to fix it in the CDX libraries so that I can use the
'same' libraries in Python or C++ programs and not have to remember it
all the time.

Howard Lightstone
EEGSoftware
howard at eegsoftware.com


From mmueller at dgfz.de  Wed Dec 13 04:13:04 2000
From: mmueller at dgfz.de (Mike Müller)
Date: Wed, 13 Dec 2000 10:13:04 +0100
Subject: unpacking Python tuple in C
Message-ID: <917e64$3bp@micky.ibh-dd.de>

Hi,

Calling Python from C I get a PythonTupleObject

PyObject *result = PyObject_CallMethod(...)

PyTuple_Check(result)  is 1 so its really a tuple!

The tuple contains only Python floats.

Using

PyArg_ParseTuple(result, "dd",  &double1, &double2)

does not change double1, or double2.

 I can get  an item of the tuple:

PyObject* floatFromPy0 = PyTuple_GetItem(result, 0);
  *double1= PyFloat_AsDouble(floatFromPy0)

since the tuple is pretty long and I need all of the tuple items
PyTuple_GetItem seems to not the best solution.

Any hints how I can get all the tuple items into C doubles?

Thanks.

Mike




From nospam at no.spam  Wed Dec 20 20:03:27 2000
From: nospam at no.spam (Carsten Geckeler)
Date: Thu, 21 Dec 2000 02:03:27 +0100 (MET)
Subject: Extracting attributes from modules
In-Reply-To: <Pine.LNX.4.10.10012210122320.1436-100000@erde.kosmos.all>
Message-ID: <Pine.LNX.4.10.10012210200130.1759-100000@erde.kosmos.all>

On Thu, 21 Dec 2000, Carsten Geckeler wrote:
[snip]
> 
> Their probably not hidden, but mySphere is "only" an instance, so there
> may be attributes or methods of the class or the base classes.  These
> attributes don't show up with the dir() command.  Try the following:
> 
> print dir(mySphere)
> print dir(mySphere.__class__)
> for bc in mySphere.__class__.__bases__:
> 	print dir(bc)

Forgot something:  There is a possibility that there are attributes not
listed in any of the above, then their lookup is done by the __getattr__
method.  In this case you have to take a look in the source (module, etc.)
or try to get a proper documentation.

Cheers, Carsten
-- 
Carsten Geckeler:  carsten dot geckeler at gmx dot de
To get proper email-address replace `dot' and `at' by the corresponding symbols.




From jcm at bigskytel.com  Mon Dec 25 14:47:08 2000
From: jcm at bigskytel.com (David Porter)
Date: Mon, 25 Dec 2000 12:47:08 -0700
Subject: __init__ question
In-Reply-To: <EvM16.270$LU6.118037@typhoon.aracnet.com>; from DanielK@aracnet.com on Mon, Dec 25, 2000 at 10:45:47AM -0800
References: <EvM16.270$LU6.118037@typhoon.aracnet.com>
Message-ID: <20001225124708.A11957@bigskytel.com>

* Daniel Klein <DanielK at aracnet.com>:

>     def __init__(self, *args):
> 
> What is the meaning of the '*args' expression?

It gathers the remaining non-keyword args into a tuple called args. Another
form is **args which gathers the remaining keyword args into a dict. These
is useful when you require an arbitrary number of arguments passed to a
function.

Note that the order must be func(normal_args, *args, **kwargs).


David



From aahz at panix.com  Sat Dec 23 11:55:56 2000
From: aahz at panix.com (Aahz Maruch)
Date: 23 Dec 2000 08:55:56 -0800
Subject: Exception Handling and Inheritance
References: <921sdr$mlq$1@newsreaderg1.core.theplanet.net>
Message-ID: <922les$bkr$1@panix2.panix.com>

In article <921sdr$mlq$1 at newsreaderg1.core.theplanet.net>,
Franz GEIGER <fgeiger at datec.at> wrote:
>
>But what about Exceptions? They are not caught by my TException handler! And
>Exception does not have my Formatted method.

No, but you could easily implement a function to do that using the
traceback module.

>Do I really have to catch them in a 2nd handler? Or can I add some
>functionality somehow to Exception instead of implementing it in a derived
>TException?

Yes, you'll need to catch other exceptions explicitly.

WARNING: it's generally considered a Bad Idea to catch exceptions
willy-nilly.  For example, do you really want to trap the keyboard
interrupt?
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

'"Crisp" is a good quality for crackers; less so for pot roast.'  --pnh


From pbjorn at uswest.net  Thu Dec 28 13:28:01 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Thu, 28 Dec 2000 11:28:01 -0700
Subject: Erlang (was Re: Good introduction to functional programming with 
 Python?)
References: <mailman.977984412.30211.python-list@python.org> <92f0il01a5u@news1.newsguy.com>
Message-ID: <3A4B8631.BC983CE@uswest.net>

Alex Martelli wrote:

> [snip]

> It sure looks unique to me for its emphasis on concurrency (if there
> are other FP's that are similarly concurrency-oriented, I have not seen
> them yet).

We had to program in Id for an assignment back when I was in school. Cute
language, although hard to find references to...
(http://www.cs.nott.ac.uk/~gmh//faq.html#id
http://citeseer.nj.nec.com/gupta90incremental.html).

-- bjorn



From qdlaty at wielun.dhs.org  Sun Dec 24 19:08:45 2000
From: qdlaty at wielun.dhs.org (QdlatY)
Date: Mon, 25 Dec 2000 01:08:45 +0100
Subject: html filtering
Message-ID: <3A46900D.BD118A0A@wielun.dhs.org>

Hello!

Big problem in writeing message boards is filtering html tags. I use
very simple method: Each occurence of "<" sign or ">" sign I replace
with &lt; and &gt;. Is this safe method? Maybe You've got better ideas?

Yours
QdlatY


From rturpin at my-deja.com  Sun Dec 24 23:28:04 2000
From: rturpin at my-deja.com (rturpin at my-deja.com)
Date: Mon, 25 Dec 2000 04:28:04 GMT
Subject: One Python 2.1 idea
References: <mailman.977590932.19335.python-list@python.org> <9237md$orm$1@nnrp1.deja.com> <924bjt0e3n@news1.newsguy.com> <925baq$5sj$1@nnrp1.deja.com> <925tak01tpu@news1.newsguy.com>
Message-ID: <926ici$vi9$1@nnrp1.deja.com>

In article <925tak01tpu at news1.newsguy.com>,
  "Alex Martelli" <aleaxit at yahoo.com> wrote:
> I don't think a 10% difference matters substantially, either
> way!  _This_ may be the crux of our disagreement.  Sure,
> if the infrastructure can be enhanced to get a few percents
> here and there, why not.  But that's not going to change my
> programming style, nor make much of a difference to the
> effectiveness of the applications I write.

I agree. And if we were discussing an across the board
10% difference between Python and language X, or of any
one change that had ill repercussions, the discussion
would end there.

But we're not.

Performance improvement very often is a matter of 10%
here, 10% there, and 10% in three other constructs, and
then the two sections of your code that were consuming
95% of execution time, each of which exhibit these five
problems, run twice as fast. And two times IS a
significant improvement. It's not any ONE 10% that
matters. It is their cumulative effect. Cumulative both
over time, as these improvements are introduced, and
cumulative over large programs where such improvements
often are synergistic.

Turpin:
>> One BIG benefit to abstractions like JOIN and list
>> comprehension is that performance improves without the
>> application programmers doing anything. The next release
>> of the programming environment makes everything run

Martelli:
> But why shouldn't said 'next release' be able to optimize
> the 'lower-abstraction' approach, too?  As long as the
> language allows and encourages the coder to express
> the code's intent exactly and precisely, optimizations are
> feasible anyway.  In other words, I do NOT consider
> potential optimizations a 'BIG' benefit of higher-level
> abstractions over lower-level ones.

The benefit of abstraction is that it shifts work from
the programmer to the language implementation. The programmer
says LESS about how to produce a result, but still precisely
specifies what result is needed. Because the implementation
is less constrained, it potentially can choose a faster
execution path. Again, consider the JOIN example. The SQL
compiler has a broad range of choices of how to implement
JOIN and which keys to use. It can make use of significant
data to inform this choice (sometimes including the sizes of
tables and distribution of key data in the target database.)
By using this abstract construct, the application programmer
not only saves the trouble of specifying a particular
implementation route, but likely gets a more efficient route
than he had the knowledge to specify. By going to more detail,
the application programmer can defeat the smart SQL compiler.

Of course, this all depends. Maybe the compiler is dumb,
and the application programmer is pushed to using specific
techniques for the sake of performance. Maybe the compiler
is very smart, and even if the programmer hand codes a JOIN,
it sees that the end result is just a JOIN, that the
intermediate results are not used, and it does the fast
thing anyway. Languages have developed differently in how
they mix responsibilities between application programmer
and language implementation.

I am pushing two principles in this regard. (1) Languages
should provide a powerful set of abstractions as simply
as possible. (2) Application programmers and language
implementors should adhere to a separation of
responsibilities. Application programmers should use the
abstractions that best solve their problem, functionally,
assuming that these are implemented well. Language
implementors should strive to make sure that the
implementation for an abstraction performs well in all
the ways it is used. This route provides significant gains
in productivity. But it only works if BOTH groups take up
their mantle.

> When I compare working in Python with working in
> Haskell, say, or perl, or, I _think_, ML dialects (not
> enough practical real-world experience to be sure!),
> it does not seem to me that I'm working at a higher
> level of abstraction; the abstraction-level seems quite
> comparable.  (Actually, this goes for C++, as well,
> most of the time -- but, admittedly, lower-level issues
> _do_ keep interfering often enough in that case).

Perl provides abstractions of similar power, but has
greater language comlexity. In some sense, the practical
power of a language is its level of abstraction divided
by its innate complexity. That denominator puts Perl
much lower on the totem pole than Python.

C++ does NOT provide the same level of abstraction as
Python. It does not have sequences, dictionaries, nor
memory management as built in parts of the language.
Raw pointers are exposed. Classes are not first class
objects. Instances are statically composed. There is
no run-time compilation of code. All this is friction
and detail I am happy to leave behind. I do not have
enough familiarity with the other two to speak. To me,
one of the strong attractions of Python is the notion
of power I just described: it has a very powerful and
practical set of abstractions, while remaining simple.

> But if Python had such a 'listbuffer' object, and
> used it internally for list-comprehensions, I'd
> _also_ like to see it exposed for direct programmer
> use... why not? .. Say I want to build up and return
> a list of Fibonacci numbers, however many are needed
> to get up to the first one that is larger than an
> argument N.  The simplest approach:
>
> def listFib(N):
>     if N<1: return [1]
>     result = [1, 1]
>     next = 1
>     while next <= N
>         next = result[-1]+result[-2]
>         result.append(next)
>     return result
>
> .. if listbuffer objects existed, I might get some
> substantial performance gain through them, without
> (IMHO) affecting readability much, if at all:
>
> def listFib(N):
>     if N<1: return [1]
>     result = listBuffer.listBuffer(1, 1)
>     next = 1
>     while next <= N
>         next = result[-1]+result[-2]
>         result.append(next)
>     return list(result)

I think this approach is misguided. A listBuffer, as
far as I can tell, is functionally identical to a
sequence. We've cluttered the language with another
abstraction, whose only purpose is to help the compiler
decide an implementation. Now, everyone learning Python
has to learn about listBuffers, in order to read your
code. This same mistake has been made already with
xrange(). Given enough time, we'll have to teach
programmers about a dozen different things that all
behave like Python sequences. Blech!

In my mind, the ideal is very different. The Python
implementation should have a dozen or more ways of
implementing sequences, depending on how they are
constructed, what kinds of objects are put into them,
the contexts in which they are indexed, whether they
are used for iteration, etc. The vast majority of the
time, these things can be determined at compile time.
There shouldn't be a different language abstraction
for sequences that are not stored. xrange() is an
abomination. The compiler should simply use this
implementation when it fits.

ONLY ONE SEQUENCE ABSTRACTION SHOULD BE EXPOSED TO THE
APPLICATION PROGRAMMER. It would be a mistake -- a BIG
mistake -- to continue the process of cluttering up
the language with new abstractions, merely to give
implementation directive to the compiler. I vote we
reverse this process, by deprecating xrange(). IF we
want implementation directives, they should be
syntactically distinct annotations, appended to the
relevant code:

    stopWords = {}     #PYDIR: nodeletes

At least that way, programmers reading and debugging
the code can (in theory) ignore the annotation. (The
compiler can also ignore the directive, if it is wrong
headed.) In my opinion, you get far more bang for the
buck if the compiler makes these optimizations
automatically.

(Apologetic prologue: This mind dump is occuring after
Christmas eve celebrations. I hope everyone reads it
generously, and that I am not too embarrassed when I
read it tomorrow.)

Russell


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


From henry_crun3583 at my-deja.com  Sat Dec 30 22:39:34 2000
From: henry_crun3583 at my-deja.com (henry_crun3583 at my-deja.com)
Date: Sun, 31 Dec 2000 03:39:34 GMT
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <05C6FUhLDNUU-pn2-tuTnzOCokxem@bill> <Sgw36.208288$U46.6385684@news1.sttls1.wa.home.com>
Message-ID: <92m9pm$d19$1@nnrp1.deja.com>

It does seem to me that "do: while" represents a critically different
algorithm (ie act, then test information that becomes available as a
consequence of the action) from "while: ..." (test pre-existing
information for need or possibility of action, then act)

Fudging "while: " to do the same thing offers the potential or probable
result that subsequent changes will take it literally as a test-then-
act.

As an aside sometimes I think having both while and until in one
statement would simplify it while still making logical sense..

while file.data_available :
  a=file.read;
  b=process(a)
  until b==DesiredValue


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


From erik_who at my-deja.com  Thu Dec  7 15:43:32 2000
From: erik_who at my-deja.com (Erik Who)
Date: Thu, 07 Dec 2000 20:43:32 GMT
Subject: Python naming conventions
References: <wAQX5.28$Mk1.9001@typhoon.aracnet.com>
Message-ID: <90ospi$1c6$1@nnrp1.deja.com>

> Are there any formal online documents describing Python naming
> conventions and taboos?
>

This is the closest thing that I've come across:

    http://www.python.org/doc/essays/styleguide.html

-Erik


Sent via Deja.com http://www.deja.com/
Before you buy.


From nickatvideosystemdotcodotuk  Fri Dec 15 05:09:02 2000
From: nickatvideosystemdotcodotuk (Nick Trout)
Date: Fri, 15 Dec 2000 10:09:02 -0000
Subject: what is a closure?
References: <3a39033b@news.xtml.co.uk> <S18_5.3929$Kd1.455131@newsb.telia.net>
Message-ID: <3a39edee@news.xtml.co.uk>

What a great resource! Thanks Fred.


"Fredrik Lundh" <fredrik at effbot.org> wrote in message
news:S18_5.3929$Kd1.455131 at newsb.telia.net...
> Nick Trout wrote:
> > Sorry, this is probably not the place to ask this but you lot are pretty
> > friendly :-)
>
> So you don't want me to flame you for not knowing about
> the free on-line dictionary of computing (foldoc)? ;-)
>
>     http://www.foldoc.org/
>
> > I have seen the term "closure" used... what is it exactly? Is it just a
> > class member function?
>
>     http://foldoc.doc.ic.ac.uk/foldoc/foldoc.cgi?query=closure
>
>     a closure is a data structure that holds an expression
>     and an environment of variable bindings in which that
>     expression is to be evaluated. The variables may be
>     local or global
>
> In Python, you have to use explicity syntax to move variables
> from an outer scope (environment) to the inner scope.
>
> In the following example, the "a=a" default argument binding
> binds add's local name "a" to the value of the "a" argument
> from make_adder.
>
>     def make_adder(a):
>         def add(b, a=a):
>             return a + b
>         return add
>
>     >>> m = make_adder(10)
>     >>> m(20)
>     30
>
> In a language with full lexical scoping, you could leave it to the
> compiler to figure this out.
>
> But in today's Python, it's probably better to use a class instead.
>
> </F>
>
> <!-- (the eff-bot guide to) the standard python library:
> http://www.pythonware.com/people/fredrik/librarybook.htm
> -->
>
>




From fredrik at effbot.org  Sun Dec 31 10:35:24 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sun, 31 Dec 2000 15:35:24 GMT
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <slrn94smds.150.grey@teleute.rpglink.com> <FOt36.17$vd1.2224@e420r-atl2.usenetserver.com> <slrn94t1vj.150.grey@teleute.rpglink.com> <mqw36.50$9s1.10042@e420r-atl2.usenetserver.com> <slrn94tqf3.150.grey@teleute.rpglink.com> <3A4F3671.AAA18BFF@stroeder.com> <slrn94uht3.150.grey@teleute.rpglink.com>
Message-ID: <0nI36.2368$AH6.353327@newsc.telia.net>

Steve Lamb wrote:
> I don't plonk people for disagreements.  I plonk them for stupidity.

No wonder you have "despair" in your mail address...

Now go back and read all the posts in this thread.  Repeat until
you understand what I, Rainer, Peter, Remco, and everyone else
has tried to tell you.

Then post an apology.

</F>




From fredrik at effbot.org  Thu Dec 21 02:10:16 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Thu, 21 Dec 2000 07:10:16 GMT
Subject: Greetings from the Interpreter!!!
References: <mailman.977314029.15733.python-list@python.org> <slrn942oun.eoe.grey@teleute.rpglink.com> <G5wCy5.783@world.std.com> <slrn9430ue.ft7.grey@teleute.rpglink.com>
Message-ID: <s1i06.770$Qb7.67152@newsb.telia.net>

Steve Lamb wrote:
> Right, I get that part.  When I C&P&run it all I get is the Please
> Wait..., nothing more.

please wait three more days...

(or, if you want your presents before christmas, look at the
code again.  remember that gunter might be in a different time
zone from you, but he's probably close enough...)

</F>




From robin at alldunn.com  Sat Dec 16 02:37:20 2000
From: robin at alldunn.com (Robin Dunn)
Date: Fri, 15 Dec 2000 23:37:20 -0800
Subject: Bug in BSDDB?
References: <I9z_5.9426$0U2.369913@news20.bellglobal.com>
Message-ID: <UYE_5.521$SKV4.80150683@news.randori.com>

"Bryan Mongeau" <bryan at eevolved.com> wrote in message
news:I9z_5.9426$0U2.369913 at news20.bellglobal.com...
> Greetings Python gurus,
>
> I was doing some benchmarks of the dbhash module when I came across some
> unsettling behavior in Windows 98:
>
> 1 - When shelve creates a new database with the "c" flag, it can only add
> 85652 keys to the dictionary before raising an error in bsddb.
>
> 2- If that error is caught and the write operation tried a second time,
the
> key is written just fine.
>
> 3- A dbhash file with 100000 simple records is 140 MB in Windows, compared
> to 21 MB in Unix. If you attempt to add more than 100000 keys in Windows,
> the size of the file quickly balloons to > 1 GIG. Keep in mind I was using
> unicode strings and cPickle.
>
> 4- An attempt to write more than 100000 keys will result in errors
> described in point 1, with ever-increasing frequency until eventually
> python itself will explode (illegal operation).
>

I don't know what's going on with bsddb, but I tweaked your samples to run
with my new bsddb3, and to use the dbshelve module included in that package.
(You can get it at http://pybsddb.sourceforge.net/)  The first sample uses
the default DB_HASH access method, and the second uses a DB_BTREE.  This was
run on a Win2K system so there may be some behavior differences for win98,
but I'm not sure to what extent.

As you can see from the outputs below the tests performed without crashing
or getting exceptions all the way up to 250000 and 500000 records.  The only
anomalous behavior is that the time for the hash test of 250000 records was
an order of magnitude larger than expected given the time progression for
the previous tests.  It may be that there were a lot of hash collisions...
Eventhough the keys were being added in sequential order the BTree tests
performed very well.

Other advantages to using bsddb3 are things like database cursors for
iterating over all the records in an efficient manner, the ability for
multiple processes and or threads to access the DB at once with automatic
single writer, multi reader locking semantics (you just have to set a flag),
or full transactional capabilities with logging, commit and rollback.

Here's my test run:

[c:\Temp\Others] python dbshelvetest1.py 10000
writing 10000 records
write took  2.31299996376  seconds
reading 10000 records
read took  1.34200000763  seconds

[c:\Temp\Others] python dbshelvetest1.py 50000
writing 50000 records
write took  14.2899999619  seconds
reading 50000 records
read took  7.41100001335  seconds

[c:\Temp\Others] python dbshelvetest1.py 100000
writing 100000 records
write took  38.7259999514  seconds
reading 100000 records
read took  15.0110000372  seconds

[c:\Temp\Others]  python dbshelvetest1.py 250000
writing 250000 records
write took  311.909000039  seconds
reading 250000 records
read took  33.1779999733  seconds

[c:\Temp\Others] dir testdb.*

 Volume in drive C is unlabeled      Serial number is A06A:886B
 Directory of  C:\Temp\Others\testdb.*

12/15/00  22:54       1,368,064  testdb.10000
12/15/00  22:55       5,193,728  testdb.50000
12/15/00  22:56      10,469,376  testdb.100000
12/15/00  23:04      41,361,408  testdb.250000
     58,392,576 bytes in 4 files and 0 dirs    58,392,576 bytes allocated
  4,029,243,392 bytes free

[c:\Temp\Others]



[c:\Temp\Others] python dbshelvetest2.py 10000
writing 10000 records
write took  1.08200001717  seconds
reading 10000 records
read took  0.570999979973  seconds

[c:\Temp\Others] python dbshelvetest2.py 50000
writing 50000 records
write took  4.43599998951  seconds
reading 50000 records
read took  2.8939999342  seconds

[c:\Temp\Others] python dbshelvetest2.py 100000
writing 100000 records
write took  8.70200002193  seconds
reading 100000 records
read took  5.84899997711  seconds

[c:\Temp\Others] python dbshelvetest2.py 250000
writing 250000 records
write took  21.1699999571  seconds
reading 250000 records
read took  14.8210000992  seconds

[c:\Temp\Others] python dbshelvetest2.py 500000
writing 500000 records
write took  40.699000001  seconds
reading 500000 records
read took  29.7120000124  seconds

[c:\Temp\Others] dir testdbbt.*

 Volume in drive C is unlabeled      Serial number is A06A:886B
 Directory of  C:\Temp\Others\testdbbt.*

12/15/00  23:09       1,466,368  testdbbt.10000
12/15/00  23:10       7,299,072  testdbbt.50000
12/15/00  23:10      14,598,144  testdbbt.100000
12/15/00  23:11      37,601,280  testdbbt.250000
12/15/00  23:12      75,956,224  testdbbt.500000
    136,921,088 bytes in 5 files and 0 dirs    136,921,088 bytes allocated
  3,892,310,016 bytes free

[c:\Temp\Others]


--
Robin Dunn
Software Craftsman
robin at AllDunn.com
http://wxPython.org     Java give you jitters?
http://wxPROs.com        Relax with wxPython!





begin 666 dbshelvetest1.py
M(R$@+W5S<B]B:6XO<'ET:&]N#0H-"FEM<&]R="!S>7,L('1I;64-"F9R;VT@
M8G-D9&(S(&EM<&]R="!D8G-H96QV90T*#0IN=6U%<G)O<G,],0T*<F5C8V]U
M;G0@/2!I;G0H<WES+F%R9W9;,5TI#0H-"F1B(#T at 9&)S:&5L=F4N;W!E;B at B
M=&5S=&1B+B5D(B E(')E8V-O=6YT+")C(BD-"G1S(#T@=&EM92YT:6UE*"D-
M"@T*<')I;G0@(G=R:71I;F<@)60@<F5C;W)D<R(@)2!R96-C;W5N= T*#0IF
M;W(@>"!I;B!R86YG92AR96-C;W5N="DZ#0H@(&1A=&$]6W4B<V]M92!T97-T
M(&1A=&$B+"!U(G-O;64@;6]R92!T97-T(&1A=&$B+"!S='(H>"E=#0H@(&ME
M>3UX#0H@('=H:6QE(#$Z#0H@(" @=')Y. at T*(" @(" @9&);<W1R*&ME>2E=
M(#T at 9&%T80T*(" @(" @8G)E86L-"B @("!E>&-E<'0Z#0H@(" @("!P<FEN
M=" B97)R;W(B+"!S='(H;G5M17)R;W)S*2P@(F%T(')E8V]R9"(L('-T<BAX
M*0T*(" @(" @;G5M17)R;W)S*STQ#0H-"F1B+G-Y;F,H*0T*#0IT9B ]('1I
M;64N=&EM92 at I#0IP<FEN=" B=W)I=&4@=&]O:R B+"!S='(H=&8M=',I+" B
M('-E8V]N9',B#0H-"G!R:6YT(")R96%D:6YG("5D(')E8V]R9',B("4@<F5C
M8V]U;G0-"G1S(#T@=&EM92YT:6UE*"D-"F9O<B!X(&EN(')A;F=E*')E8V-O
M=6YT*3H-"B @9&%T82 ](&1B6W-T<BAX*5T-"@T*=&8@/2!T:6UE+G1I;64H
M*0T*<')I;G0@(G)E860@=&]O:R B+"!S='(H=&8M=',I+" B('-E8V]N9',B
4#0H-"F1B+F-L;W-E*"D-"@T*#0H`
`
end

begin 666 dbshelvetest2.py
M(R$@+W5S<B]B:6XO<'ET:&]N#0H-"FEM<&]R="!S>7,L('1I;64-"F9R;VT@
M8G-D9&(S(&EM<&]R="!D8G-H96QV92P at 9&(-"@T*;G5M17)R;W)S/3$-"G)E
M8V-O=6YT(#T@:6YT*'-Y<RYA<F=V6S%=*0T*#0ID8B ](&1B<VAE;'9E+F]P
M96XH(G1E<W1D8F)T+B5D(B E(')E8V-O=6YT+" B8R(L(&9I;&5T>7!E/61B
M+D1"7T)44D5%*0T*=',@/2!T:6UE+G1I;64H*0T*#0IP<FEN=" B=W)I=&EN
M9R E9"!R96-O<F1S(B E(')E8V-O=6YT#0H-"F9O<B!X(&EN(')A;F=E*')E
M8V-O=6YT*3H-"B @9&%T83U;=2)S;VUE('1E<W0 at 9&%T82(L('4B<V]M92!M
M;W)E('1E<W0 at 9&%T82(L('-T<BAX*5T-"B @:V5Y/7 at -"B @=VAI;&4@,3H-
M"B @("!T<GDZ#0H@(" @("!D8EMS='(H:V5Y*5T@/2!D871A#0H@(" @("!B
M<F5A:PT*(" @(&5X8V5P=#H-"B @(" @('!R:6YT(")E<G)O<B(L('-T<BAN
M=6U%<G)O<G,I+" B870@<F5C;W)D(BP@<W1R*'@I#0H@(" @("!N=6U%<G)O
M<G,K/3$-"@T*9&(N<WEN8R at I#0H-"G1F(#T@=&EM92YT:6UE*"D-"G!R:6YT
M(")W<FET92!T;V]K("(L('-T<BAT9BUT<RDL("(@<V5C;VYD<R(-"@T*<')I
M;G0@(G)E861I;F<@)60@<F5C;W)D<R(@)2!R96-C;W5N= T*=',@/2!T:6UE
M+G1I;64H*0T*9F]R('@@:6X@<F%N9V4H<F5C8V]U;G0I. at T*("!D871A(#T@
M9&);<W1R*'@I70T*#0IT9B ]('1I;64N=&EM92 at I#0IP<FEN=" B<F5A9"!T
M;V]K("(L('-T<BAT9BUT<RDL("(@<V5C;VYD<R(-"@T*9&(N8VQO<V4H*0T*
$#0H-"@``
`
end



From francois.granger at free.fr  Fri Dec  8 15:32:33 2000
From: francois.granger at free.fr (=?ISO-8859-1?Q?Fran=E7ois_Granger?=)
Date: Fri, 08 Dec 2000 20:32:33 GMT
Subject: Q for regexp experts
References: <yv2aea77xw2.fsf@lionsp093.lion-ag.de>
Message-ID: <1elc25p.eh6l6mvurk68N@paris11-nas10-39-20.dial.proxad.net>

Harald Kirsch <kirschh at lionbioscience.com> wrote:

> Even with perl-compatible regular expressions it seems to be
> impossible to specify a pattern which matches 
> 
>    `the longest substring not containing separator' SEP
> 
> Put another way, from a given starting point, the pattern shall match
> everything until just before the next occurence of SEP. A first
> attempt is something like
> 
>    [^SEP]*

Stupid idea:

temp = string.split(string, 'SEP')
for string in temp:
    'do what you need with string'
for i in range(len(temp)):
    temp[i]

-- 
"Nous autres, civilisations, nous savons maintenant que nous sommes 
mortelles"
- Paul Val?ry 1919


From fredrik at effbot.org  Tue Dec  5 16:48:15 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 05 Dec 2000 21:48:15 GMT
Subject: CGI POST binary file upload?
References: <90j8j0$cmt$1@nnrp1.deja.com> <p1dX5.3468$Er5.2143@news.indigo.ie>
Message-ID: <zodX5.1943$O%.340817@newsc.telia.net>

Ilariu Raducan wrote:
> In class _sockobject in the file socket.py change the 'r' mode with 'rb'
>
>   def makefile(self, mode='r', bufsize=-1):
>         #MODIFIED   return _fileobject(self._sock, 'rb', bufsize)
>         return _fileobject(self._sock, 'rb', bufsize)

please.  what's wrong with changing the *call* to the
function, instead of the function itself?

btw, your "fix" breaks SocketServer.py, and god knows
how many other programs...

</F>




From aleaxit at yahoo.com  Tue Dec 12 18:24:59 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 13 Dec 2000 00:24:59 +0100
Subject: A question.
References: <9152a0$acb$1@panorama.wcss.wroc.pl> <9156h002uh2@news2.newsguy.com> <915a7j$fe0$1@panorama.wcss.wroc.pl> <915elr0681@news2.newsguy.com> <9166uv$8rs$1@animus.fel.iae.nl>
Message-ID: <916cd20b3b@news1.newsguy.com>

"Carel Fellinger" <cfelling at iae.nl> wrote in message
news:9166uv$8rs$1 at animus.fel.iae.nl...
> Alex Martelli <aleaxit at yahoo.com> wrote:
>
> ...excelent tutorial stuf skipped
>
> now that the tim-bot is exposed as a fake (he has a *face* !:)
> and stopped work^H^H^H^Hposting and C4E still the BDFL's pet
> I just wondered:
>
>   have they sprouted a new timbot, more geared towards newbies,

Just because I was a newbie so recently myself, I think.

>   more polite and friendly maybe,

*More polite and friendly*?!  And here I thought I was getting into
twice as many flamewars as anybody else on this group... typo for
'less', by any chance?  [I don't wink, I don't adverbialize, I iconize
only very occasionally, the polysyllabic eruditeness of my hallmark
circumlocutions pushes my Gunning Fog AND my Flesch-Kincaid up
in the high 20's -- what more could I do to shew hostility...?]


>   with a touch of human fallibility
>   (hence the occasional slip of the keyboard) and named it Alex?
>
> In case you hadn't noticed he too is fond of numbers and seems

Very (particularly very high-precision ones, http://gmpy.sourceforge.net).

> to respond educationally to more posts then others can read.

Just a trick to lure y'all into trusting me (as is the simulated
human fallibility, which seems to be working...).


Alex





From MarkH at ActiveState.com  Wed Dec  6 07:16:33 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Wed, 06 Dec 2000 12:16:33 GMT
Subject: win32evtlogutil: How does it works ?
References: <3A2D556F.6240C661@club-internet.fr> <3A2D8223.4070408@ActiveState.com> <lNmX5.963$K6.104400@juliett.dax.net>
Message-ID: <3A2E2CD7.2090602@ActiveState.com>

Dag Sunde wrote:

> The dll-part of the EventLog _is_ a little bit strange... :-)
> 
> You see... The number you give when logging an event is supposed
> to reference an id of a string-recource in the dll you registered
> 
> So you have to create a resource-file with all the strings(messages)
> you need, include it into a dll, and then reference those strings
> by their id from the outside. This is done for nationalization-
> purposes, I think...

Yes - but win32evtlog.pyd _should_ be such a DLL.  It should magically 
register itself as a valid source of such messages.

All of the messages it comtains are generic - eg, often simply of the 
form "%1" - meaning that they are pretty pointless from an 
internationalization point-of-view - but they _should_ still work (and 
have been known to) for the simple use they were intended.

Mark.



From jcollins at endtech.com  Wed Dec 20 17:44:18 2000
From: jcollins at endtech.com (Jeff Collins)
Date: Wed, 20 Dec 2000 15:44:18 -0700 (MST)
Subject: Python in embedded systems?
In-Reply-To: <_w706.11237$TU6.1069571@ptah.visi.com>
Message-ID: <Pine.LNX.4.21.0012201532420.7529-100000@euclid.endtech.com>

On Wed, 20 Dec 2000, Grant Edwards wrote:

> I'd love to use Python, but I don't know if it will fit into
> the embedded system environment -- I probably can't allocate
> more than about 1MB of memory for the scripting feature.

The Python interpreter for the Palm currently uses approximately 200K for
storage of the VM and will run in less than 100K of heap, including the
compiler.  (Alternatively, Python source could be compiled on the desktop,
eliminated the need for the compiler code and the additional heap required
by the accelerators).  Some of the features have been eliminated to reduce
code storage.  An early port is available at
http://www.isr.uci.edu/projects/sensos/python/ and the next revision
should be available in early January 2001.

Though the port may not be of interest, the overall reduction in resource
usage may.

-- 
Jeffery D. Collins, Ph.D.
Sr. Software Developer
Endeavors Technology, Inc.




From fredrik at effbot.org  Tue Dec 12 13:10:05 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 12 Dec 2000 18:10:05 GMT
Subject: Inheriting the @ sign from Ruby
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <915l04$l3i$1@nnrp1.deja.com>
Message-ID: <1StZ5.3572$Kd1.416037@newsb.telia.net>

Simon wrote:
> Isn't that Java? It *certainly* isn't Python - *everything* is by
> reference in python. Unless I'm *really* confused...

using Roy's definitions, "call by reference" means that you
can change whatever the reference points to.

that's not true in Python.

better avoid the terms completely, since Python's model isn't
neither strict call-by-value (arguments aren't copied), nor strict
call-by-reference (not all arguments can be modified in place).

</F>




From quinn at dinar.ugcs.caltech.edu  Sat Dec 16 21:14:21 2000
From: quinn at dinar.ugcs.caltech.edu (Quinn Dunkan)
Date: 17 Dec 2000 02:14:21 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <r0pzoi11lhz.fsf@black.cs.washington.edu> <5jvZ5.3588$Kd1.418770@newsb.telia.net> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com> <kuu28971kp.fsf@lasipalatsi.fi> <3A380E4E.8537A0FB@san.rr.com> <91a1vr02b02@news2.newsguy.com> <Co9_5.25465$x6.13756579@news2.rdc2.tx.home.com> <91bbq102mtk@news1.newsguy.com> <rWd_5.26820$x6.14119694@news2.rdc2.tx.home.com> <91d5el0me7@news1.newsguy.com> <rOr_5.31120$x6.15241856@news2.rdc2.tx.home.com>
Message-ID: <slrn93o8br.tu0.quinn@dinar.ugcs.caltech.edu>

On Fri, 15 Dec 2000 16:38:47 GMT, Rainer Deyke <root at rainerdeyke.com> wrote:
>Yes, in-place modification should be supported.  I had something like this
>in mind:
>
>class UserInt:
>  def __init__(self, value):
>    self.data = value
>  def __add__(self, other):
>    if isinstance(other, UserInt):
>      return self.__class__(self.data + other.data)
>    else:
>      return self.__class__(self.data + other)
>  __radd__ = __add__
>  def __iadd__(self, other):
>    if isinstance(other, UserInt):
>      self.data += other.data
>    else:
>      self.data += other
>  # Repeat for all other operations
>
>UserInt as written is more properly UserNumber, since it supports longs ints
>and floats.

Just as a matter of style, would it be superior to add a __coerce__ method
that knows how to coerce ints, longs, and floats into UserInts, or make an
explicit check on each method?  __coerce__ saves a lot of typing, but I'm
generally wary of implicit things like coercion.  When I implemented a class
like this (a vector, actually), I used __coerce__, but after trying to wrap my
brain around the control flow, I was thinking maybe I wouldn't next time, even
though it seems to work fine.

Is this fear unfounded?  Is there a better way?


From tbryan at python.net  Wed Dec 27 19:12:24 2000
From: tbryan at python.net (Thomas A. Bryan)
Date: Wed, 27 Dec 2000 19:12:24 -0500
Subject: Just starting out!
References: <92buip$kcp$1@nnrp1.deja.com>
Message-ID: <3A4A8568.1578EA06@python.net>

birdman wrote:
> 
> I just started teaching myself to write code and wonder if Python is the
> best way to go. Also, Is there a better book than "Learning Python" by
> Mark Lutz and David Ascher available for someone at my level.

I think that Python makes a good first language.  I'm teaching 
my wife Python, and I have been fairly happy with how easy it 
is in Python to get past the language and get a beginner thinking 
about good code and intelligent design.  I was rather disappointed 
in _Learning_Python_ for someone with no programming experience.  
Since ordered it before I had a chance to thoroughly review it, 
I ended up writing supplemental notes and exercises.  They're in 
a pretty rough form right now, but I plan to make them available 
at some point.  If you'd like to see the current drafts, e-mail off 
the list, and I'll see what I can put together.

There's a SAMS book, something like _Learning_Python_in_24_Hours_.  
Perhaps you should skim it in a bookstore to see if it's more to your 
style.  

Non-book resources include this newsgroup, www.python.org, and the 
Python tutor mailing list.

---Tom


From nadavh at envision.co.il  Fri Dec 29 18:58:26 2000
From: nadavh at envision.co.il (Nadav Horesh)
Date: Sat, 30 Dec 2000 01:58:26 +0200
Subject: Accitental hang up
Message-ID: <3A4D2522.A45A718@envision.co.il>

I occasionally work with very large arrays (>100000 elements):
sometimes, when I accidentally type a name of an array followed by <CR>
python hangs while trying to format the array for output.  Is there a
nice way to abort the process without closing the python session?

    Nadav.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: nadavh.vcf
Type: text/x-vcard
Size: 349 bytes
Desc: Card for Nadav Horesh
URL: <http://mail.python.org/pipermail/python-list/attachments/20001230/32e759a4/attachment.vcf>

From aleaxit at yahoo.com  Thu Dec 21 09:04:46 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 21 Dec 2000 15:04:46 +0100
Subject: Events in win32com
References: <Y1g=Op6IfWSoJn1ISFtOyuvqwBb3@4ax.com> <FY0=OlZFLy71DtnK62u5TEPlhOts@4ax.com> <3A40060B.10504@ActiveState.com> <cYxAOmxi9FgNERgBo29zuCl5c2h1@4ax.com> <ik724tc2v3cpcqfbt7i1ookvaona0mk5ql@4ax.com> <3A418D02.2040302@ActiveState.com>
Message-ID: <91t2mf02dob@news2.newsguy.com>

"Mark Hammond" <MarkH at ActiveState.com> wrote in message
news:3A418D02.2040302 at ActiveState.com...
> Paul Moore wrote:
>
> > Is it possible to do anything like this in Python? If not, can I do
> > the same thing another way? (I have created ie via
> > DispatchWithEvents(), but the HTML document is a property of this. Can
> > I rewrap it somehow, say as
> >   doc = DispatchWithEvents(ie.document, DocEvents)
>
> This is (should be) as simple as instantiating a sub-class of the makepy
> generated event class, and passing the document.

Yes, but -- how does this play with the bForDemand flag, which
is really a must when the module is as huge as MSHTML...?


> eg, something like:
>
> mod = EnsureModule(...)
>
> class MyEvents(mod.IDocumentEvents):
>    # your methods here....
>
> handler = MyEvents(ie.document)
>
> # handler should start recieving events.

Sure, once I managed to somehow grab the right module it's very
easy; but I think I've not yet grasped how to do that grabbing
in the current 'bForDemand' architecture (indispensable for
really huge type libraries, of course).  For example, consider
the following attempt:


import win32com.client, win32com.client.gencache, pythoncom

def EnsureModuleForTypelibOfObject(obj, bForDemand=1):
    try:
        ti = obj._oleobj_.GetTypeInfo()
        clsid = ti.GetTypeAttr()[0]
        tlb, index = ti.GetContainingTypeLib()
        tla = tlb.GetLibAttr()
        return win32com.client.gencache.EnsureModule(
            tla[0], tla[1], tla[3], tla[4], bForDemand=bForDemand)
    except pythoncom.com_error:
            raise TypeError, "This COM object can not automate the makepy
process - please run makepy manually for this object"

ie = win32com.client.gencache.EnsureDispatch("InternetExplorer.Application")
ie.Navigate('c:/index.htm')
ie.Visible=1

html_module = EnsureModuleForTypelibOfObject(ie.Document)

print html_module

class Handler(html_module.HTMLDocumentEvents2):
    def __init__(self, obj):
        self.seen = 0
        html_module.HTMLDocumentEvents2.__init__(self, obj)
    def Ononclick(self, *args):
        print "clic!", args
        self.seen += 1

h = Handler(doc)

while h.seen < 5:
    pythoncom.PumpWaitingMessages()


What I see on the output:

D:\PySym>python tev1.py
<module 'win32com.gen_py.3050F1C5-98B5-11CF-BB82-00AA00BDCE0Bx0x4x0' from
'd:\py
thon20\win32com\gen_py\3050F1C5-98B5-11CF-BB82-00AA00BDCE0Bx0x4x0\__init__.p
yc'>

Traceback (most recent call last):
  File "tev1.py", line 22, in ?
    class Handler(html_module.HTMLDocumentEvents2):
AttributeError: HTMLDocumentEvents2


and indeed, the generated directory:

D:\Python20\win32com\gen_py\3050F1C5-98B5-11CF-BB82-00AA00BDCE0Bx0x4x0

only contains files named __init__ and DispHTMLDocument (.py and .pyc
for each, of course).

Reading the CLSIDToPackageMap in __init__.py, I can see among
the many, many other lines, the relevant one:

 '{3050F613-98B5-11CF-BB82-00AA00BDCE0B}' : 'HTMLDocumentEvents2'


But how do I instantiate and call genpy.Generator, or arrange for
it to be instantiated and called, to ensure the HTMLDocumentEvents2.py
gets built in the bForDemand setting...?  Is there a canonical/
recommended way...?

I can get the file _built_ by faking, once, a
    DispatchWithEvents('htmlfile',SomeJunkClass),
which I'm sure is not the 'clean' way, but that doesn't insert
HTMLDocumentEvents2 into the html_module's object directory,
anyway.

So, what *is* the beautiful way I'm still missing after all
of my study of sources...?  Pretty please...?


Alex





From aleaxit at yahoo.com  Fri Dec 29 19:30:11 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 30 Dec 2000 01:30:11 +0100
Subject: Question re: eval()
References: <978109654.339714327@news.silcom.com>
Message-ID: <92jah506eb@news1.newsguy.com>

"Clarence Gardner" <clarence at netlojix.com> wrote in message
news:978109654.339714327 at news.silcom.com...
>
> I want to eval() an expression, but have some of the variables in it be
> lazily evaluated.  E.g.,
>
> class D:
>     def __getitem__(self, name):
>         if name == 'n':
>             return 3
>
> d = D()
> x = eval('n + 1', d)
>
> but the eval function accepts only a dictionary as the second parameter.
> Is there any amazing way to do this?

Funny, that was one of my first serious questions to this forum
as I really cranked up my Python use, in 1999 -- and I was soundly
flamed for asking this (timbot was apparently adamant that it was
a better design to have the user enter the expression as 'd.n + 1',
where d.n is easily lazy-evaluated via __getattr__).  I see that this
time you HAVE been given the answer -- the co_names attribute
of the code object that you get by compiling the expression is
almost good enough (not real lazy-evaluation, but at least you do
get to learn which variable names the expression uses -- in some
cases, precomputing those can be OK; this only matters in terms
of performance if a LOT of and/or use (shortcircuiting) in your
typical expressions COULD potentially save you _many_ costly
variable-fetches -- in practice, in my case, using co_names was
more or less OK, though not quite as good as the 'd.n + 1' idea).


Alex





From aleaxit at yahoo.com  Mon Dec 25 03:20:51 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 25 Dec 2000 09:20:51 +0100
Subject: classes (was Re: Same again please for OOP)
References: <mailman.977623149.8184.python-list@python.org> <9257mf0sl1@news1.newsguy.com> <roy-E5B925.13324224122000@news.panix.com> <925khi019hs@news1.newsguy.com> <kubsu1t81x.fsf@lasipalatsi.fi>
Message-ID: <92708d0g4g@news1.newsguy.com>

"Erno Kuusela" <erno-news at erno.iki.fi> wrote in message
news:kubsu1t81x.fsf at lasipalatsi.fi...
> >>>>> "Alex" == Alex Martelli <aleaxit at yahoo.com> writes:
>
> | "Roy Smith" <roy at panix.com> wrote in message
> | news:roy-E5B925.13324224122000 at news.panix.com...
> [...]
> || The mp3 class might return a URL for its location() method, and the
> || CD or Vinyl classes might return a string, "Third shelf from the
> || bottom,
>
> | Looks like a typical case of state mismodeled as behavior.  Why
> | should 'location' be a method -- what does it *DO*?!  It doesnt _do_
> | anything; rather, it's an 'accessor' for a part of the object's
> | state.
>
> i don't understand what is wrong with this - i like making them all
> methods, so i don't have to remember if it's an attribute or method
> later. also overrinding __getattr__ later, if it turns out to be
> needed, is tedious and ugly.

"Tedious and ugly"?!  Just add inheritance from the usual mixin:

class AccessorsMixin:
    def __getattr__(self, name):
        if name.startswith('__'): raise AttributeError,name
        return getattr(self, '__get_'+name)()

and implement method __get_location(self) the same way as you'd
implement location(self) in your style.  What could be easier?  If
you already have a __getattr__, it's equally OK to add this couple
of lines to it, or delegate to this one (delegating has pluses).

Giving accessor methods uniform names is hardly a minus, and
starting them with something suggesting 'get' is particularly good.

It's also easy and fast to get caching of computed attributes this
way -- a slightly modified mixin can cache away the values of
attributes it's accessing, and fetch them from cache later (with,
of course, an 'invalidate-cache' internal method too, to be called
when attributes will have to be recomputed).  This is not always
appropriate, but it's quite handy at times.


> why does the user want/need to know if "location" is stored directly
> as an attribute or if computation is needed to to reconstruct it
> from some other information?

He/she neither wants or needs to -- and what does this have to
do with the price of tea in China?  Access to location is obtained
from client-code by 'myobject.location' anyway: in the common
case where this state is stored (or cached) in an attribute, it's
accessed directly (remember __getattr__ isn't called for those
attributes which are in the object's __dict__!) -- in the rarer case
where accessor-computations are needed, they're transparently
performed.

As for 'what is wrong' with hiding state access behind methods
(most of which don't actually DO anything but fetch state), a
first example is:
    obj.price *= 1.07    # 7% increase
versus
    obj.setPrice(obj.getPrice() * 1.07)    # 7% increase

Why force client code to go through the latter, more complicated
form, when the simpler first form is available so cheap?

Further: Python classes are normally 'expando' (a term that
comes, I believe, originally from Javascript, but the concept
is of course wider) -- attributes can be added unless one takes
pains to disallow this.  Why force client-code to know and
remember which attributes are actually 'in' the original object
(and thus must be accessed through accessor methods, in
your style) and which ones aren't (and thus must be accessed
directly, unless one goes to the serious trouble of building up
an accessor method and sticking it on on-the-fly -- blah!)...?
Uniform attribute-access notation does away with the issue.


These are pragmatical considerations that make direct, simpler
notation preferable.  On a _conceptual_ plane, the key issue
is not confusing state and behavior.  "Methods BEHAVE -- they
DO stuff" and "properties JUST ARE -- I can 'read' them, and
some of them I may also be allowed to 'write'" is a good pair
of rules-of-thumbs to use as guidelines in deciding how to
model something -- method, or property (attribute)?  It's a
very similar distinction to that between functions and variables.

Would you like it if the language didn't let you write
    x = 23
and
    y = 6 + x
but forced you to code everything up as functions, e.g.
    set_x(23)    # instead of x = 23
or
    set_y(6 + x())    # instead of y = 6 + x
...?  I think this is part of why some of us prefer Python to
Scheme -- setq(y, +(6, x)) ("everything as functions") is
not all that natural to many of us.

So why should the author of client-code be any happier with
writing
    obj.set_y(6 + obj.x())
rather than
    obj.y = 6 + obj.x
...?  Other languages may well-nigh force one to frame
things up in terms of accessors/mutators, but that's an
issue with *those* languages -- not one we should copy
in Python, which lets us be more direct and simple.

I note in passing that 'property-notation' is a VERY popular
extension to all sorts of OO languages, *particularly* for
accessors even more than for mutators; e.g., while C++
does not allow for it in its standard, Borland first, and later
Microsoft, each added it as a non-standard extension to
their own versions of C++.  It may be framed as a silly little
piece of quite-dispensable syntax sugar (and I might agree),
but, in terms of mental separation between behavior and
state, it's conceptually commendable (although _in C++_
I would NOT use it in practice -- violating the standard in
this way has costs; but, in Python, it IS standard and easy).


I would not normally touch these subjects in a didactical
introduction to classes in Python -- they're far removed from
what should be a beginner's concerns regarding how to use
Python classes and instances most directly, simply, and
effectively.  In my view, the introduction should start with
'classes as simple "containers" of attributes [data, state]',
move on to initialization-behavior, then to other more
general cases of behavior, thence to polymorphism, and
on to inheritance, contain-and-delegate, and special
methods that allow instances to 'act like' numbers, sequences,
and so on.  __getattr__ and similar techniques would come
in pretty late in the curriculum -- at a point where the general
ideas of state and behavior are well-understood, and these
techniques fit in as very natural ways to maintain them.

But I guess it's good for the 'student' to be exposed early
on to the fact that controversy exists on these themes!-)


Alex





From Blair.Lowe at compeng.net  Sat Dec  2 15:46:47 2000
From: Blair.Lowe at compeng.net (Blair Lowe)
Date: Sat, 2 Dec 2000 13:46:47 -0700
Subject: SSL with python
In-Reply-To: <9097pc+pspc@eGroups.com>
References: <9097pc+pspc@eGroups.com>
Message-ID: <v04220804b64f0e47b847@[192.168.11.33]>

-----BEGIN PGP SIGNED MESSAGE-----

Hi,

I am new with python, and I was wondering if anyone has got it to 
work with SSL connections.

We are trying to do some socket level manipulation using urllib, and 
httplib, and we are getting some messages saying that we are using 
straight http over an SSL connection.

Does anyone know of any information about SSL and python (web or printed work)?

Thanks,
Blair.
Computer Engineering Inc. http://www.compeng.net
Phone: 780 499 5687 (9 - 5 MST) Fax:   780 435 0693 (24 Hours)

-----BEGIN PGP SIGNATURE-----
Version: PGPfreeware 6.5.8 for non-commercial use <http://www.pgp.com>

iQCVAwUBOilgDUZu3Sp72OwBAQFHRQP8CKBpVM8Wll7N2DSjbhtgueqsWohYkjmc
4nymzTfsTPGebMjjE+AnozMe+yzUvpSaXedABw3vGfJME/r807H/uIOrZbMLYhsw
gp0fWL06DJpcNRuAYLHZ6QTD5VjqWxsgm8BZSNA3vzR2I40JYh5BOaVOVjy9S4mA
alP+cyWsFA0=
=38Vk
-----END PGP SIGNATURE-----



From stefanb at mail.vega.bg  Wed Dec 13 11:55:12 2000
From: stefanb at mail.vega.bg (Stefan Boiadjiev)
Date: Wed, 13 Dec 2000 18:55:12 +0200
Subject: Modifying running code? [HACK ALERT]
References: <912m65$pu1$1@news.rz.uni-karlsruhe.de>
Message-ID: <3A37A9F0.369666AA@mail.vega.bg>

hi,

that is one of the features that is missing in the Python in comparison to
Smalltalk.
Smalltalk implementation of this is easier because it deals with 'image' where
resides an actual code, not with files as in Python and in Smalltalk you can
resume from the exception point.

This modifying running code can be used not only to debug but even to develop
some code - imagine that you develop initially only classes and method names
and then on the fly fill the methods with proper code :)
hth
Stefan Boiadjiev

Bjoern Giesler wrote:

> Hi,
>
> I'm writing a kind of mini runtime environment that lets me debug / correct
> buggy Python code "on the fly". This is supposed to work as follows:
>
> 1) an exception gets thrown by the buggy code and is caught by a toplevel
> exception handler
>
> 2) the toplevel handler figures out the function that caused the exception
>
> 3) the function source is loaded into an editor and can be corrected by the
> user
>
> 4) if the code has been saved, it is recompiled and inserted instead of the
> original function
>
> 5) the program is then resumed at the point where the original function had
> been called
>
> Can this be done in Python at all? Can it be done in regular Python or only
> in Stackless Python (which supports continuations and would therefore, I
> believe, make step 5 easy)? Is the whole thing complete bull doodoo? Any
> hints or suggestions?
>
> Thanks in advance for any input,
>                 --Bj?rn



From etsang at my-deja.com  Thu Dec  7 14:04:41 2000
From: etsang at my-deja.com (etsang at my-deja.com)
Date: Thu, 07 Dec 2000 19:04:41 GMT
Subject: building python 2.0 problem!!! Help anyone!!!
Message-ID: <90on04$s4r$1@nnrp1.deja.com>

Hi, when I build Python-2.0.
I enountered the follinw problems:

after building python 2.0, when I first run it, I found I cannot import
any of the standrad modules like sys, string, os etc..

I found out from the FAQ that there is a work around in assocaited
PYTHONPATH with the full path of Lib that solves the problem. however,
when I bring up the new python and print out the sys.path, they give me
wrong paths!
Steps done:
import sys
print sys.path
It prints out path containing lib, there is no lib but Lib in python
distribution.
Even after I changed the lib to Lib in variables LIBDIR and SCRIPDIR of
the Makefile in top direcotry, they still show lib is present.
My question where can I make sys.path pointing to the correct paths?


Sent via Deja.com http://www.deja.com/
Before you buy.


From peter at engcorp.com  Fri Dec  8 02:02:51 2000
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 08 Dec 2000 02:02:51 -0500
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu> <slrn92tu7t.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com> <3A2FE630.5C113AB@bioeng.ucsd.edu>
Message-ID: <3A30879B.C518D9FE@engcorp.com>

Curtis Jensen wrote:
> 
> A while back, someone wrote in and asked what a text editor was.  The
> answers to this were polite and informative, and non-demeaning.  I'm
> curious why this one and why others get ripped on, and others don't.
> Luck of the draw?

My first thought was like Andrei's but I actually followed those links,
as Fredrik did, and realized it wasn't really true that Python info was
easier to find than the other with a quick google search.

So I went back to the original posting to reread it to see if I could
understand why we reacted as we did, and whether the reaction was
warranted.  Let's see, he said:

   "I am doing some research and I would be grateful if 
    anybody could give me some history on the origins of 
    Python."

Hmmm... My immediate gut reaction to that is "you say you are *doing
some research* but you show absolutely no evidence of having done any
before asking us for help".  In other words, his time is more valuable
than ours and it isn't worth his even *trying* to find something, since
we're sitting here waiting for his questions because we have nothing
better to do.

Had the original poster said even "I am trying to research the origins
of Python and haven't been able to find any links that help" I suspect
he would have received little from anyone but helpful links to the real
thing.

Those of us who have made even the mildest attempt to actually *do* such
research have found the information rather easy to find.  Maybe not at
the top of a search engine result, but easily found nonetheless.  I just
tried going to www.python.org, for example, and within about thirty
seconds found that clicking on "Documentation", then "FAQ" led me to
http://www.python.org/doc/FAQ.html#1.15, which contains a good start.

In other words, if the original poster didn't just want us to do his
research for him, as implied, he shouldn't have suggested he hadn't done
any himself.  Or perhaps it was simply a poor understanding of the
sentiment of many people (some of us old-timers?) who have long since
become tired of "do my homework" type postings.


From fredrik at effbot.org  Wed Dec 20 14:45:19 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 20 Dec 2000 19:45:19 GMT
Subject: object deletion order question
References: <3a41016d.3397213@news-server.socal.rr.com>
Message-ID: <j%706.700$AH6.70137@newsc.telia.net>

howard at eegsoftware.com wrote:
> However, that led my to question:
>         What ordering does Python use for object deletions?

any order it likes.  it may remove an object as soon as
possible, or never at all.

> When Python terminated (or at other deletions), it called the
> delete method of the SoundBuffer  objects after deleting the
> Sound objects, leaving DirectX calling into unknown data space
> (I know..it's a Microsoft tradition).

Direct X uses reference counting just like Python.  if you mess
up the reference count in Python, it'll crash as well.

avoiding this is easy, of course.  just make sure you keep the
reference counts in sync:

-- use Python wrappers for all reference-counted Direct X
   objects (e.g. SoundObject, SoundBufferObject)

-- only access Direct X objects via the Python wrappers
   (e.g. through an LPDIRECTSOUND member in the Sound
   object class)

-- add a PyObject (or better, SoundObject) pointer to the
   SoundBuffer wrapper class, and make it point to the Sound
   object when you create the buffer (don't forget to call
   Py_INCREF on it)

-- only destroy Direct X objects in the Python wrapper's
   dealloc method.  in the SoundBuffer wrapper, make sure
   you decrement the pointer to the Sound object.

if you do this, the sound object will never be destroyed before
any of its sound buffers.

> Of course, I realize (now) I should not have *assumed* ANY
> deletion order.

why are you asking when you've already figured it out ;-)

</F>




From bernhard at abnoba.intevation.de  Thu Dec 28 04:43:57 2000
From: bernhard at abnoba.intevation.de (Bernhard Reiter)
Date: 28 Dec 2000 09:43:57 GMT
Subject: fcgi.py
References: <3a4a865c.1325844015@localhost>
    <3a4a90ab.1328483109@localhost>
Message-ID: <92f20t$re4$4@newsserver.rrzn.uni-hannover.de>

In article <3a4a90ab.1328483109 at localhost>,
	victor at prodigy.net (Victor Muslin) writes:

> I just found an old posting that says that socket.fromfd() is not
> implemented on Windows platform (though Python documentation does not
> say so). Is there a FastCGI solution for Python for WindowsNT???

Not that I know of. :(
I ran into the same problem and the solution to fix fcgi.py to use
named pipes under WindowsNT is not trivial. Just peek at the mod_fastcgi
sources themself to see what tricks they have to pull under WindowsNT.
(Another problem is that the versions of apache I have tried under
WindowsNT all were buggy regarding fcgi. They did not even run with
the fastcgi c examples and error reporting was lousy.)

The part of fcgi and apache that you want of course is the
dynamic capabilities of apache to create as many fcgi instances as
needed. Fortunatly there seems to be another solution, which still
stands to be tested here:

Use mod_python with apache and program your python script in a way
that it will do most of its stuff when importing the modules onces.
Then subsequent calls in the same apache thread will be in the same
name space of the python subinterpreter. So you can use the apache
thread load balancing of the server and answer multiple requests
fast.

Let me know if you find out more!
	Bernhard


From daniel at cs.washington.edu  Tue Dec 12 15:05:52 2000
From: daniel at cs.washington.edu (Daniel Wood)
Date: Tue, 12 Dec 2000 20:05:52 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <r0pzoi11lhz.fsf@black.cs.washington.edu> <5jvZ5.3588$Kd1.418770@newsb.telia.net>
Message-ID: <r0py9xl1k2n.fsf@black.cs.washington.edu>

"Fredrik Lundh" <fredrik at effbot.org> writes:
> 
> lambda is a reserved word, so I doubt that's what you're
> doing.  Let's call it "spam" instead.

Oops.  I was just making up an example for the post and lambda was the
first parameter that came to mind.

> Besides being a syntax error, it doesn't look very Pythonic
> to me.  If you want to change a named value, the best way
> is to make it an attribute of a class.

Hmmm, I'm open to the suggestion that I need to rethink my plan at a
larger scale, but isn't a mutable int likely to sometimes be useful?
(Just like a mutable list is sometimes useful.)  And if so, is [ 2 ]
the currently favored idiom?  (And if so, isn't there something
better?)

Thanks,
Daniel.


From aleaxit at yahoo.com  Fri Dec 22 17:56:19 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 22 Dec 2000 23:56:19 +0100
Subject: A Suggestion for Python Dictionary/Class
References: <920fe9$c20$1@hecate.umd.edu>
Message-ID: <920m9h011uh@news2.newsguy.com>

"William Djaja Tjokroaminata" <billtj at y.glue.umd.edu> wrote in message
news:920fe9$c20$1 at hecate.umd.edu...
    [snip]
> are implemented using dictionary.  The danger with dictionary is, even
> though a key does not exist, we still can make an assignment to it.  When
    [snip]
> Because Python provides a read-only list in terms of tuple, I think it
> will be nice if Python provides another class type with fixed members,
> just like a dictionary with read-only keys.  Any opinion, anyone?

My opinion is that it would be nice for Python dictionaries to
provide various kinds of 'locking' options in general, but the
specific request you have is very easy to implement in today's
Python, right as it stands.

class FixedMembersMixin:
    _expando = 0
    def __setattr__(self, name, value):
        if not self._expando and not hasattr(self,name):
            raise AttributeError, name
        self.__dict__[name]=value

Just inherit your class from FixedMemberMixin: you'll be
able to set in an instance only those attributes which are
class attributes too.

For example:

class Point2d(FixedMembersMixin):
    x=y=None
    def __init__(self, x, y):
        self.x=x
        self.y=y

Now, for example:

>>> p = Point2d(1.2,3.4)
>>> print p.x
1.2
>>> p.x *= 1.1
>>> print p.x
1.32
>>> p.z = 7.8
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "tep.py", line 5, in __setattr__
    raise AttributeError, name
AttributeError: z
>>>

The mixin only lets you set instance attributes that already
are present (as either instance or class ones).  The _expando
attribute can be set to some true value to turn this off in a
temporary way, for convenience; or, you can set the
attributes explicitly via the self.__dict__[name]=value way.
Neither is a problem as I see it -- no accidents are likely!
Anyway, you don't have to have _expando unless you _wish_
to have it -- the mixin's even simpler if you omit it, of course
(down to just 5 lines, and one of them gets shorter:-).

You may choose different conventions to encode what
attributes can be set, of course.  The key point is that Python
already has all the machinery you need to get your wish,
with just a few lines' worth of a mixin class (it doesn't
have to be a mixin, that's just the handiest way to do it).


Alex





From typeon at netsgo.com  Thu Dec 21 00:27:07 2000
From: typeon at netsgo.com (°­»ó¹è)
Date: Thu, 21 Dec 2000 14:27:07 +0900
Subject: file upload CGI error
References: <91s24b$cf8$1@nnrp1.deja.com>
Message-ID: <210SE4waAHA.299@news1.sys.netsgo.com>

try again

<form enctype="multipart/form-data" action=... method=post>
...
</form>

<sparrowhawk9909 at my-deja.com> wrote in message
news:91s24b$cf8$1 at nnrp1.deja.com...
> I was using python 1.5.1 as CGI script with IIS
> on windows NT. That was
> working fine. But when I upgrade to python 1.5.2,
> every time I upload
> binary file such as GIF files cause CGI error.
> I've tried python 1.6
> and python 2.0 also, but the result was same.
> Followings are the
> error message.
>
> CGI Error
> The specified CGI application misbehaved by not
> returning a complete set
> of HTTP headers. The headers it did return are:
>
>
> Traceback (most recent call last):
>   File "C:\InetPub\wwwroot\python\sample2
> \upload.py", line 5, in ?
>     form = cgi.FieldStorage()
>   File "c:\inetpub\wwwroot\cgi.py", line 869, in
> __init__
>     self.read_multi(environ, keep_blank_values,
> strict_parsing)
>   File "c:\inetpub\wwwroot\cgi.py", line 954, in
> read_multi
>     headers = rfc822.Message(self.fp)
>   File "c:\python\lib\rfc822.py", line 92, in
> __init__
>     self.readheaders()
>   File "c:\python\lib\rfc822.py", line 135, in
> readheaders
>     startofline = tell()
> IOError: (0, 'Error')
>
> Can anyone help me out ?
> Thanks.
>
>
>
>
> Sent via Deja.com
> http://www.deja.com/




From phlip_cpp at my-deja.com  Sat Dec 30 14:01:22 2000
From: phlip_cpp at my-deja.com (Phlip)
Date: 30 Dec 2000 19:01:22 GMT
Subject: ? The best web language ?
References: <92la49$t13$1@SOLAIR2.EUnet.yu>
Message-ID: <92lbe2$p4o@dispatch.concentric.net>

Dejan Bizinger wrote:

> We are making a survey for Broker Pal Newsletter so we are interested what
> is your opinion about the best web language. Is it Java, Perl, SQL...You
> can also tell why do you think that your choice is the right one, too.

So if you've obviously never heard of Python or PHP, and don't seem to have 
any idea that "comp.lang.java" is obsolete and "comp.lang.java.javascript" 
is an oxymoron, how credible could your preliminary research for this 
project have been???

-- 
?Phlip
======= http://users.deltanet.com/~tegan/home.html =======


From sorular at netscape.net  Sat Dec  9 00:21:27 2000
From: sorular at netscape.net (sorular)
Date: Sat, 09 Dec 2000 05:21:27 GMT
Subject: Using python-PgSQL and the doublequote ("") problem...!
Message-ID: <3a31ba9e.105555070@news-server.bigpond.net.au>

G'day

Is there a standard sintax to filter the data before entering it into
PgSQL database. to prevent parsing errors ?

hk



From dale at out-think.NOSPAMco.uk  Wed Dec 13 08:45:41 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Wed, 13 Dec 2000 13:45:41 +0000
Subject: Direct pixel-by-pixel drawing (with Tk)
References: <6izoi0bmwd.fsf@ids.org.au>
Message-ID: <28ve3tok60epmertqg0jm53c0ufqbdj38o@4ax.com>

Jonathan Lange <jml at ids.org.au> wrote:

>
>Hello,
>
>I'm relatively new to Python (but not to programming), and I'd like to
>muck around by doing some fractal stuff. However, alot of this stuff
>needs (wants?) setting pixels directly.
>
>Now, of course, one solution is to draw a bunch of lines of length 1,
>but TK seems to want to remember each bloody line its drawn, which
>is a bit of overkill and a memory hog.
>
>I'm doing this on Linux and Windows, so a platform-independent
>solution would rock.
>
>Thanks,
>jml
>
>PS. I'm not afraid of reading documentation :)

Unless you really *have* to use TK, take a look at wxPython. www.wxpython.org


--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants


From fredrik at effbot.org  Wed Dec 13 18:18:53 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 13 Dec 2000 23:18:53 GMT
Subject: Direct pixel-by-pixel drawing (with Tk)
References: <6izoi0bmwd.fsf@ids.org.au>
Message-ID: <xtTZ5.3870$Kd1.447805@newsb.telia.net>

Jonathan Lange wrote:
> I'm relatively new to Python (but not to programming), and I'd like to
> muck around by doing some fractal stuff. However, alot of this stuff
> needs (wants?) setting pixels directly.

Create a PhotoImage object, fill it with data, and display it.

One way to update a PhotoImage object is to use the "put"
method.  This method takes a tuple of pixel rows, where each
row is a tuple containing color values (names or #rrggbb hex
strings).

    label = Label()
    label.pack()

    data = (
        ("#000000", "#404040"),
        ("#808080", "#FFFFFF")
    )

    im = PhotoImage(width=len(data[0]), height=len(data))
    im.put(data)

    label.config(photo=im)
    label.im = im # make sure to keep a reference!

Another approach is to use PIL's ImageTk module, and do the
drawing in a PIL Image memory.

PIL sources and docs are available from PythonWare:

    http://www.pythonware.com/products/pil/index.htm

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From china at the.web.com  Mon Dec  4 21:35:00 2000
From: china at the.web.com (Albert Chin-A-Young)
Date: 5 Dec 2000 02:35:00 GMT
Subject: urllib2, proxy, basic authentication
Message-ID: <90hk8k$fqh$1@bob.news.rcn.net>

I'm trying to use urlopen in urllib2.py with an HTTP proxy to connect
to a site requiring BASIC authentication. My code follows:

  import os, sys, urllib2
  from urllib2 import urlopen

  os.environ['http_proxy'] = 'http://[PROXY_HOST]:5865'

  authinfo = urllib2.HTTPBasicAuthHandler ()
  authinfo.add_password ('[REALM]', 'http://[URL]/', '[LOGIN]',
  '[PASSWORD]')

  opener = urllib2.build_opener (authinfo)
  urllib2.install_opener (opener)

  url = urlopen ('http://[URL]/')
  print url.info ()
  url.close ()

This is not working. Does urllib2 support this configuration?

    File "/tmp/a.py", line 15, in ?
      url = urlopen ('http://updates.thewrittenword.com/')
    File "/opt/TWWfsw/pkgutils12/lib/python20/lib/python2.0/urllib2.py",
  line 137, in urlopen
      return _opener.open(url, data)
    File "/opt/TWWfsw/pkgutils12/lib/python20/lib/python2.0/urllib2.py",
  line 325, in open
      '_open', req)
    File "/opt/TWWfsw/pkgutils12/lib/python20/lib/python2.0/urllib2.py",
  line 304, in _call_chain
      result = func(*args)
    File "/opt/TWWfsw/pkgutils12/lib/python20/lib/python2.0/urllib2.py",
  line 764, in http_open
      return self.parent.error('http', req, fp, code, msg, hdrs)
    File "/opt/TWWfsw/pkgutils12/lib/python20/lib/python2.0/urllib2.py",
  line 351, in error
      return self._call_chain(*args)
    File "/opt/TWWfsw/pkgutils12/lib/python20/lib/python2.0/urllib2.py",
  line 304, in _call_chain
      result = func(*args)
    File "/opt/TWWfsw/pkgutils12/lib/python20/lib/python2.0/urllib2.py",
  line 430, in http_error_default
      raise HTTPError(req.get_full_url(), code, msg, hdrs, fp)
  urllib2.HTTPError: HTTP Error 401: Authorization Required

-- 
albert chin (china @at@ thewrittenword .dot. com)


From dfarrell01 at earthlink.net  Tue Dec 26 10:10:16 2000
From: dfarrell01 at earthlink.net (Doug Farrell)
Date: Tue, 26 Dec 2000 15:10:16 GMT
Subject: Python and Napster
Message-ID: <sx226.18970$RC1.511898@newsread1.prod.itd.earthlink.net>

Hi,
Is there a way to build a 'Napster' like client in Python? What I'm really
looking for is a way to have a cross-platform (Windows, Mac, Sun) way of
having users share files, no necessarily music files, from machine to
machine. I'd like to do this in Python as I'm the only Windows programmer at
my company and we'd like to support the other two platforms I mentioned.
Maybe I should be looking at Java?

Thanks in advance,
Doug




From j_gergic at yahoo.com  Mon Dec 11 07:55:03 2000
From: j_gergic at yahoo.com (Jaroslav Gergic)
Date: Mon, 11 Dec 2000 13:55:03 +0100
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com>
Message-ID: <3A34CEA7.CD9D85AE@yahoo.com>

Hello,

one thing I dislike in many Python tutorials is
the style of infinite 'while' loop with 'break' statements:

> Python:
> 
>     file = open('file.input')
> 
>     count = 0
>     while 1:
>         lines = file.readlines(8192)
>         if not lines: break
>         count = count + len(lines)
> 
>     print count
> 


I hate it so much I prefer to write something like this:

line = fh.readline()
while(line != ""):
  ... do something ...
  line = fh.readline()

OK, it is ugly to write the same line of code twice,
but 'while true' or 'while 1' is very non-programmer construction
it smells like VisualBasic GOTO.

Do you know better construction to avoid both - duplicate lines
and horrible 'while 1' construction?

Something like:

while( "" != (line = fh.readline()) ):
  ... do something ...

Thanks
Gergi


From fredrik at effbot.org  Sun Dec 31 07:37:47 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sun, 31 Dec 2000 12:37:47 GMT
Subject: Tkinter: title oddity
References: <3A467C2F.548F5D42@columbus.rr.com>
Message-ID: <vMF36.2799$Qb7.322152@newsb.telia.net>

Dan Underwood wrote:
> root.title('my nice title')
> and
> root.title(string='my nice title')
>
> produce the same result... but
>
> someOtherToplevel.title('my nice title')
> works ok
> but
>
> someOtherToplevel.title(string='my nice title')
> gets the message: unexpected keyword 'string'
>
> Did I do it wrong, or is this a known inconsistency?

To give library implementers a chance to change things, please
don't use keyword arguments unless the documentation explicitly
tells you to.

(title takes a positional argument, not a keyword argument).

But it shouldn't really matter in this case -- there's only one "title"
implementation in Tkinter (in the Wm mixin), and it takes a single
argument called "string".  Maybe "someOtherToplevel" is a Toplevel
subclass where title has been overridden?

</F>




From aleaxit at yahoo.com  Mon Dec 18 11:05:51 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 18 Dec 2000 17:05:51 +0100
Subject: while true: !!!
References: <B43D149A9AB2D411971300B0D03D7E8B0FBCBB@natasha.auslabs.avaya.com> <v04220801b663769e1800@[192.168.11.33]> <mailman.977147642.18708.python-list@python.org>
Message-ID: <91lclt02qr3@news1.newsguy.com>

"D-Man" <dsh8290 at rit.edu> wrote in message
news:mailman.977147642.18708.python-list at python.org...
> > It is cheap to define true and false once,
>
> I think that some people were confused with this statement.  He meant
> that 'true' and 'false' would be constants like in C++ and Java.  It
> would be illegal to try
>
> true = false
>
> Thus while true: is equal to while 1: except that a boolean, not an
> integer is used for the test.  I, for one, would like to abolish the
> concept that an integer can denote true or false.  How many functions
> return 0 for success?  (just take a look at the standard C library).

Depending on how you define 'success', 0 is a pretty popular
'success' return-value, with a return value != 0 defining an
error-code.  That is how C's main() function works, for
example; it's also the COM binary standard (but, in the COM
mappings for languages that have the concept of 'exception',
COM error returns are mapped into exceptions -- however, the
binary standard is defined so it can be handled by C code,
and thus cannot rely on the concept of 'exception').

Few languages equate 'success' (not raising an exception, for
languages where 'exception' means something) with 'a boolean
result' (Icon is rather oriented that way, but that's because
it _backtracks_ over 'insuccess' -- a very different setup
than Python's).  And such a mapping would not be very normal,
natural, or idiomatic in Python, though you can get it, if
you want, by wrapping things up to convert true/false results
to not-raise/raise or vice versa.

An explicit Boolean type is very useful in languages where
you can define *overloads* based upon it -- i.e., reasonably
advanced languages with static typesystems; that is specifically
the reason for the design decision to add it to C++, for
example (see Stroustrup's "Design and Evolution" book), though
keeping _also_ the C convention (inevitable, for backwards
compatibility) makes things pretty awkward sometimes.

In a language without overloads, and, more generally, without
static typechecking, it's hard to see what adding such a type
buys you.  Presumably, you'd have to add a type conversion
function, say 'boolean', and most of the current if and while
statements would become "if boolean(whatever):" (assuming
you're not as careful about backwards compatibility as C++
always was -- else, nothing at all would change, and you'd
just have the slight conceptual cost of another elementary
type in the language _without_ any extra feature coming from
its presence...?).

I think it would be inconsistent, un-Pythonic, to add a
fundamental type for 'conceptual' reasons, without any
practical, measurable benefit.  For example, Python does
away with the distinction between 'a character' and 'a
string of length 1' -- a distinction that is religiously
maintained in C, Pascal, Java, etc, and does have a basis
in both 'conceptual' and machine-level considerations --
yet the strong conceptual *simplification* arising from
just doing away with the distinction appears to be quite
positive.  Practical benefits -- constructs, idioms,
patterns, of real interest, that are now awkward or
impossible, and would become easy or at least possible --
had better be exhibited, and shown not to reduce to
duplicating what is best done with exceptions (success
vs failure issues) or lead to 'type-switches' (a horrid
meta-idiom that one would love to _stamp out_, rather
than _encouraging_ it...:-).

Note that a *separate* case would have to be made for
the existence of "constants like in C++ and Java"; this
is another concept which Python does not currently
support (without much noticeable loss, altough, in
this case, I think it's more open to discussion).  I
suggest not arguing for two separate fundamental changes
to the language in the same breath -- I think getting
the BDFL to accept _one_ such change will already be
plenty hard enough:-)


Alex





From ssartor at bellatlantic.net  Sat Dec 16 16:58:23 2000
From: ssartor at bellatlantic.net (Steven Sartorius)
Date: Sat, 16 Dec 2000 21:58:23 GMT
Subject: Can't get Tkinter to run from command line
References: <MEM_5.1372$L3.323917@typhoon2.ba-dsg.net> <91gg43$7vi$1@animus.fel.iae.nl>
Message-ID: <3AR_5.859$TC3.442356@typhoon2.ba-dsg.net>

In article <91gg43$7vi$1 at animus.fel.iae.nl>, "Carel Fellinger"
<cfelling at iae.nl> wrote:

> Steven Sartorius <ssartor at bellatlantic.net> wrote:
> 
> you allmost got it, it's just that main is spelled differently:)
> 
>> if __name__ == 'main': gui().mainloop()
> 
> try:
> 
>   if __name__ == '__main__': gui().mainloop()
> 
> 

It's always the obvious.......!!

Thanks,

Steve


From junaftnoon at nospamplzyahoo.com  Fri Dec  8 19:08:04 2000
From: junaftnoon at nospamplzyahoo.com (June Kim)
Date: Sat, 9 Dec 2000 09:08:04 +0900
Subject: The secrets of operator.add
References: <g08zpq51vy.fsf@scumbag.ecs.soton.ac.uk> <90r69r02hmn@news2.newsguy.com>
Message-ID: <90rt6p$3s1$1@news.nuri.net>

"Alex Martelli" <aleaxit at yahoo.com> wrote in message
news:90r69r02hmn at news2.newsguy.com...
> "Jacek Generowicz" <jmg at ecs.soton.ac.uk> wrote in message
> news:g08zpq51vy.fsf at scumbag.ecs.soton.ac.uk...
>     [snip]
> > It looks as if multiple calls to operator.add are avoided somehow
> > . . . or am I failing to understand the profiler's output ?
>
> I think the profiler may be inserting its tracefunction in some
> places and not others, thus perhaps distorting measurement.
>
[snip]

How come the huge difference in its total running time between the two?




From wware at world.std.com  Mon Dec 11 07:03:34 2000
From: wware at world.std.com (Will Ware)
Date: Mon, 11 Dec 2000 12:03:34 GMT
Subject: making a DLL
References: <G5ADuH.GC8@world.std.com> <XgnY5.2712$O%.428036@newsc.telia.net>
Message-ID: <G5EK5y.73v@world.std.com>

Fredrik Lundh (fredrik at effbot.org) wrote:
> or you could build your own libraries.  since you have 5.0,
> all you need to do is to fire up the developer studio, open
> the project file (in the PCBuild directory), press "build", and
> wait a couple of minutes...

Thanks, this was very helpful. I never managed to get the DLL to work,
but I only had one short function I wanted to add, and I made it an
extra method of complex numbers so that it got built into python.exe.
-- 
# - - - - - - - - - - - - - - - - - - - - - - - -
# Resistance is futile. Capacitance is efficacious.
# Will Ware	email:    wware @ world.std.com


From aleaxit at yahoo.com  Mon Dec 18 04:46:29 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 18 Dec 2000 10:46:29 +0100
Subject: Singleton vs Proxies DP (was Re: Solution: Direct access to Printer   I/O lines)
References: <917k9b$9i9$1@nnrp1.deja.com> <slrn93hrjn.2vp.ssthapa@localhost.localdomain> <91cnp3$g59$1@nnrp1.deja.com> <91d23f0jk7@news1.newsguy.com> <91gfnp$7r4$1@animus.fel.iae.nl> <91gkh801e49@news2.newsguy.com> <3A3CB83D.45C8775E@darwin.in-berlin.de> <91iu95013ta@news2.newsguy.com> <3A3D1D49.1A453B51@darwin.in-berlin.de>
Message-ID: <91kmfd020hp@news1.newsguy.com>

"Dinu C. Gherman" <gherman at darwin.in-berlin.de> wrote in message
news:3A3D1D49.1A453B51 at darwin.in-berlin.de...
> Alex Martelli wrote:
> >
> > I agree that unique id is part of the Singleton Pattern:
> > that is what makes it problematic.  One does not really
> > _care_ about that identity, only about the state and
> > behavior it represents.
>
> For me it sounds like being part of the essence of that
> pattern... but you're free to disagree.

Identity is 'part of the essence' only if a typical use case
includes client-code doing identity checks -- in Python:
    same = a is b
or
    same = id(a) == id(b)
or, in C++:
    bool same = dynamic_cast<void*>(a) == dynamic_cast<void*>(b)
or, in COM:
    IUnknown *pUa, *pUb;
    a->QueryInterface(IID_IUnknown, (void**)&pUa);
    b->QueryInterface(IID_IUnknown, (void**)&pUb);
    bool same = pUa == pUb;
    pUa->Release();
    pUb->Release();
and so on, and so forth.

This is so far from typical that the language-idiom for
object-identity-checking is considered one of the 'obscure'
parts of C++.  It's not quite so obscure in Python, given
the 'is' keyword and easy exposure of 'id', and the COM
idiom is well-known, since it's what _defines_ the notion
of object-identity in that object-model... which is very
relevant to (e.g.) remoting-infrastructure supplied by it
(proxy/stub and marshaling issues).

But _usage in 'normal', "application"-level, client-code_
is still pretty unusual, whatever the implementation language;
indeed, if I had a situation in COM where identity-checks
_did_ come up as an important use-case, I would much rather
choose to expose an Identity *property* from the relevant
object, to decouple "application-level" checks from
infrastructure-level identity notions and implications (I'd
rather keep my ability to, e.g., tweak/optimize delicate
marshaling situations which might arise in different remoting
and deployment scenarios, than be constrainted in such
implementation level fine-tuning optimizations by
application-level semantic needs!).

Using a GUID as the type/representation for 'identity' would
also be more flexible in several ways, rather than having
to have application-level code keep the library-side object
alive just to be able to later check 'if it's still the same':
GUID's can be easily persisted and depersisted, do not imply
continuing connection between client and server (and thus,
maybe, 'keep-alive' protocol overhead in a remote scenario),
and don't go against the grain of MTS/COM+ or other cases of
transactioning (where object-identity as naively seen by many
OO analists/designers/coders is often problematic, or can at
the very least carry substantial scalability costs).  Similar
advantages for system-level-id / application-level-identity
decoupling appear in Python and C++, though the semi-transparent
remoting possibilites of COM & friends (and the system-level
services 'keyed' on system-level-identity) make it stand out.

To reiterate, that's even in the _rare_ case where application
semantics DO require checking equality of identities for
system-supplied 'unique' ('singleton') objects; such requirements
are *not* typical.


> > No!  I'm not saying two clients can't hold their
> > 'handles' (lw proxies) at the same time.  It IS
> > pretty typical.  But it's not at all typical for client
> > code to be interested in ensuring _identity_ --
> > state and behavior are typical.
>
> In the context of Singletons I would even reduce that to
> state only, like with counters of some system-wide entities
> of I-don't-know-what. A client is "typically" not able to
> compare *bahaviour*, but state, yes. In all cases where

A client is not able to *compare* behavior, but it IS,
typically, quite interested in getting server-side objects
with "just THAT behavior", just as it is interested in
getting them with "just THAT state" -- even the portions
of the state which it's not meant to access directly, in
as much as they do affect behavior.  "Is not meant to"
may mean relatively little in a pure-Python scenario, as
client-code *may* choose to dig into object internals
even though it _shouldn't_, but object models providing
stronger encapsulation may even make portions of state
_totally unaccessible_ by the client (e.g., that IS quite
common for COM, Corba, and other object models with some
orientation to remoting).

So client code *ASKS* in some way for the state and
behavior it needs (in some abstract terms) and it's the
server-side (library, etc) job to supply an object that
works right for those requests.  "Works right" and "has
a specified identity which the client can compare to
other object identities" do not need to be coupled, and
such coupling as enforced in the Singleton Pattern as
the GoF wrote it up is typically a design weakness.

> this comparison is difficult or time-consuming, comparing
> or, probably more often in fact, *accessing* the current
> state via a "globally identical" object is what a Singleton
> is about - if I understood anything at all about this
> pattern.

I do agree that the GoF-Singleton-Pattern IS "about"
global identity -- that's exactly why I'm dissing it,
in favour of patterns that decouple the actual identity
of the objects that the library/server/whatever supplies,
from the state-and-behavior issues that are client-side's
typical needs and interests.

My claim, in your terms, is that *client-side needs* are
NOT "about" verifiable ('system-level') object-identity,
and quite rarely 'about' object-identity at all; thus,
it is not optimal to answer such needs with a design
pattern that _hinges_ on object-identity.


Alex





From prestonlanders at my-deja.com  Wed Dec 20 17:41:21 2000
From: prestonlanders at my-deja.com (Preston Landers)
Date: Wed, 20 Dec 2000 22:41:21 GMT
Subject: mail processing module?
Message-ID: <91rcig$rgu$1@nnrp1.deja.com>

Hi all.

I'm trying to find out if there's any interest in a kitchen-sink Python
module for mail handling.

I've got some code (for my company's internal product) that, with a
little bit of polishing, could handle most people's mail sending and
receiving needs in a pretty straightforward way.  Also I have not got
managerial clearance for releasing this code, as I'd like to see if
there's any interest before approaching them.

This module lets you send mail (using smtplib) and retrieve mail from
either a POP3 or IMAP4 server.  It handles encoding and decoding MIME
messages with attachments.  I wrote it because I needed something to
intelligently handle sending multipart MIME attachments.  It uses the
standard Python smtp/pop/imap/MIME modules, just basically ties them all
together.

Probable code license would be BSDish.  Post here or email me privately
at pibble at yahoo dot com if you are interested.

thanks
Preston

--
|| Preston Landers <prestonlanders at my-deja.com> ||


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


From u_schlaepfer at bluewin.ch  Sat Dec  9 12:22:06 2000
From: u_schlaepfer at bluewin.ch (Ueli Schlaepfer)
Date: 09 Dec 2000 18:22:06 +0100
Subject: wxPython & floating point numbers ?
Message-ID: <m2puj18q81.fsf@hobbes.dyn.dhs.org>

Hi everyone,

when using the otherwise excellent wxPython package, I observe the
following strange behaviour:

>>> print 0.1
0.1
>>> import wxPython.wx
>>> print 0.1
0.0

Version information:
Python 1.5.2 (#1, Nov  9 1999, 12:15:14)  [GCC egcs-2.91.66 19990314/Linux (egcs- on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> wxPython.wx.wxMAJOR_VERSION
2
>>> wxPython.wx.wxMINOR_VERSION
2
>>> wxPython.wx.wxRELEASE_NUMBER
2

Has anybody seen the same error?  What is the solution, if there is any?

Thanks,

Ueli


From fredrik at effbot.org  Wed Dec 13 12:27:51 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 13 Dec 2000 17:27:51 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <WntZ5.3566$Kd1.414873@newsb.telia.net> <918ath$15f$1@violet.singnet.com.sg>
Message-ID: <rkOZ5.3821$Kd1.435478@newsb.telia.net>

Ng Pheng Siong wrote:
> According to Fredrik Lundh <fredrik at effbot.org>:
> > Realize that there are more than one way to do it.
>                      ^^^
> "There are more than one way..."

http://www.dogma-2000.com/

</F>




From timr at probo.com  Wed Dec 20 01:53:02 2000
From: timr at probo.com (Tim Roberts)
Date: Tue, 19 Dec 2000 22:53:02 -0800
Subject: Walk through dictionary keys?
References: <3A3F8B89.1DD9E01F@schlund.de> <3A3F8E24.8040307@fenx.com>
Message-ID: <apl04ts8h4vl52cdv9tcbva2n1gcabtuqd@4ax.com>

Emile van Sebille <emile at fenx.com> wrote:

>I'm not sure when or how long this has been in, but the cvs version of 
>dict's has a popitem method that looks like it does what you want.
>
>d = {1:11,2:22,3:33,4:44}
>for i in xrange(len(d)):
>     print d.popitem()

But is popitem destructive?  That is, when that loop is complete, is d
empty?
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From pbjorn at uswest.net  Mon Dec 18 09:38:52 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Mon, 18 Dec 2000 07:38:52 -0700
Subject: while true: !!!
References: <mailman.977127721.4398.python-list@python.org> <slrn93rsl2.qb2.grey@teleute.rpglink.com>
Message-ID: <3A3E217C.5F2F4996@uswest.net>

Steve Lamb wrote:

> [snip]

> IMHO trying to cater to the novice at the expense of maintainability for
> the experts is not something to strive for in a language.  Introducing a
> design where a statement is contradictory is also not something to strive for.
> That is why we have "while 1:" because it is obvious to all who see it what it
> is beyond, say, the first day or so with the language and it doesn't have the
> problems "for ever" and "while true" impose.

Isn't catering to the novice what cp4e is/was all about <wink>? You're correct of course, that "while
1:" is pretty trivial to understand, and so is "x = x + 1", however neither is telling the reader
exactly what you're doing. The most common reason for using "while 1:" (at least in newspostings ;-) is
because saying what you _really_ mean <0.5 wink>:

   for line in f.readlines():
      ...

has the potential of crashing your machine with an out-of-memory error if you don't know the sizes of
files you're operating on. Using the fileinput module, or any of the other possible wrappers that have
been posted is just too slow. So, the status quo is that because the straight-forward way of doing it is
too slow, we're all using an optimized idiom that while easy to understand if coming from a CS
background, doesn't model what you're doing (at least not until you've programmed enough Python that
reading a file seems like an  endless loop <wink>).

-- bjorn



From arturo_perez at wanadoo.es  Wed Dec  6 14:55:25 2000
From: arturo_perez at wanadoo.es (Arturo Pérez Mulas)
Date: Wed, 06 Dec 2000 19:55:25 GMT
Subject: [DAO Database in win32] What is more efficient
References: <0g8X5.152$7U6.3922@m2newsread.uni2.es> <PXlX5.117656$U46.3885672@news1.sttls1.wa.home.com> <ssnX5.41$rZ6.2186@m2newsread.uni2.es> <3a2e8ce3.1084445@news.demon.co.uk>
Message-ID: <NQwX5.753$rZ6.21711@m2newsread.uni2.es>

Thanks for the hint. I will stick to ADO for the moment!

Arturo.

"Andy Robinson" <andy at reportlab.com> escribi? en el mensaje
news:3a2e8ce3.1084445 at news.demon.co.uk...
> "Arturo P?rez Mulas" <arturo_perez at wanadoo.es> wrote:
>
> >> If you aren't using ADO you should be. DAO was replaced by ADO some
time
> >> ago.
> >
> >By the way, is mxODBC any better?
> >
> >Thanks
> >
> If you just want to do simple queries, there is no need to upgrade. I
> have used both modules in production systems and had no trouble
> with either.   mxODBC is best if
> (a) you want to work a lot with dates and times,
> (b) if you want your Windows and Unix code to have EXACTLY the same
> database interface (you can build mxODBC for many database on Unix)
> or (c) if it is a high-load application and you want to be sure of
> using prepared statements.
>
> - Andy Robinson




From tim.one at home.com  Sun Dec 31 06:04:01 2000
From: tim.one at home.com (Tim Peters)
Date: Sun, 31 Dec 2000 06:04:01 -0500
Subject: why no "do : until"?
In-Reply-To: <POy36.102241$x6.45606433@news2.rdc2.tx.home.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEPJIFAA.tim.one@home.com>

To summarize 10 years of fruitless discussion on this issue:  yes,

    while 1:
        ...
        if finished:
            break

is the intended way to spell "do ... until" in Python.

"do ... until F" is better than "do ... while F" because the truth of F
conjoined with the loop invariant and loop precondition should imply the
loop postcondition, and the "while" form artificially negates the F of
interest, thus making reasoning harder.

"do ... while F" is better than "do ... until F" because the truth of F
conjoined with the loop invariant and loop precondition should imply that
the loop invariant still holds, and the "until" form artificially negates
the F of interest, thus making reasoning harder.

Python's indentation rules do indeed not fit well with either form.  If
there were an obvious way to make them fit well, Guido would have done so a
long time ago.  In the meantime, if you think you see one, write a PEP!

and-no-of-course-uninitialized-vrbls-exist-in-python-ly y'rs  - tim




From syver.enstad at sensewave.com  Tue Dec  5 18:32:04 2000
From: syver.enstad at sensewave.com (Syver Enstad)
Date: Wed, 6 Dec 2000 00:32:04 +0100
Subject: Global object in ASP (calling other scripting languages in ASP pages)
Message-ID: <90judk$edp$1@troll.powertech.no>

I tried calling dir() in on an ASP page and writing out the output.

The names I didn't know from before was 'ScriptingNamespace' and
'ASPGLOBALTLB'

If you qualify your function calls with ScriptingNamespace you get the
string of the foreign function instead of calling it, but at least the
function is visible to Python. If I could manage to call it too I'd be very
pleased.

Any ideas?






From costas at springmail.com  Tue Dec 12 19:54:49 2000
From: costas at springmail.com (Costas Menico)
Date: Wed, 13 Dec 2000 00:54:49 GMT
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu>
Message-ID: <3a36c85e.4409136@News.CIS.DFN.DE>


<snip>..........
>
>-- Russell
>
>Having said that Python is not really a great substitute for Smalltalk, 
>I will admit that I am working on a reasonable scale GUI project in 
>Python because even though I adore ST I have doubts about ST's long-term 
>viability (cincom seems to be back to ParcPlace's old tricks on insane 
>pricing) and it'd be an incredibly hard sell to my management. I may be 
>setting up for failure, due to Python's mediocre GUI, but I'm having a 
>go at it.

Cincom is not the only Smalltalk out there. What is wrong with the
other lower cost ones?

Costas


From grey at despair.rpglink.com  Tue Dec 19 19:45:05 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Wed, 20 Dec 2000 00:45:05 -0000
Subject: while true: !!!
References: <mailman.977127721.4398.python-list@python.org> <slrn93rsl2.qb2.grey@teleute.rpglink.com> <3A3E217C.5F2F4996@uswest.net> <slrn93sn65.tvg.grey@teleute.rpglink.com> <3A3F7336.36F2330C@uswest.net>
Message-ID: <slrn94008h.6qq.grey@teleute.rpglink.com>

On Tue, 19 Dec 2000 07:39:50 -0700, Bjorn Pettersen <pbjorn at uswest.net> wrote:
>Well, if you have it, and use it, you're hardly wasting it <wink>.

    Not a good attitude to have if one is programming in a highly shared
environment.  Considering the primary machine I program on is a utility box in
a large ISP that has dozens of processes running at once I prefer to be a
little more conservative with what I do in regards to memory.  As it stands I
had to oversee a process that took 300Mb at a pop.

>So you're arguing that since Python is so unreadable in it's use of else
>clauses, "while 1:" hardly matters? <wink>

    Nope, I'm pointing out that there are other portions of the language which
are far more confusing than while 1: and while 1: is hardly worth the time to
argue about except for those who are just looking for something to argue about
in their spare time.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From anders.eriksson at morateknikutveckling.se  Mon Dec 11 05:09:54 2000
From: anders.eriksson at morateknikutveckling.se (Anders M Eriksson)
Date: Mon, 11 Dec 2000 11:09:54 +0100
Subject: UML and Python
References: <90o91k$eks$1@desperado.vew-telnet.net> <90stth$mea$1@newsreaderg1.core.theplanet.net> <m3aea5te8t.fsf@wilma.localdomain>
Message-ID: <rj993tc5ig1kn6tcvovkajrgarpcgevubs@4ax.com>

On 09 Dec 2000 14:35:14 -0800, Harry George <hgg9140 at seanet.com>
wrote:

>You could also consider Dia.  It doesn't (yet) have a model repository

A link would be nice!

// Anders



From aahz at panix.com  Wed Dec 20 20:38:45 2000
From: aahz at panix.com (Aahz Maruch)
Date: 20 Dec 2000 17:38:45 -0800
Subject: How to create GUID UUID like with pythoncom.CreateGuid
References: <3a4136bb.2555949@192.168.0.1>
Message-ID: <91rmv5$d7v$1@panix3.panix.com>

In article <3a4136bb.2555949 at 192.168.0.1>,
Georg Gogo. BERNHARD <gogo at bluedynamics.com> wrote:
>
>The algorithm should be platform independend which means that  a
>python only solution is needed. I am starting to think that that is
>impossible without extending python with some C / C++ functions that
>have to be compiled on every platform they are needed, because the MAC
>adress of a compuiter is part of the GUID, and there is no way to get
>it in python.

I think it does need to be a C extension.  Microsoft is simply using the
OSF DCE GUID/UUID spec.  While there are ways of creating a GUID in the
absence of a MAC (for machines without a network card), I'd suggest that
it probably makes more sense to just rip off the standard algorithm if
you can find it.  There may be a GNU library somewhere.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

I don't really mind a person having the last whine, but I do mind
someone else having the last self-righteous whine.  --Aahz


From greg at cosc.canterbury.ac.nz  Thu Dec 14 20:44:24 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Fri, 15 Dec 2000 14:44:24 +1300
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <tcmay-FCA22F.14073713122000@news.got.net> <918v7a02scn@news1.newsguy.com> <tcmay-04E80C.16181313122000@news.got.net> <91a1ed02a4j@news2.newsguy.com> <3A389B1A.4F33FFD4@cfmu.eurocontrol.be> <91ae8s02llf@news2.newsguy.com>
Message-ID: <3A397778.2E6D84D3@cosc.canterbury.ac.nz>

Alex Martelli wrote:
> 
> I am not aware of any OOA approaches in which ... let 
> some objects be an instance of ... several classes
> at the same time.

Just to throw another language into the mix,
the TADS (Text Adventure Development System)
lets you declare objects with multiple
classes, e.g.

   bed: Furniture, Surface, Scenery
      ...

declares a singleton object called 'bed'
which belongs to the classes Furniture,
Surface and Scenery.

This is really just syntactic sugar, since
you could just as well declare a new class
inheriting from all those and then instantiate
it. But it's extremely handy in a language
designed for applications where the vast
majority of objects tend to be singletons.

To mention something vaguely Python-relevant:
if Python had a convenient syntax for declaring 
singletons, it could be a pretty good language for
Interactive Fiction...

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


From teyc at bigfoot.com  Thu Dec 14 00:18:15 2000
From: teyc at bigfoot.com (Chui Tey)
Date: Thu, 14 Dec 2000 15:18:15 +1000
Subject: Rummaging around inside an Access DB
References: <c5es2ts2l66bhg8qb7ne28cj0qa3gi9595@4ax.com> <hefs2t8tpsu3f9h746o84ns6sspa2p99dp@4ax.com>
Message-ID: <919l9t$tvk$1@bunyip.cc.uq.edu.au>

Check your installation of Access, the developer help is not installed by
default. There is a detailed description of the Access model, otherwise
you'd probably find it at http://msdn.microsoft.com/library/default.asp

Chui






From erno-news at erno.iki.fi  Tue Dec 26 14:44:23 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 26 Dec 2000 21:44:23 +0200
Subject: python telnet script for openvms host
References: <92ainp$jab$1@nnrp1.deja.com> <kuito7rrg7.fsf@lasipalatsi.fi> <92apa7$oac$1@nnrp1.deja.com> <D8D70FCFA8CC11C0.4BEE4AB5F2A3BCF7.EAD0DA106483FAE8@lp.airnews.net>
Message-ID: <kuelyvrmqw.fsf@lasipalatsi.fi>

Cameron Laird <claird at starbase.neosoft.com> writes:

| 1.  OpenVMS term negotiation is hairy--or at least
|     I always forget the magic after about a year,
|     and also where I put the answer the last time.
|     Based on my experience, I'd guess it *very*
|     likely that all your difficulties have to do
|     with that initial negotiation.

this is probably the problem then, looks like telnetlib
just refuses/denies all options the server proposes.

  -- erno


From grante at visi.com  Thu Dec 28 12:25:51 2000
From: grante at visi.com (Grant Edwards)
Date: Thu, 28 Dec 2000 17:25:51 GMT
Subject: Q: Is there any good documentation on MAPI using Python
References: <wpm06.791$AH6.82177@newsc.telia.net>
Message-ID: <zIK26.3737$4c.280470@ruti.visi.com>

In article <wpm06.791$AH6.82177 at newsc.telia.net>, Henrik M?rtensson wrote:

>I would like to use the MAPI with the MS office suite
>programming in Python, but I'm having a little bit of a problem
>with finding good documentation.
>
>Does anybody have any leads?

I'm using a Python program to fetch e-mail messages (and
associated attachements) via MAPI.  You're welcome to it:

  ftp://ftp.visi.com/users/grante/python/mfetch.py

-- 
Grant Edwards                   grante             Yow!  Do you think the
                                  at               "Monkees" should get gas on
                               visi.com            odd or even days?


From billy at nospamforme.com  Mon Dec 11 14:43:12 2000
From: billy at nospamforme.com (Billy)
Date: Mon, 11 Dec 2000 19:43:12 +0000
Subject: Control-C Stops normal output to the terminal
Message-ID: <c3aZ5.5741$ma1.122916@news2-win.server.ntlworld.com>

I've just done "My First Python Program"(tm) and I have a little odd
problem...
I'm calling an external program using os.system() and checking the
result to catch control-c's (and other abnormal termination).  When
the python program exits the shell no longer echo's output.
I can reset the terminal and get back to normal, but I'd rather have
it work "right"...
I'm using python 1.52 at the moment on a debian linux setup.
I've heard of problems with linux & readline - is this related to the
problems I'm hitting?
Oh, I can call the external program from a terminal and cntl-c it,
and I get a normal shell back...

Billy.


From joconnor at cybermesa.com  Thu Dec 21 14:57:33 2000
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Thu, 21 Dec 2000 19:57:33 +0000
Subject: A Suggestion for Python Colon Syntax
 
Message-ID: <3A4260AD.1AFCA0D3@cybermesa.com>

>It is really useful (needed) if you wanted to write 2 statements on 1
>line.  I don't know why that would be needed, but . . .

This is useful for small anonymous functions such as used for filter()
and map().

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

"God himself plays the bass strings first when He tunes the soul"



From fredrik at effbot.org  Fri Dec 29 18:48:11 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 29 Dec 2000 23:48:11 GMT
Subject: Interesting behaviour of the assignment
References: <92f4gk$qk6$2@news.nuri.net> <92fhsr01u88@news1.newsguy.com> <mailman.978059501.14500.python-list@python.org> <92j4n1$mhu$0@216.39.162.232> <92j7gm$6v2$1@panix2.panix.com>
Message-ID: <%o936.2527$Qb7.277358@newsb.telia.net>

Aahz Maruch wrote:
> If you drop "is", you can't check for object identity.

id(x) == id(y)

</F>




From cfelling at iae.nl  Sat Dec 16 14:27:53 2000
From: cfelling at iae.nl (Carel Fellinger)
Date: 16 Dec 2000 20:27:53 +0100
Subject: Solution: Direct access to Printer I/O lines
References: <917k9b$9i9$1@nnrp1.deja.com> <slrn93hrjn.2vp.ssthapa@localhost.localdomain> <91cnp3$g59$1@nnrp1.deja.com> <91d23f0jk7@news1.newsguy.com>
Message-ID: <91gfnp$7r4$1@animus.fel.iae.nl>

Alex Martelli <aleaxit at yahoo.com> wrote:
...
> This gives me a good chance for one of my favourite rants,
> '"Singleton" design-pattern considered harmful'!-)  But
> 'Singleton' is not being ideally applied, so I'll skip that.

And in an other thread you wet my appetite and direct me here!
So let us hear that rant, mister, and enlight us mere mortals
who still think that there is legitimate use of this singleton
pattern.
-- 
groetjes, carel


From ben.hutchings at roundpoint.com  Fri Dec 15 12:52:05 2000
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 15 Dec 2000 17:52:05 +0000
Subject: Capturing stdout from a console process?
References: <91b8q3$aqu$1@nnrp1.deja.com>
Message-ID: <uitolzi62.fsf@roundpoint.com>

Sam Schulenburg <samschul at pacbell.net> writes:

> I am using <os.system("xcopy c:*.* d: /v")> within a python program to
> copy files from one directory to another under windows NT. When the
> program executes, a new console window is created with the results of
> the xcopy command being displayed in the console window. Is their a way
> to capture this output by python so I can log error messages?

Yes, but you might find it easier to use something based on the source
for shutil.copytree (which is commented as being `an example rather
than a tool').

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From bryan at eevolved.com  Fri Dec 15 20:01:28 2000
From: bryan at eevolved.com (Bryan Mongeau)
Date: Sat, 16 Dec 2000 01:01:28 GMT
Subject: Bug in BSDDB?
Message-ID: <I9z_5.9426$0U2.369913@news20.bellglobal.com>

Greetings Python gurus,

I was doing some benchmarks of the dbhash module when I came across some 
unsettling behavior in Windows 98:

1 - When shelve creates a new database with the "c" flag, it can only add 
85652 keys to the dictionary before raising an error in bsddb.

2- If that error is caught and the write operation tried a second time, the 
key is written just fine.

NB. To reproduce the above behaviour, consult the attached file hashtest1.py


3- A dbhash file with 100000 simple records is 140 MB in Windows, compared 
to 21 MB in Unix. If you attempt to add more than 100000 keys in Windows, 
the size of the file quickly balloons to > 1 GIG. Keep in mind I was using 
unicode strings and cPickle.

4- An attempt to write more than 100000 keys will result in errors 
described in point 1, with ever-increasing frequency until eventually 
python itself will explode (illegal operation).

NB. To reproduce the above behaviour, consult the attached file hashtest2.py
*save your work first* ;-)


I hope someone can shed some light on the situation. I am aware of some 
bugs in bsddb, but none like this. ?It is important to note that the code 
in question works *beautifully* on Unix.

Any advice/replacement db module recommendations would be greatly 
appreciated. 
-- 
<==================================>
Bryan Mongeau
Lead Developer, Director
eEvolved Real-Time Technologies Inc.
www.eevolved.com
<==================================>

"Let us understand what our own selfish genes are up to, because we may 
then at least have a chance to upset their designs, something that no other 
species has ever aspired to do." -- Richard Dawkins

-------------- next part --------------
A non-text attachment was scrubbed...
Name: hashtest1.py
Type: application/x-python
Size: 712 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20001216/f380a140/attachment.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: hashtest2.py
Type: text/x-java
Size: 777 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20001216/f380a140/attachment.java>

From skidrow at nownuri.net  Sun Dec 10 06:02:24 2000
From: skidrow at nownuri.net (¼­Áö¿ø)
Date: Sun, 10 Dec 2000 20:02:24 +0900
Subject: How can I use printer port in python?...
Message-ID: <90vnol$l5p$1@news1-2.kornet.net>

As the title says, I was wondering if there's anyway I can

use printer port in python...




From zeitlin at seth.lpthe.jussieu.fr  Wed Dec 20 12:52:09 2000
From: zeitlin at seth.lpthe.jussieu.fr (Vadim Zeitlin)
Date: 20 Dec 2000 17:52:09 GMT
Subject: Using the fonts of different encodings in wxPython
References: <91nho6$hk2$1@news.nuri.net>
Message-ID: <slrn941se9.eui.zeitlin@seth.lpthe.jussieu.fr>

On Tue, 19 Dec 2000 20:44:10 +0900, June Kim <junaftnoon at nospamplzyahoo.com> wrote:
>This problem is of wxPython 2.2.2 and Python 2.0 on Win98.
>I'm using 8 bit character set (CJK) on windows. When I call the
>wxFontEnumerator, what I get is just a partial list of the fonts on my
>system -- only English fonts. Moreover, the wxPython doesn't seem to have
>wxFONTENCODING_xxxx that fits my system's character set. What can I do to
>use the fonts of different charater sets?

 Unfortunately, the support for Far East encodings is incomplete in wxWindows,
mainly because the people who implemented encoding support didn't know much
about them. It shouldn't be too complicated to make wxWindows support them but
we need someone who can either do it with our help or help us to do it (first
variant is preferred :-)

 If you're interested, please join wxwindows-devel mailing list.

 Regards,
VZ

-- 
GCS/GM d? H+ s++:-- p2 au--- a- w+ v C+++ UBLS+++ P- L++ N++ E--- W++++ M? V--
-po+ R++ G`` !tv b+++ D--- e++++ u++ h--- f+ r++ n- y?


From ldw at us.ibm.com  Thu Dec 14 10:19:29 2000
From: ldw at us.ibm.com (Larry Whitley)
Date: Thu, 14 Dec 2000 09:19:29 -0600
Subject: Deleting a stored file in a ZipFile
Message-ID: <91aoe1$1cno$1@news.rchland.ibm.com>

Looking over ZipFile, it looks like it's not possible to delete a file from
the archive.  True?

Larry




From matt at virtualspectator.com  Wed Dec 20 19:58:49 2000
From: matt at virtualspectator.com (matt)
Date: Thu, 21 Dec 2000 13:58:49 +1300
Subject: IP adresses are not unique enough.
In-Reply-To: <3a4148ad.7150467@192.168.0.1>
References: <3a413a17.3416402@192.168.0.1> <3A41475A.5DC4ABD8@san.rr.com> <3a4148ad.7150467@192.168.0.1>
Message-ID: <00122114180307.00886@localhost.localdomain>

As long as you never want to create the same one again.  There are tradeoffs
between using some persistent things to make an ID as opposed to involving a
dynamic compnent such as time.  For many applications it is nice to be able to
re-create the unique identifier rather than storing it somewhere.

Matt


On Thu, 21 Dec 2000, Georg Gogo. BERNHARD wrote:
> Well, the thing is that we want to use the mechanism with a ms-sql
> server and its "uniqueidentifier" datatype. This GUID is some sort of
> standard and is used within DCOM or RPC algorithms and seems to be
> very relyable. 
> 
> There is a big disadvantage if I use IP adresses since they can be
> "unrouteable" as you called, which could mean that they are within a
> LAN but using a firewall. The MAC adresses are unique, as I
> understood. The Windows CoGreateGUID COM API function combines a
> timestamp and a MAC adress. I want to do the same thing in python,
> which is maybe not possible.
> 
> Thank you for your help,
> Gogo.
> 
> 
> 
> 
> On Wed, 20 Dec 2000 23:57:14 GMT, Darren New <dnew at san.rr.com> wrote:
> 
> >> A plain python algorithm itself would be very useful, since one could
> >> use it for RPC / COM as well as for creating indexes for distributed
> >> databases.
> >
> >Well, concatenating one of the IP addresses assigned to the machine to the
> >clock would give you something that should be globally unique, if you had an
> >IP address that wasn't unroutable. Generate a random number to tack on the
> >end, and you're set?
> >
> >-- 
> >Darren New / Senior MTS & Free Radical / Invisible Worlds Inc.
> >San Diego, CA, USA (PST).  Cryptokeys on demand.
> >Personal malapropism generator free with purchase!
> >Steganography: The manual is hidden in the source code.
> >Gold mining is only profitable because of the gold found.
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list



From loewis at informatik.hu-berlin.de  Fri Dec  8 15:15:39 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 08 Dec 2000 21:15:39 +0100
Subject: import site error!!: in building python 2.0
References: <90m19p$lhr$1@nnrp1.deja.com>
Message-ID: <p6qlmtqk6tw.fsf@informatik.hu-berlin.de>

etsang at my-deja.com writes:

> What is going on???

Hard to say; it works for me. I recommend to use truss/strace to
analyse this further; site.py should be found in Lib/site.py.

Regards,
Martin



From grey at despair.rpglink.com  Sat Dec 16 11:45:07 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Sat, 16 Dec 2000 16:45:07 -0000
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <915jcf0c6o@news2.newsguy.com> <9162p3$7t2$1@animus.fel.iae.nl> <MBWZ5.334$h67.37748@newsread1.prod.itd.earthlink.net> <3A397C74.6C8906E1@cosc.canterbury.ac.nz> <91d5520m6j@news1.newsguy.com> <wvw_5.10652$h67.670217@newsread1.prod.itd.earthlink.net> <slrn93mh41.1nk.scarblac@pino.selwerd.nl>
Message-ID: <slrn93n70i.fuu.grey@teleute.rpglink.com>

On 16 Dec 2000 10:34:06 GMT, Remco Gerlich <scarblac at pino.selwerd.nl> wrote:
>Yes! And you don't need the ugly "break" statement, you can simply do
>"true = 0" and the loop will end! (Remember to set it back to 1 later)

    Am I the only one who doesn't think that "break" is no more ugly than
"continue"?  They are flow control, nothing more, nothing less, and a natural
part of the language.  I certainly find the former more pleasing than the
latter because it is how things should be

while 1:
  if false:
    break
  do something

true = 1
while true:
  if false:
    true = 0
  else
    do something

    But, hey, that might just be my perlism showing through since even though
I dislike constructs like "next if (true)" because I've had to rewrite too
many "next if ()"'s to "if () {next}" so I could put in other operations.  IE,
there was a practical reason other than someone's notion of asthetics
dictating the form and function.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From jasonic at nomadicsltd.com  Thu Dec 21 16:23:34 2000
From: jasonic at nomadicsltd.com (Jason Cunliffe)
Date: Thu, 21 Dec 2000 16:23:34 -0500
Subject: Reality check PYZOMA = python / zope for  Mac ?
Message-ID: <t44t20nmcgena7@corp.supernews.com>

Hello

I am wondering what the prospects are for Python and Zope on Macintosh:

Is the mac getting left behind or will the new MacOSX in fact bring Python
adn Zope more easily into the fold?

I am quite concerned that projects planned for Python and Zope will not be
in fact be usable on Macs.. What are the realistic prospects during the next
6-12 months do you think?

Does anybody even care anymore?

Thanks for any opinions and links.. brief searching for Python and has been
a frustrating and confusing experience


- Jason
___________________________________________________________
Jason CUNLIFFE = NOMADICS['Interactive Art and Technology']




From m.hadfield at niwa.cri.nz  Wed Dec  6 20:38:10 2000
From: m.hadfield at niwa.cri.nz (Mark Hadfield)
Date: Thu, 7 Dec 2000 14:38:10 +1300
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu>
Message-ID: <976153091.257544@clam-ext>

"Nathan Gray" <nospam at caltech.edu> wrote in message
news:3A2EE2AF.AD8A51E2 at caltech.edu...
> What a rude response to a polite and legitimate question.

Well I think a more legitimate question would have been:

"I want to know about Pythons origins. I've looked in the obvious places,
i.e. www.python.org, and read the stuff there but I want to know more about
this: ..."

---
Mark Hadfield
m.hadfield at niwa.cri.nz  http://katipo.niwa.cri.nz/~hadfield/
National Institute for Water and Atmospheric Research
PO Box 14-901, Wellington, New Zealand





From abrahams at mediaone.net  Sun Dec 10 19:33:44 2000
From: abrahams at mediaone.net (David Abrahams)
Date: Mon, 11 Dec 2000 00:33:44 GMT
Subject: [ANN] Boost C++ Python Library (formerly py_cpp)
Message-ID: <IhVY5.94094$751.2244355@typhoon.ne.mediaone.net>

I am proud to announce the re-release of py_cpp as the Boost Python Library
at www.boost.org.

Among its many features:

* Automatic function argument type checking and conversion
* Subclassing wrapped C++ types in Python
* [Member] Function overloading
* Automatic wrapping of numeric operators
* Pickling support
* Support for C++ smart pointers

For those who have been following py_cpp, significant improvements have
occurred since the last release. An excerpt from the release notes follows:


2000-11-22 10:00
    Ullrich fixed bug in operator_dispatcher<op_long>.

2000-11-21 10:00
    Changed all class and function names into lower_case.

    Ullrich updated documentation for operator wrapping.

2000-11-20 10:00
    Ullrich renamed ExtensionClass:register_coerce() into
    ExtensionClass:def_standard_coerce() and made it public

    Ullrich improved shared_pod_manager.

2000-11-17 15:04
    Changed allocation strategy of shared_pod_manager to make it portable.

    Added pickling support + tests thanks to "Ralf W. Grosse-Kunstleve"
    <rwgk at cci.lbl.gov>

    Added a specialization of Callback<const char*> to prevent unsafe usage.

    Fixed Ullrich's operator_dispatcher refcount bug

    Removed const char* return values from virtual functions in tests; that
    usage was unsafe.

    Ullrich changed Module::add() so that it steals a reference (fix of
refcount bug)

    Ullrich added operator_dispatcher::create() optimization

    Ullrich changed design and implementation of TypeObjectBase::enable()
(to eliminate low-level
    code) and added shared_pod_manager optimization.


2000-11-15 12:01
    Fixed refcount bugs in operator calls.

    Added callback_adjust_refcount(PyObject*, Type<T>) to account for
different ownership
    semantics of Callback's return types and Caller's arguments (which both
use from_python())
    This bug caused refcount errors during operator calls.

    Moved operator_dispatcher into extclass.cpp
    Gave it shared ownership of the objects it wraps

    Introduced sequence points in extension_class_coerce for
exception-safety

    UPPER_CASE_MACRO_NAMES

    MixedCase template type argument names

    Changed internal error reporting to use Python exceptions so we don't
force the
    user to link in iostreams code

    Changed error return value of call_cmp to -1

    Moved unwrap_* functions out of operator_dispatcher. This was
transitional: when
    I realized they didn't need to be declared in extclass.h I moved them
out, but
    now that operator_dispatcher itself is in extclass.cpp they could go
back in.

    Numerous formatting tweaks

    Updated the BoundFunction::create() optimization and enabled it so it
could actually be used!

2000-11-15 00:26

    Made Ullrich's operators support work with MSVC

    Cleaned up operators.h such that invalid define_operator<0> is no longer
needed.

    Ullrich created operators.h to support wrapping of C++ operators
(including the "__r*__" forms).
    He added several auxiliary classes to extclass.h and extclass.cpp (most
importantly,
    py::detail::operator_dispatcher and py::operators)

2000-11-13 22:29

    removed obsolete ExtensionClassFromPython for good.

    removed unused class ExtensionType forward declaration

2000-11-12 13:08

    Added enum_as_int_converters for easier enum wrapping

    Introduced new conversion namespace macros:
        PY_BEGIN_CONVERSION_NAMESPACE,
        PY_END_CONVERSION_NAMESPACE,
        PY_CONVERSION

    callback.h, gen_callback.py:
        Added call() function so that a regular python function (as opposed
to
         method or other function-as-attribute) can be called.

         Added newlines for readability.

    class_wrapper.h:
        Fixed a bug in add(), which allows non-method class attributes

    Ullrich has added def_raw for simple varargs and keyword support.

    Fixed version number check for __MWERKS__

    Added tests for enums and non-method class attributes

    objects.h/objects.cpp:
        Added py::String operator*= and operator* for repetition

        Change Dict::items(), keys(), and values() to return a List

        Added template versions of set_item, etc., methods so that users can
optionally
        use C++ types that have to_python() functions as parameters.

        Changed various Ptr by-value parameters to const Ptr&





From moshez at zadka.site.co.il  Thu Dec 21 01:30:27 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Thu, 21 Dec 2000 08:30:27 +0200 (IST)
Subject: real life examples
In-Reply-To: <u48q19.cth.ln@adsl.bapnt>
References: <u48q19.cth.ln@adsl.bapnt>
Message-ID: <20001221063027.DA379A84F@darjeeling.zadka.site.co.il>

On Wed, 20 Dec 2000 11:51:37 +0100, huber <nospam at wanted.com> wrote:

> i try to teach python to my son (13 years old)!
> but instead of stupid for loops, classes nobody ever needs and so on .....
> are there some good real life examples a child could be interested in ?

If there are daily-comic sites that he likes, have him write a program
that connects to one of them and downloads the picture only at 6am,
so he can look at it when he wakes up without waiting for it to
download or watching the advertisements.
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From aleaxit at yahoo.com  Sun Dec 24 10:45:11 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 24 Dec 2000 16:45:11 +0100
Subject: Speed of Python 1.5.x vs 2.0
References: <4vj54t4v0hvjsnstfmtmf36q9irotp59sj@4ax.com> <mailman.977475665.3482.python-list@python.org> <dud64ts8mb7mlc41kht410q5g87ekb2nc5@4ax.com> <91vfc6029r5@news2.newsguy.com> <mailman.977653989.1950.python-list@python.org>
Message-ID: <9255p80qjm@news1.newsguy.com>

"Peter Schneider-Kamp" <peter at schneider-kamp.de> wrote in message
news:mailman.977653989.1950.python-list at python.org...
> Hei Alex!
>
> If you used the code you posted, I'm not surprised by
> the results. In the first module you have:

[timing withMap both times!]

Yep, thanks -- that's why I corrected it a few hours ago:-).

Timing the right functions, comprehension is shown to be
slower than map by a solid 10%, with a simple loop using
neither faster than either.

The correct version & results are now posted in the thread
"One Python 2.1 idea" in a response to Russell Turpin.


Alex





From aahz at panix.com  Thu Dec 21 17:11:18 2000
From: aahz at panix.com (Aahz Maruch)
Date: 21 Dec 2000 14:11:18 -0800
Subject: A Suggestion for Python Colon Syntax
References: <91tc8h$2pp$1@hecate.umd.edu> <91tfg1$282$1@hecate.umd.edu> <EEr06.842$Qb7.73987@newsb.telia.net> <91trir$52l$1@hecate.umd.edu>
Message-ID: <91tv66$nqb$1@panix2.panix.com>

In article <91trir$52l$1 at hecate.umd.edu>,
William Djaja Tjokroaminata  <billtj at z.glue.umd.edu> wrote:
>
>I saw many replies that give the reasons why the colons are
>*needed*.  My suggestion is to make it *optional*, like the semicolon at
>the end of statements.
>
>Yes, for people who cares about making it easier to read for newbies, they
>can always put colons.  Yes, for people who use emacs to automatically
>indent the next line, they can put colons too.

Did you *read* the URL Fredrik gave you?  What part of forcing
programmers to write intelligible code do you not understand?
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

I don't really mind a person having the last whine, but I do mind
someone else having the last self-righteous whine.  --Aahz


From jepler at inetnebr.com  Mon Dec 18 22:41:55 2000
From: jepler at inetnebr.com (Jeff Epler)
Date: Tue, 19 Dec 2000 03:41:55 GMT
Subject: [ANN] xreadlines patch on sourceforge
Message-ID: <slrn93tm82.5rk.jepler@potty.housenet>

I've just uploaded my "xreadlines" patch to Sourceforge.  You can
retrieve it from the Patch Manager, and it is my hope that it will
be considered for inclusion in Python 2.1.

xreadlines, inspired by xrange, permits a simple 'for' loop to iterate
over the contents of a file without reading the entire file at once.

xreadlines is both a new extension module, as well as an additional
method on file objects.  It includes an entry for the test suite,
as well as a modified 'fileinput' module which has a >2x speed
improvement by using the xreadlines function.

example:

def wc(filename):
	l=w=c=0
	for line in open(filename).xreadlines():
		c=c+len(line)
		w=w+len(line.split())
		l=l+1
	return (l,w,c)


Jeff Epler
jepler at inetnebr.com


From loewis at informatik.hu-berlin.de  Sat Dec  9 05:54:11 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 09 Dec 2000 11:54:11 +0100
Subject: exec'ing an import
References: <3A315C1A.90CB8D98@seebelow.org>
Message-ID: <p6qpuj1amr0.fsf@informatik.hu-berlin.de>

Grant Griffin <not.this at seebelow.org> writes:

> Does anybody have an explanation?  

Perhaps script.py is not on sys.path, so import won't find it.

Regards,
Martin


From DavidL.nono.nospammy at raqia.com  Wed Dec  6 16:56:39 2000
From: DavidL.nono.nospammy at raqia.com (David Lees)
Date: Wed, 06 Dec 2000 16:56:39 -0500
Subject: Python: what all the hype is about?
References: <90locf$d80$1@nnrp1.deja.com>
Message-ID: <3A2EB617.4E32817D@raqia.com>

See the language architects thoughts on this at:
http://www.python.org/doc/essays/comparisons.html


david lees

aivkov at my-deja.com wrote:
> 
> Dear all,
> 
> Excuse my ingnorance in the subject area. Can someone explaing me very
> briefly, why one should choose Python over Java or is it better to
> compare it with Javascript/ECMA?
> 
> I hope I am not trying to compare apples with oranges.
> 
> Thanks,
> Anatoliy.
> 
> Sent via Deja.com http://www.deja.com/
> Before you buy.


From fredrik at effbot.org  Wed Dec  6 15:55:49 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 06 Dec 2000 20:55:49 GMT
Subject: CGI File Upload
References: <p8ce0tcgtt2bksb3adrf9cfr2i8egr4ff5@4ax.com> <8u7fn7$jpf$1@la-mail4.digilink.net> <90j10o$5p5$1@nnrp1.deja.com> <ulmttsihs.fsf@cs.uu.nl> <90m06a$kg1$1@nnrp1.deja.com>
Message-ID: <pJxX5.1851$Kd1.283467@newsb.telia.net>

Bill Seitz wrote:
> Yes, this turned out to be the issue. I had it set for stdout, but not
> stdin.
>
> msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
> msvcrt.setmode(sys.stdin.fileno(), os.O_BINARY)

an easier (and more portable) solution is to do what the
FAQ says, and run Python with the -u option:

    http://www.python.org/doc/FAQ.html#7.12

    "7.12. cgi.py (or other CGI programming) doesn't
    work sometimes on NT or win95!"

    Be sure you have the latest python.exe, that you are using
    python.exe rather than a GUI version of python and that you
    have configured the server to execute

         "...\python.exe -u ..."

    for the cgi execution. The -u (unbuffered) option on NT
    and win95 prevents the interpreter from altering newlines
    in the standard input and output. Without it post/multipart
    requests will seem to have the wrong length and binary
    (eg, GIF) responses may get garbled (resulting in, eg,
    a "broken image").

</F>






From rzilavec at tcn.net  Fri Dec 15 18:03:16 2000
From: rzilavec at tcn.net (Richard Zilavec)
Date: Fri, 15 Dec 2000 23:03:16 GMT
Subject: new with python..
References: <91dt4e$d3u$1@barni-1.tehnicom.net>
Message-ID: <3a3aa306.276820025@news.tcn.net>

On Fri, 15 Dec 2000 20:55:39 +0100, "Milos Hadzic" <milhad at ptt.yu>
wrote:

>i am new with this proggraming tool and i would appriciate if you would tell
>me where can i get a good tutorial.
>

I'm also new and have been reading through the tutorial at python.org.

--
 Richard Zilavec
 rzilavec at tcn.net


From emile at fenx.com  Fri Dec 29 15:22:24 2000
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 29 Dec 2000 12:22:24 -0800
Subject: Question re: eval()
References: <978109654.339714327@news.silcom.com> <92iivt$6uk0r$1@ID-11957.news.dfncis.de> <978118012.3281608@news.silcom.com>
Message-ID: <92irqp$7cf6o$1@ID-11957.news.dfncis.de>

I'm still not following.  Will the user be typing in
something that needs specifically to be eval'd?  If so, eval
won't take a UserDict or class.  You may be able to populate
a dictionary by determining the variables in the expression
(see sample below).  Or will it be (or could it be) a
properly formed SQL statement?  Or will they be providing
parameters for information to be accessed, such as a
parameter names and range values?

#--- for simple expressions

def getEvalVars(s, sgVars=None, slVars=None):
    if sgVars == None:
        sgVars={}
    if slVars == None:
        slVars={}
    stillErrors = 1
    newVars = []
    while stillErrors:
        try:
            eval(s, sgVars, slVars)
            stillErrors = 0
        except NameError, err:
            stillErrors = 1
            errDesc = str(err)
            missingVar = errDesc.split("'")[-2]
            slVars[missingVar] = 1
            newVars.append(missingVar)
        except:
            stillErrors = 0
            pass
    return newVars

print getEvalVars('these + names + are + missing')


--

Emile van Sebille
emile at fenx.com
-------------------


"Clarence Gardner" <clarence at netlojix.com> wrote in message
news:978118012.3281608 at news.silcom.com...
> On Fri, 29 Dec 2000, Emile van Sebille wrote:
> >What are you trying to do that using a dict doesn't
address?
>
> The lazy evaluation.  Not only will the values that I want
to supply
> require large database accesses to retrieve, but I don't
even know
> with what names they will be requested!
>
> Clarence Gardner
> clarence at netlojix.com




From fredrik at effbot.org  Mon Dec 11 14:37:40 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Mon, 11 Dec 2000 19:37:40 GMT
Subject: Python2 distutils question: how to best "autoconfig"...?
References: <9130e60obc@news2.newsguy.com>
Message-ID: <82aZ5.3458$Kd1.395204@newsb.telia.net>

Alex Martelli wrote:
> The specific issue is "how many bits of significance
> does a Python float have".  (Or is there somewhere in
> the Python headers a #define for this that I missed...?)

why not just do things on the Python side?  if not else,
it's a sure way to avoid optimization head-aches:

def get_base_and_digits():
    def precise(x):
        return x + 1.0 - x - 1.0 == 0.0
    a = b = c = 1.0
    base = digits = 0
    while precise(a):
        a = a + a
    while not base:
        b = b + b
        base = a + b - a
    while precise(c):
        digits = digits + 1
        c = c * base
    return base, digits

>>> print get_base_and_digits()
(2.0, 53)

(if anyone runs Python on a box that returns anything
but (2.0, 53), let me know!)

</F>




From ullrich at math.okstate.edu  Sun Dec 24 14:28:55 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Sun, 24 Dec 2000 19:28:55 GMT
Subject: <> vs != was Re: A Suggestion for Python Colon Syntax
References: <2kM06.120$Lm2.7518@news.clarkson.edu> <LNBBLJKPBEHFEDALKOLCKEDEIFAA.tim.one@home.com> <mailman.977594956.22001.python-list@python.org> <OR616.67063$x6.31192247@news2.rdc2.tx.home.com>
Message-ID: <925ipj$b3d$1@nnrp1.deja.com>

In article <OR616.67063$x6.31192247 at news2.rdc2.tx.home.com>,
  "Rainer Deyke" <root at rainerdeyke.com> wrote:
> "Thomas Wouters" <thomas at xs4all.net> wrote in message
> news:mailman.977594956.22001.python-list at python.org...
> > On Sat, Dec 23, 2000 at 12:15:25AM -0500, Tim Peters wrote:
> >
> > > Best guess is that "<>" will go away in P3K.  I jokingly told
Guido it
> was
> > > clearly wrong for an inequality operator to look so symmetric, and
he
> forgot
> > > to laugh <wink>.
> >
> > Well, excuse me, but it's clearly wrong for an equality test
operator to
> > look so much like an augmented assignment statement ! I definately
prefer
> > '<>' myself, anyway, I just wish I could use it in C without turning
> towards
> > preprocessor macros :)
>
> I dislike <> because to me A <> B implies A > B or A < B.  This does
not
> logically apply to some data types where no ordering is defined.

Well to _me_ A!=B says that A is not only equal to B, A
is _emphatically_ equal to B, no doubt about it.

>  Of
course
> in Python, A <> B (and A != B) does imply A > B or A < B because of
the way
> comparison is implemented.
>
> --
> Rainer Deyke (root at rainerdeyke.com)
> Shareware computer games           -           http://rainerdeyke.com
> "In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor
>
>

--
Oh, dejanews lets you add a sig - that's useful...


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


From sragsdale at my-deja.com  Tue Dec 19 14:00:23 2000
From: sragsdale at my-deja.com (sragsdale at my-deja.com)
Date: Tue, 19 Dec 2000 19:00:23 GMT
Subject: Swap Detector: resource.rusage problems
Message-ID: <91ob81$aod$1@nnrp1.deja.com>

I'm writing a wrapper for a very CPU and RAM hungry program.  Every so
often this program gets a little too hungry and I need to be warned
about it.

Is there any way to get resource statistics on a program WHILE it's
running?  I know that resource.getrusage() can be used to get resource
statistics after a program has finished, but at that point it's too late
for me.

Here's my sample programs that I've been messing around with.
consumer.py just consumes 10 seconds of CPU and increasing amounts of
RAM while reporting its usage, then it exits.  foo.py forks a process,
runs 'consumer' in the child and monitors resource usage in the parent.
It doesn't see the child's resource usage until the child exits.

ObSystemSetups:

1-Python 1.5.2 (#8, Jul 13 1999, 17:39:45) [C] on irix646
  IRIX64 bach 6.5 07201608 IP28

2-Python 1.5.2 (#1, Feb  1 2000, 16:32:16)  [GCC egcs-2.91.66
19990314/Linux (egcs- on linux-i386
  Linux fivefifty 2.2.15-3SGI_39smp #1 SMP Mon Jul 24 08:45:16 EDT 2000
i686 unknown

############### consumer.py #################
import time,resource

ltime = time.time()
foo = ['x'] * 10000
bar = foo[:]
ticks = 0
while 1:
    if time.time() > ltime + 1:
        ticks = ticks + 1
        bar.extend(foo)
        ltime = time.time()
        print "consumer>",resource.getrusage(resource.RUSAGE_SELF)
        print
        if ticks > 10:
            break

#################### foo.py ###################
import resource
import os,sys,time

pid = os.fork()

# original process
if pid:
    # monitor our usage and the use of the children
    while 1:
        # don't use much CPU to do it
        time.sleep(1)
        val = os.waitpid(pid,os.WNOHANG)
        print "1c>",resource.getrusage(resource.RUSAGE_CHILDREN)
        print "1s>",resource.getrusage(resource.RUSAGE_SELF)
        print
        if val[0]:
            print "Done.  Exiting."
            sys.exit(0)

# child process
else:
    os.system("python consumer.py")
    print "2c>",resource.getrusage(resource.RUSAGE_CHILDREN)
    print "2s>",resource.getrusage(resource.RUSAGE_SELF)
    os._exit(0)


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


From bartig at pinpoint.com  Mon Dec 11 14:21:39 2000
From: bartig at pinpoint.com (bartig at pinpoint.com)
Date: Mon, 11 Dec 2000 19:21:39 GMT
Subject: threads and dictionaries
Message-ID: <9139fq$ovj$1@nnrp1.deja.com>

I'm writing a program that passes dictionaries into a subroutine that's
threaded.  Are dictionaries thread safe?  If so how would I go about
making changes to the dictionaries inside of a thread so that other
threads can take advantage this?


Sent via Deja.com http://www.deja.com/
Before you buy.


From matthewm at ihug.co.nz  Sun Dec 10 23:10:37 2000
From: matthewm at ihug.co.nz (Matthew)
Date: Mon, 11 Dec 2000 17:10:37 +1300
Subject: Newbie Thread question
Message-ID: <3A3453BD.B9A54981@ihug.co.nz>

Hi Everyone,

I've got thread stuff working by sub-classing threading.Thread, but
haven't be able to figure out how to use them by calling a runnable
target. Something like:

dog1 = threading.Thread(None, Dog().bark())

I suspect I'm horribly off the mark with this one!

I checked the faq and a few other places and there seems to little to no
docs to get one started - any urls i've missed.

Any help really appreciated. Thanks, Matthew.



From cpr at emsoftware.com  Wed Dec 13 02:03:27 2000
From: cpr at emsoftware.com (Chris Ryland)
Date: Wed, 13 Dec 2000 07:03:27 GMT
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu>
Message-ID: <3bFZ5.382$Y2.6323@news>

Russell--

Overall, good show! Some questions/observations.

"Russell E. Owen" <owen at astroNOJNK.washington.edu.invalid> wrote in message
news:915qlh$89fa$1 at nntp6.u.washington.edu...
> [...]
> On the other hand, it is not as powerful, clean and regular as
> smalltalk. Limitations include:
> - The language is somewhat cluttered and irregular because it has
> procedures, functions and methods. Hence arbitrary decisions had to be
> made in some areas as to how to implement things, which can lead to some
> hunting through the manuals.

But, as we all know, languages that give you an "on ramp" from other more
familiar paradigms often succeed against "more clean/powerful" languages.
(E.g., Java was an easy "reach" for many C++ programmers.)

I.e., Smalltalk's fierce O-O cleanliness may work against it for general
acceptance. (That, and the fact that every Smalltalk vendor in the world
(all 2 of them ;-) fell down and played dead at the first appearance of the
(apparent) Java juggernaut.)

Python's familar procedures and functions make it easier for newbies, and
they're a very small step from methods ("just put them in a class, and add a
self parameter").

> [...]
> - Collection classes are good, but no match for Smalltalk. (But then
> what language comes close to ST in this area?). It does have
> dictionaries and lists, both of which are great. But that's about it. It
> also has "tuple", a minor and annoying (to me) variant of a list that is
> immutable (more or less) and so can be used as a key in dictionaries.

But surely someone could add all the ST collection classes to Python, or is
it that without blocks, they wouldn't be nearly as powerful?

> [...]
> - Rather ugly implementation of "package" (multi-file modules). This is
> one of the few areas where I find the implementation to be ugly and
> hastily done; generally it's a very clean language.
> - Like most languages, it uses text files to contain sets of code. I
> really miss Smalltalk's more integrated approach.

Or, rather, maybe this is one of the most tasteful choices in a language
implementation whose code objects must live in a traditional file system.
(See "on ramp" argument above.)

But I've often mused that it would be lovely to see a ST-like development
environment for Python (Zope, other than the fact that it has a very clunky
UI, comes close).
 --
Cheers!
Chris Ryland, President * Em Software, Inc. * www.emsoftware.com




From gherman at darwin.in-berlin.de  Thu Dec 14 07:57:49 2000
From: gherman at darwin.in-berlin.de (Dinu C. Gherman)
Date: Thu, 14 Dec 2000 13:57:49 +0100
Subject: EXIF images with PIL?
Message-ID: <3A38C3CD.F6625D9A@darwin.in-berlin.de>

Hello,

does anybody know if PIL supports EXIF images? EXIF(*) images
are basically like JPEG, but contain additional data that
usually stems from the generating device like digital cameras
and the like. 

I'd like to know if using PIL it is possible to extract the
EXIF data segments?

Thanks in advance!

Dinu

(*) http://www.sentex.net/~mwandel/jhead/exif-e.html

-- 
Dinu C. Gherman
................................................................
"The only possible values [for quality] are 'excellent' and 'in-
sanely excellent', depending on whether lives are at stake or 
not. Otherwise you don't enjoy your work, you don't work well, 
and the project goes down the drain." 
                    (Kent Beck, "Extreme Programming Explained")


From BgPorter at NOartlogicSPAM.com  Tue Dec 19 14:05:32 2000
From: BgPorter at NOartlogicSPAM.com (Brett g Porter)
Date: Tue, 19 Dec 2000 19:05:32 GMT
Subject: MIDI extensions (WIN32, Mac)?
References: <3A3FACEF.DAF77229@cableinet.co.uk>
Message-ID: <0kO%5.54693$w35.9313503@news1.rdc1.nj.home.com>

"Richard Dobson" <rwd at cableinet.co.uk> wrote in message
news:3A3FACEF.DAF77229 at cableinet.co.uk...
> Hello,
>
>
> I have just started looking at Python seriously, with a focus on music
> applications; e.g. algorithmic composition. Python seems to me to be
> idally suited as a first programming language for Music Technology
> students who have never done programming before. So far the net trawling
> I have done has failed to turn up any already-done extensions supporting
> MIDI i/o. I could probably code this myself at least for WIN32 if need
> be, but it is the sort of thing I feel sure somebody must already have
> done. Can anyone suggest anything? Or has it indeed not been done yet?
>
I'm doing some work on this, but the idea of getting anything approaching
usable timing has made me focus instead on writing MIDI files and using a
standalone sequencer for actual performance. MIDI timing under Win9x is bad
enough as it is -- the additional latencies of Python aren't going to
help...

It would be easy enough to wrap the Windows MIDI i/o calls, if they're not
already available through one of Mark Hammond's extensions...

--
//  Today's Oblique Strategy (? Brian Eno/Peter Schmidt):
//  Honor thy error as a hidden intention
//
//  Brett g Porter * BgPorter at acm.org **  http://members.home.com/bgporter





From pbjorn at uswest.net  Mon Dec 25 14:23:15 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Mon, 25 Dec 2000 12:23:15 -0700
Subject: Some Python 2.1 ideas
References: <mailman.977590932.19335.python-list@python.org> <Cw616.23054$qN6.3308205@typhoon.we.rr.com> <3A452881.7686E717@uswest.net> <%vb16.187233$U46.5801214@news1.sttls1.wa.home.com> <3A464D5D.B541C04E@uswest.net> <1XE16.190885$U46.5872195@news1.sttls1.wa.home.com>
Message-ID: <3A479EA3.7F4A08FE@uswest.net>

Greg Jorgensen wrote:

> [snip]

> I always use rstrip because I never care about trailing whitespace. If I did
> have a reason to preserve trailing spaces I would write my own code to strip
> the trailing newlines.

Good enough, but it's error prone to have everyone write their own version of
chomp:

    def chomp(s):
        if s.endswith('\r\n'): # dos
            return s[:-2]
        if s.endswith('\n'): # unix
            return s[:-1]
        if s.endswith('\r'): # mac
            return s[:-1]
        return s # no terminator

(I actually forgot the case where the string doesn't have a line terminator the
first time I wrote it...)

-- bjorn



From daniel at cs.washington.edu  Tue Dec 12 16:24:05 2000
From: daniel at cs.washington.edu (Daniel Wood)
Date: Tue, 12 Dec 2000 21:24:05 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <r0pzoi11lhz.fsf@black.cs.washington.edu> <5jvZ5.3588$Kd1.418770@newsb.telia.net> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com>
Message-ID: <r0pwvd51gga.fsf@black.cs.washington.edu>

2 would not be mutable.  Certainly.

I'm talking about an *object* that is both an int and mutable.

class MutableInt:
	def __init__( self, default ):
		self.value = default

	def set( self, val ):
		self.value = val

	def get( self ):
		return self.value

	# And maybe
	def __int__( self ):
		return self.value


My question is: am I confused if I think that this might be useful?
And if it is useful then what's the most pythonic way to do this.
(And if the best thing is a list of one element, why isn't that much
more "hack" than "idiom"?)

Thanks,
Daniel.

Darren New <dnew at san.rr.com> writes:

> Daniel Wood wrote:
> > Hmmm, I'm open to the suggestion that I need to rethink my plan at a
> > larger scale, but isn't a mutable int likely to sometimes be useful?
> 
> No. That's why people invented variables.
> 
> Having 
>  x = 2 + 2
>  print x
> result in anything except "4" is probably a *really* bad idea.
> AFAIK, only FORTH and really old Fortrans allow this.


From jschmitt at vmlabs.com  Sat Dec 16 20:24:49 2000
From: jschmitt at vmlabs.com (John Schmitt)
Date: Sat, 16 Dec 2000 17:24:49 -0800
Subject: while true: !!!
Message-ID: <008F0A63472BD311AF9800104BCD102561CA88@minirex.vmlabs.com>

I'm enjoying this thread quite a bit.  But I am a little surprised that this
hasn't turned into a structured programming debate.  I fully expected people
to bring up the issue that a loop should have exactly one entry point and
exactly one exit point and that those points should be exactly the beginning
or the end of the loop.  That's the kind of stuff I was taught when I was
learning to program.  Has that gone out of fashion in the 6 years I've been
out of school?  Is my memory of these ideas wrong?

I don't feel qualified to be suggesting new syntax to Python, but I don't
think two kinds of loops would be unpythonic in anyway:

while condition:
	stuff

and the complement of that would be

do:
	stuff
while condition

I think being able to test at the end of the loop is pretty convenient and
would help out the situations where you have a primer to the loop.

break and continue and goto are all exactly in the same class with exactly
the same prestige/stigma/usefulness, IMNSHO.  



From dean.hall at computer.org  Fri Dec  1 17:13:00 2000
From: dean.hall at computer.org (Dean Hall)
Date: Fri, 01 Dec 2000 22:13:00 -0000
Subject: Gatspy (GA module) available on sourceforge
Message-ID: <9097pc+pspc@eGroups.com>

I have put Gatspy, a Genetic Algorithm framework module (with docs!), 
on sourceforge:
    http://sourceforge.net/projects/gatspy

Gatspy evolved out of my thesis work.  The name orginally comes from 
"a Genetic Alrgorithm Training System in PYthon."  Since it was used 
to evolve the weights of a neural network.

You can use Gatspy to solve many types of problems where numbers, 
bits, or character strings are the solution.  If you feel loopy, give 
it a whirl, and send me some feedback.

!!Dean





From sdm7g at virginia.edu  Tue Dec 19 14:41:03 2000
From: sdm7g at virginia.edu (Steven D. Majewski)
Date: Tue, 19 Dec 2000 14:41:03 -0500 (EST)
Subject: Python interpreter in python?
In-Reply-To: <20001219134253.B15468@westsidecnc.rh.rit.edu>
Message-ID: <Pine.NXT.4.21.0012191430020.324-100000@localhost.virginia.edu>

On Tue, 19 Dec 2000, D-Man wrote:

> 
> The problem with such a thing is that to run your interpreter you
> would need an existing interpreter (written in C/Java or other
> language).  The only languages that can be implemented in themselves
> are compiled languages like C and C++.
> 

That is absolutely untrue!

Lisp and Scheme are frequently implemented in themselves -- a 
'metacircular interpreter' is a fairly typical student exercise. 

I don't know of a Python-in-Python, but I don't know of any reason
you couldn't do one. 

The only question is: how deeply do you want to emulate it? i.e.
are you allowed to use the existing parser module or do you want
to write one from scratch? 

-- Steve Majewski <sdm7g @ Virginia.EDU>






From wware at world.std.com  Tue Dec 12 07:00:32 2000
From: wware at world.std.com (Will Ware)
Date: Tue, 12 Dec 2000 12:00:32 GMT
Subject: Q: Print whithout \n
References: <aboZ5.3538$Kd1.410322@newsb.telia.net>
Message-ID: <G5GEow.IEx@world.std.com>

Henrik M?rtensson (henrik.martensson at frontconsulting.se) wrote:
> How do you print a string whithout a linefeed in the end.

Use a comma at the end, which which will give you a space
instead of a linefeed. For instance,

print "These two things",
print "will appear on the same line"

will print

These two things will appear on the same line

-- 
# - - - - - - - - - - - - - - - - - - - - - - - -
# Resistance is futile. Capacitance is efficacious.
# Will Ware	email:    wware @ world.std.com


From erno-news at erno.iki.fi  Sun Dec 17 11:25:53 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 17 Dec 2000 18:25:53 +0200
Subject: determining how much data is on a socket
References: <91gu7q$6pj$1@bob.news.rcn.net>
Message-ID: <kun1dv3tgu.fsf@lasipalatsi.fi>

In article <91gu7q$6pj$1 at bob.news.rcn.net>, "Z 3 Penguin"
<z3penguin at penguinpowered.com> writes:

[reformatted - please keep your lines under 73 characters long]

| is there any way to find out how much data is waiting on a given
| socket?
[...]
| i would like to know if the bytes are waiting before i read, so that
| if they are not, my program can continue and try to read it later.
| any ideas?

you can't do this portably. most unixish systems support the
FIONREAD ioctl on sockets which does this.

the normal way to deal with this situation is to use select and/or
non-blocking sockets.

  -- erno


From gustafl at algonet.se  Fri Dec 22 00:16:42 2000
From: gustafl at algonet.se (Gustaf Liljegren)
Date: 22 Dec 2000 05:16:42 GMT
Subject: How use urllib.urlopen()?
Message-ID: <9012313F4gustaflalgonetse@194.213.69.148>

Can't get the urlopen() function to work. I have a script on a UNIX web 
server that should go to another webpage when the user clicks a button. To 
isolate the problem, I made a test script:

#!/usr/local/bin/python

import urllib
u = urllib.urlopen("http://www.yahoo.com")
print u.read()

When running this I get HTTP 500 - Internal server error. :-( 

What's missing?

Gustaf


From anderson at boi.hp.com  Tue Dec 12 15:20:35 2000
From: anderson at boi.hp.com (Bill Anderson)
Date: Tue, 12 Dec 2000 13:20:35 -0700
Subject: NewBie question: cannot import TKinter and os
References: <90m7fc$r6m$1@nnrp1.deja.com> <p6qitouk6ow.fsf@informatik.hu-berlin.de>
Message-ID: <3A368893.FDC9CAC7@boi.hp.com>

Martin von Loewis wrote:
> 
> etsang at my-deja.com writes:
> 
> > 1. at the very first beginning, it says import site error. I know this
> > module site is no longer required in Python 2.0 after dereading some
> > doc in pytho.org. how can I make thie error go away??
> 
> That is incorrect. Where did you read that? site.py is still needed.

Perhaps from the site.py included in 2.0?
"""
Append module search paths for third-party packages to sys.path.

****************************************************************
* This module is automatically imported during initialization. *
****************************************************************

In earlier versions of Python (up to 1.5a3), scripts or modules that
needed to use site-specific modules would place ``import site''
somewhere near the top of their code.  Because of the automatic
import, this is no longer necessary (but code that does it still
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
works).

"""
(emphasis mine)

I, too am having problems importing site.py and os, at least Zope is
anyway. I can do it just fine from the interpeter... it seems that not
allcode that does it still works. :(

-- 
Bill Anderson               Linux Specialist
Modular Network Storage     R&D           
Random Quote:
    Portable: survives system reboot.


From fredrik at effbot.org  Tue Dec 12 13:12:30 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 12 Dec 2000 18:12:30 GMT
Subject: [Q] Are Exceptions used that much in practice?
References: <3A35D546.A4731B5B@enteract.com> <914mfd02ibt@news2.newsguy.com> <915p3q$p1i$1@nnrp1.deja.com>
Message-ID: <iUtZ5.3573$Kd1.415992@newsb.telia.net>

David C. Ullrich wrote:
>     Python and Delphi both have try-except and try-finally
> but no try-except-finally. I once saw the person responsible
> for this in Delphi explain why there was no try-except-finally,
> in reply to a complaint about its non-existence. I'm not going
> to try to convince you he was right, but at least in Delphi
> it wasn't just an oversight, they considered it and finally
> decided against it.

quoting myself from a recent post:

$ cd Python-2.0
$ more Misc/HISTORY

    ...

    Release 0.9.6 (6 Apr 1992)

    ...

    New features in 0.9.6:
        - stricter try stmt syntax: cannot mix except and
        finally clauses on 1 try

    ....

for more info, see e.g.
http://www.deja.com/getdoc.xp?AN=644277741

</F>




From tyler at tylereaves.com  Thu Dec 21 22:53:32 2000
From: tyler at tylereaves.com (Tyler Eaves)
Date: Thu, 21 Dec 2000 22:53:32 -0500
Subject: Speed of Python 1.5.x vs 2.0
Message-ID: <4vj54t4v0hvjsnstfmtmf36q9irotp59sj@4ax.com>

	I'm soon going to attemp to talk my webhost into upgrading to
2.0, but I figure the only way I'm going to be able to make that
happen is if I can show results that show 2.0 being faster / more
effcient than 1.5.x. Their servers run on Red Hat 7.

Thanks for any help.

---
Tyler Eaves

Visit Ultra Coaster Central!
The internet's largest repository
of Ultra Coaster Tracks!

http://www.ultracoastercentral.com


-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----


From roy at panix.com  Fri Dec 29 10:49:12 2000
From: roy at panix.com (Roy Smith)
Date: Fri, 29 Dec 2000 10:49:12 -0500
Subject: Great New Book
References: <t4o1vr8hktk531@corp.supernews.com> <4534AFCCD1BB9D46.A2643D7A58D216F5.CCC99D2604D602BA@lp.airnews.net>
Message-ID: <roy-F534EE.10491229122000@news.panix.com>

claird at starbase.neosoft.com (Cameron Laird) wrote:
> Is "deceptively clear" a good thing?

I havn't seen the book, but I would interpret "deceptively clear" to 
mean, "He makes it sound so simple, you don't even realize how much 
stuff you're learning when you read it".

A rare quality in technical authors.


From gb at a3design.de  Fri Dec  8 02:44:17 2000
From: gb at a3design.de (gunter )
Date: Fri, 08 Dec 2000 07:44:17 -0000
Subject: fetchone vs. fetchmany with pygres...
Message-ID: <90q3gh+c4am@eGroups.com>

Hello there!

I am using pygres (postgres-python-interface).

There are two functions for cursors:

cursor.fetchone(): returns one result-row.
cursor.fetchmany(): returns all result-rows.

the problem:
multiple calls of "fetchone" return the same data every call. (It
seems to
be written for "non-table-returns" only)
the call of "fetchmany" returns a list of all rows. This is ok but
imagine
a very large return-set. You have the whole return-set in the memory
instead of every line one by one.

Is there a solution with something like "fetchone" but with every call
to
get the next result-row?

yours
gunter






From phrxy at csv.warwick.ac.uk  Wed Dec 13 12:40:19 2000
From: phrxy at csv.warwick.ac.uk (John J. Lee)
Date: Wed, 13 Dec 2000 17:40:19 +0000
Subject: unpacking Python tuple in C
In-Reply-To: <917t88$fi3@micky.ibh-dd.de>
References: <917e64$3bp@micky.ibh-dd.de> <m31yvc38hm.fsf@atrus.jesus.cam.ac.uk> <917t88$fi3@micky.ibh-dd.de>
Message-ID: <Pine.SOL.4.30.0012131710310.28239-100000@mimosa.csv.warwick.ac.uk>

On Wed, 13 Dec 2000, Mike M?ller wrote:
> >Stepping back from your immediate problem, what are you actually
> >trying to do?  I can infer it from what you've said here, I'm afraid.
>
> I am writing my program in Python (it's a water quality model). There is a
> "main" model written by somebody else in FORTRAN which calls  submodels as
> DLLs. My Python model is one of these submodels. Now I try to speeze
> everything into a DLL. So C is the "glue" and just has to pass some args
> back and force from the main to the submodel. I do NOT program any
> functionallity in C (just the argument passing via the DLL). Besides this it
> seems I need to improve my knowledge of C.

Things that may be of interest: Numeric Python of course, Pyfort, FPIG,
SWIG.


John



From lg at rgz.ru  Wed Dec 20 07:12:57 2000
From: lg at rgz.ru (lg at rgz.ru)
Date: 20 Dec 2000 15:02:57 +0250
Subject: real life examples
References: <u48q19.cth.ln@adsl.bapnt>
Message-ID: <3a409ff1@news.edunet.ru>

huber <nospam at wanted.com> wrote:
> i try to teach python to my son (13 years old)!
> but instead of stupid for loops, classes nobody ever needs and so on .....
> are there some good real life examples a child could be interested in ?
in my 13 I was real fan of recursion technology
try to explain him that python can _really_ helps with solving 
problems .. such as math prlems or somethink else, let he try to do some homework
using python 

-- 
zev


From reichowa at tcfreenet.org  Wed Dec 13 00:16:17 2000
From: reichowa at tcfreenet.org (Aaron Jon Reichow)
Date: Tue, 12 Dec 2000 23:16:17 -0600
Subject: Smalltalk and Python
In-Reply-To: <kuzoi15d60.fsf@lasipalatsi.fi>
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <kuzoi15d60.fsf@lasipalatsi.fi>
Message-ID: <Pine.BSF.4.10.10012122310510.48551-100000@tcfreenet.org>

On 13 Dec 2000, Erno Kuusela wrote:

> the thing that struck me as a little foreign was the apparent
> isolation of the smalltalk image from the rest of the operating system
> and the filesystem. i want to be able to grep code!

While I'm not a proponent of Smalltalk isolationism, grep is a useless
tool (more or less), simply an artifact, ye, a symptom of the
sickness known as file-based development.

> but i guess this is just an artifact of that particular
> implementation. (it was the only free one that i came across however)

A lot of Smalltalks have better integration with the host OS, but Squeak
is also the most portable, and runs on the most platforms.  

> | - Mediocre IDE. I hope this will be improved over time, along with the 
> | GUI support.
> 
> emacs!

Emacs (even Xemacs, with OO-browser) is a poor IDE.  It's little more than
an editor.  Sure, you have fancy syntax-highlighting, a news/email/web
client, but not the dynamic, incredibly integrated environment that are
common to all Smalltalks, less GNU Smalltalk and Little Smalltalk.

> | It's an absolutely stunning language for free, and although not as fast 
> | as Perl
> 
> (it's as fast in my experience)

According to some benchmarks I've seen, Squeak is faster. ;P  (although it
doesn't feel like it sometimes, with Morphic)

Aaron

Aaron Reichow                   :: "In essence, Smalltalk is a programming
Twin Ports ACM Vice President   ::  language focused on human beings rather
http://www.d.umn.edu/~reic0024/ ::  than the computer." - Alan Knight




From snowfox at yipyap.net  Sat Dec  2 11:51:29 2000
From: snowfox at yipyap.net (Brian McGroarty)
Date: Sat, 2 Dec 2000 10:51:29 -0600
Subject: [OT] Apple IIe help please
In-Reply-To: <20001201212507.A14226@westsidecnc.rh.rit.edu>; from dsh8290@rit.edu on Fri, Dec 01, 2000 at 09:25:07PM -0500
References: <20001201212507.A14226@westsidecnc.rh.rit.edu>
Message-ID: <20001202105129.B2371@vixen.yipyap.net>

On Fri, Dec 01, 2000 at 09:25:07PM -0500, D-Man wrote:
> 
> Hi all.  I apologize for the off topic-ness of the message and the cross-post. 
> I have a friend who has a program she likes (written in BASIC) on an Apple IIe. 
> I have a way to get into the code and list it on the screen.  What I am looking
> for is someone who has had some experience using an Apple IIe who can tell me
> how I can get the listing to be redirected to the printer port.  Also if it is
> possible to get a copy of the program file onto a disk for an IBM compatible
> comptuer (Windows or Linux) that would be great.

I think PR #1 sounds right for routing output to the printer, as
someone else said.

For the transfer, your best option is probably to use a modem for the
transfer - you'll need to set up a host that can receive X/Y or
Z-modem and take it from there.

Alternatively, if you can find someone with a working 3.5" floppy for
a //c or ][gs (did they make one for the ][e ever?) there is software
available for writing MS-DOS format files, and you can take it from
there. Mind, you may want to find a way to convert the BASIC code to a
non-tokenized format on the Apple or it's not going to be too useful
to you on the PC.




From maxm at normik.dk  Mon Dec 11 08:37:36 2000
From: maxm at normik.dk (=?iso-8859-1?Q?Max_M=F8ller_Rasmussen?=)
Date: Mon, 11 Dec 2000 14:37:36 +0100
Subject: REPOST:SQLlike tables in XML via Python
Message-ID: <7BD10B680501D411B9DF009027E06F32197B4E@exchange>

I have send this befor but apparantly it didn't get through

Maybe somebody on the list can help me. I spend hours the other day
searching the web, but without any luck.

I remember reading a year ago about using a subset of XML to make SQL-like
tables, saving the data in xml format.

There was an api and code examples, and I seem to remember that there was
Python examples too.

Does anybody have a link??? or an idea what i'm talking about.

Regards Max M



From fredrik at effbot.org  Thu Dec  7 15:29:59 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Thu, 07 Dec 2000 20:29:59 GMT
Subject: Simple HTTP client bug?
References: <3A2FEE2A.D1A0A645@raqia.com> <hnSX5.2109$Kd1.301769@newsb.telia.net>
Message-ID: <brSX5.2110$Kd1.301499@newsb.telia.net>

> just curious, but what's Programming in Python with Medusa
> and the Async Sockets Library"?

never mind, I found it:

http://www.nightmare.com/medusa/programming.html

</F>




From moshez at zadka.site.co.il  Tue Dec  5 23:59:15 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Wed,  6 Dec 2000 06:59:15 +0200 (IST)
Subject: Mutual import?
Message-ID: <20001206045915.8B109A818@darjeeling.zadka.site.co.il>

On Mon, 04 Dec 2000 13:15:33 +0000, Dale Strickland-Clark 
<dale at out-think.NOSPAMco.uk> wrote:
> Is it considered safe and acceptable for two modules to import each other or 
> are there likely to be unwelcome side effects of that?
> 
> The problem is that both require access to values and routines in the other.

There shouldn't be any real problem, but usually, when I do that, I find
that it usually tries to point to a problem you have in your design.
Perhaps even something as simple as there should be a third module
which they both import.
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From gregj at pobox.com  Wed Dec 20 01:49:18 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Wed, 20 Dec 2000 06:49:18 GMT
Subject: while (thread.alive()):
References: <mailman.977283904.26696.python-list@python.org>
Message-ID: <ODY%5.172281$U46.5447371@news1.sttls1.wa.home.com>

"Jay O'Connor" <joconnor at cybermesa.com> wrote in message
news:mailman.977283904.26696.python-list at python.org...

> Why are we still maintaining these leftover idioms such as using
> integers as booleans.  If we are to carry the state of the art of
> programming beyond it's current level, we need to be programming a t a
> higher level of abstraction than is easily supported while still
> dragging anachronisms such as this around.
>
> Those who claim that 'while 1:' is obvious, it's not.
> It's only obvious if you've programmed "C" and other such languages.
> It's makes no sense really, unless you've learned the 'code

No one claimed "while 1" is obvious, only that it's a idiom of the language
(and idiom shared more or less with C and C++). It's something you figure
out once in your programming career, and then it's "obvious" forever after.
I don't think that the complaints about "while 1" spring from confusion
about what it means, but a mistaken aesthetic offense at Python's idiomatic
"while 1: ... if condition: break ..." loop. Other languages handle this in
their own way, and have their own idioms, but they are no more or less
"obvious" than Python. Blaming C and C++ isn't fair; in those languages
expressions have a value, so richer looping expressions are possible and
"while (1)..." is less common.

Very little of any programming language is obvious until you've learned the
syntax and the idioms. Just because Ada and Smalltalk have boolean types
doesn't make Ada or Smalltalk code any more obvious to someone who doesn't
know the language

> My first professional language was Ada83.  From there I went to
> Smalltalk.  When I finally started working with languages such as C,
> C++, etc... I was at first confused ( while (1)...what does that mean)
> and then revolted.
>
> I'm continuously amused that people try to take language design forward
> and still drag obsolete ideas along for the ride.

I am on safe ground pointing out that C and C++ have enjoyed somewhat
greater success and wider adoption than Ada and Smalltalk, so perhaps those
obsolete ideas aren't so bad after all.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From fgeiger at datec.at  Thu Dec 28 14:28:52 2000
From: fgeiger at datec.at (Franz GEIGER)
Date: Thu, 28 Dec 2000 20:28:52 +0100
Subject: Language localization?
Message-ID: <92g4ev$5v3$1@newsreaderm1.core.theplanet.net>

Regardless whether I run scripts within a console or within PythonWin, the
umlauts (? = ae, ? = oe, ? = ue) and other special chars are not displayed
correctly.
E.g. \n seems to be "translated" to \012.
My Win NT 4.0 Svr is English speaking, but the regionals are set to German.

I use Python now for almost half a year and always displayed English dialogs
(until now I always could work around the \012-problem  somehow
with strip() and replace() ). But now I have to display German dialogs,
where I have to use umlauts. So this grows serious now.

Any hints?

Thanks in advance and
Best regards
Franz GEIGER





From s2mdalle at titan.vcu.edu  Wed Dec 20 23:14:42 2000
From: s2mdalle at titan.vcu.edu (David Allen)
Date: Thu, 21 Dec 2000 09:14:42 +0500
Subject: A Suggestion for Python Colon Syntax
References: <91tc8h$2pp$1@hecate.umd.edu>
Message-ID: <91tld7$nf3$1@bob.news.rcn.net>

In article <91tc8h$2pp$1 at hecate.umd.edu>, "William Djaja Tjokroaminata"
<billtj at y.glue.umd.edu> wrote:
> My suggestion is as follows:
> 
> 1) If the if (or for, etc.) clause is followed by a block, then the
> colon
> is optional (for backward compatibility):
> 
>     if condition is true
>         statement_1 statement_2
>         ....
> 
> 2) If the if (or for, etc.) clause is followed only by a single
> statement,
> then the colon can be used to write a single line of code:
> 
>     if condition is true: statement_1 next_statement
>     ....
> 
> Any comments, anyone?

Me, I always want the colon to be there for one 
simple reason:  python-mode.el and emacs knows that
when I have a : on the end of a line, the next line
should be indented, and it does it for me.  :)

Sure, that's not a *language* reason for wanting the
colon there, but it's a reason.  :)

-- 
David Allen
http://opop.nols.com/
----------------------------------------
Q: What can a goose do, a duck can't, and a lawyer should?
A: Stick his bill up his ass.


From robin at alldunn.com  Mon Dec  4 18:00:45 2000
From: robin at alldunn.com (Robin Dunn)
Date: Mon, 4 Dec 2000 15:00:45 -0800
Subject: Python+Emacs+GUD ?
References: <yfmn1eciew4.fsf@collet.i-did-not-set--mail-host-address--so-shoot-me>
Message-ID: <uwVW5.71$1yD3.11337832@news.randori.com>

<luthi at vaw.baug.ethz.ch> wrote in message
news:yfmn1eciew4.fsf at collet.i-did-not-set--mail-host-address--so-shoot-me...
> According to the Python archives this question has been posted 6 years
> ago... still without an answer?
>
> Q: Has someone accomplished Python integration within Emacs GDB-Mode (GUD,
>    Grand Unified Debugger)?
>
> With this, Emacs Python-mode mode would be a very handy integrated
development
> environment.
>

The current gud.el already supports it.  Mine has this in it:

(defgroup gud nil
  "Grand Unified Debugger mode for gdb and other debuggers under Emacs.
Supported debuggers include gdb, sdb, dbx, xdb, perldb, pdb (Python), and
jdb."
  :group 'unix
  :group 'tools)

Works great.  You just need to have a executable on your PATH named pdb.  It
can be as simple as a symlink to pdb.py.

--
Robin Dunn
Software Craftsman
robin at AllDunn.com
http://wxPython.org     Java give you jitters?
http://wxPROs.com        Relax with wxPython!





From Randy.L.Kemp at motorola.com  Thu Dec 28 13:59:56 2000
From: Randy.L.Kemp at motorola.com (Kemp Randy-W18971)
Date: Thu, 28 Dec 2000 12:59:56 -0600
Subject: Python vs. PHP (& Java?)
Message-ID: <69BC4C6AE83ED311BC9400805FA7B14AF84D36@il93exp01.css.mot.com>

A gentleman after my own image.  Some mornings I drink tea, other mornings coffee, and occasionally, espresso.  Some evenings I take a shot of whiskey, but other evenings, I indulge in a shot of Sherry, or a glass of beer.  It's all OK, as long as I can distinguish between my mornings and evenings.

-----Original Message-----
From: davidw at prosa.it [mailto:davidw at prosa.it]
Sent: Thursday, December 28, 2000 9:46 AM
To: python-list at python.org
Subject: Re: Python vs PHP (& Java?)


"Scott Hathaway" <slhath at home.com> writes:

> Also, I have fallen in love with python, but I would not use it for
> server-side web scripting.  PHP is very easy and extremely powerful
> for this task.  Why not use both?

Because you are using Tcl?;-)

-- 
David N. Welton
     Personal:           http://www.efn.org/~davidw/  
Free Software:           http://people.debian.org/~davidw/
   Apache Tcl:           http://tcl.apache.org



From offer at sgi.com  Tue Dec 26 16:37:43 2000
From: offer at sgi.com (Nhi Vanye)
Date: 26 Dec 2000 21:37:43 GMT
Subject: mail processing module?
References: <91rcig$rgu$1@nnrp1.deja.com> <m266kd45m0.fsf@freak.kaiserty.com> <91vumo$di3$1@nnrp1.deja.com> <92ap6g$o7n$1@nnrp1.deja.com>
Message-ID: <92b337$h50ik$1@fido.engr.sgi.com>

$ from prestonlanders at my-deja.com -#118967 | sed "1,$s/^/> /"
>
>
>In article <91vumo$di3$1 at nnrp1.deja.com>,
>  Preston Landers <prestonlanders at my-deja.com> wrote:
>
>I just want to see if people have had a chance to glance at this, the
>current interface to the maillib draft and tell me a good reason why
>it's horribly broken, or need some tweaking, or is just fine.  I've
>gotten tenative permission to post code, but I need to sort out a few
>more details.  I want to see if I can get some preliminary feedback on
>the module interface.  I will be making some tweaks to make it fully
>agnostic of transport mechanisms, but my basic plan is still to unify
>all the lower level modules into one maillib module (with several
>internal helper classes...)

You have a fetch_first_msg() API, is that all ?

What about non POP/IMAP mailboxes (does the API work for those ? ).

How does the API scale (particularly for unix mailboxes where you don't
really want to have to keep the entire box in memory if you can avoid it).

Does this API make sense as a wrapper around the c-client code ? (C-client
has a syntax for specifying which 'host/port/mailbox' to use), if you
were to pick up c-clients syntax for naming mailboxes that would help
any future work.

>
>thanks
>Preston


richard.
-- 
Richard Offer          Widget FAQ --> http://reality.sgi.com/widgetFAQ/
{X,Motif,Trust} on {Irix,Linux}
__________________________________________http://reality.sgi.com/offer/



From Daniel.Kinnaer at AdValvas.be  Tue Dec 26 18:20:43 2000
From: Daniel.Kinnaer at AdValvas.be (Daniel)
Date: Tue, 26 Dec 2000 23:20:43 GMT
Subject: saving NNTP bodies to disk
Message-ID: <3a4925f7.51675327@news.skynet.be>

Hello. I've come up with a small Python script but ran into trouble.
Here's what I'm trying to do : get a NNTP connection, get the last 10
headers AND bodies and save them to disk.  It seems I made an error: I
try to write a list to disk using f.write() and Python complains about
this with following report 

 f.write(TheBody[i])
TypeError: read-only character buffer, list

Below is the script. Can someone please explain me what I'm doing
wrong?  All help is greatly appreciated.  Thanks

print "Start script..."
from nntplib import *
print "### Ok! Starting to read from Skynet..."
s=NNTP('news.skynet.be')
resp, count, first,last,name = s.group('comp.lang.python')
print 'Group', name, 'has', count, 'articles, range', first, 'to',
last
resp, subs = s.xhdr('subject', first + '-' + last)

print "showing and saving only last 10 headers + bodies..."
for id, sub in subs[-10:]:
    print id, sub
    #save body to disk
    f=open( id,'w')      #how do I add an extension 'txt' ???
    TheBody=s.body(id)    
    f.write(sub)
    for i in range(0,len(TheBody)):
        print i,TheBody[i] #show on screen
        f.write(TheBody[i])
    f.close()
s.quit()
print "End of Script..."


best regards,

Daniel


From erno-news at erno.iki.fi  Mon Dec 25 10:34:48 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 25 Dec 2000 17:34:48 +0200
Subject: classes (was Re: Same again please for OOP)
References: <mailman.977623149.8184.python-list@python.org> <9257mf0sl1@news1.newsguy.com> <roy-E5B925.13324224122000@news.panix.com> <925khi019hs@news1.newsguy.com> <kubsu1t81x.fsf@lasipalatsi.fi> <92708d0g4g@news1.newsguy.com>
Message-ID: <kuy9x4seef.fsf@lasipalatsi.fi>

>>>>> "Alex" == Alex Martelli <aleaxit at yahoo.com> writes:

|| i don't understand what is wrong with this - i like making them all
|| methods, so i don't have to remember if it's an attribute or method
|| later. also overrinding __getattr__ later, if it turns out to be
|| needed, is tedious and ugly.

| "Tedious and ugly"?!  Just add inheritance from the usual mixin:

| class AccessorsMixin:
|     def __getattr__(self, name):
|         if name.startswith('__'): raise AttributeError,name
|         return getattr(self, '__get_'+name)()

you forgot __setattr__, it's more fun.

but ok, after these, it's merely ugly :). tastes too much like
operator overloading to me.

| and implement method __get_location(self) the same way as you'd
| implement location(self) in your style.  What could be easier?  If
| you already have a __getattr__, it's equally OK to add this couple
| of lines to it, or delegate to this one (delegating has pluses).

| Giving accessor methods uniform names is hardly a minus, and
| starting them with something suggesting 'get' is particularly good.

even easier than __get_location(self) is get_location(self) ;)

i rarely already have a __getattr__, and from what i've seen,
99% of classes written by other people don't either.

| As for 'what is wrong' with hiding state access behind methods (most
| of which don't actually DO anything but fetch state), a first
| example is: obj.price *= 1.07 # 7% increase versus
| obj.setPrice(obj.getPrice() * 1.07) # 7% increase

| Why force client code to go through the latter, more complicated
| form, when the simpler first form is available so cheap?

because, the latter is simpler. yes, it is more tedious to type. i
count this against augmented assignment :)

| Why force client-code to know and remember which attributes
| are actually 'in' the original object (and thus must be accessed
| through accessor methods, in your style) and which ones aren't (and
| thus must be accessed directly, unless one goes to the serious
| trouble of building up an accessor method and sticking it on
| on-the-fly -- blah!)...?

interesting point. i've never done this, seems like dubious practice
at first sight.

| These are pragmatical considerations that make direct, simpler
| notation preferable.  On a _conceptual_ plane, the key issue is not
| confusing state and behavior.  "Methods BEHAVE -- they DO stuff" and
| "properties JUST ARE -- I can 'read' them, and some of them I may
| also be allowed to 'write'" is a good pair of rules-of-thumbs to use
| as guidelines in deciding how to model something -- method, or
| property (attribute)?  It's a very similar distinction to that
| between functions and variables.

so, you'd prefer most of the string methods that take no args in
python 2.0 were actually attributes? i don't know, somehow
doesn't seem pythonic to me.

[rest snipped]

  -- erno



From lists at TheWebsons.com  Sat Dec 23 09:31:38 2000
From: lists at TheWebsons.com (Ben Ocean)
Date: Sat, 23 Dec 2000 06:31:38 -0800
Subject: Why Won't This Work?
Message-ID: <LOBBJMHOIFOLFIPFANHJKEFGCFAA.lists@TheWebsons.com>

Hi;
Because I'm building Zope, I'm trying to install Python-1.5.2. I've
successfully run:
./configure --with-thread
make
and everything appears to be in order. But when I type
python
at a command prompt, I get a bash error. What have I not done?
TIA,
BenO




From eliotm at pacbell.net  Wed Dec 13 15:33:53 2000
From: eliotm at pacbell.net (Eliot Miranda)
Date: Wed, 13 Dec 2000 12:33:53 -0800
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <3a36c85e.4409136@News.CIS.DFN.DE> <owen_nospam-1312001131590001@cs204-90.fsmodem.washington.edu>
Message-ID: <3A37DD31.C9A5F18C@pacbell.net>


"Russell E. Owen" wrote:
> 
> In article <3a36c85e.4409136 at News.CIS.DFN.DE>, costas at springmail.com (Costas Menico) wrote:
> 
> ><snip>..........
> >>
> >>-- Russell
> >>
> >>Having said that Python is not really a great substitute for Smalltalk,
> >>I will admit that I am working on a reasonable scale GUI project in
> >>Python because even though I adore ST I have doubts about ST's long-term
> >>viability (cincom seems to be back to ParcPlace's old tricks on insane
> >>pricing) and it'd be an incredibly hard sell to my management. I may be
> >>setting up for failure, due to Python's mediocre GUI, but I'm having a
> >>go at it.
> >
> >Cincom is not the only Smalltalk out there. What is wrong with the
> >other lower cost ones?
> 
> I need something that is:
> -- reasonably well supported
> - likely to be around for awhile
> - runs on Mac, unix and possibly PC
> - can write GUI code with reasonably native look and feel
> 
> I would also prefer a language that is fairly clean and not evolving very quickly (which I dearly hope Squeak eventually becomes).
> 
> I don't think there is a Smalltalk that fits these requirements other than Cincom, is there?
> - Squeak runs on everything but its GUI stuff is slow and uses a rooted main window, making it pretty hopeless for native feel, at least on Mac
> - IBM and Dolphin are mature but only runs on PC and (IBM) unix
> - I tried QKS years ago but got tired of it always being in beta, always promises for a real release some year.
> 
> Also, as I said, it'd be a tough fight. The conservatism in astronomy is astonishing. If it's not C or Perl then it's going to be a struggle to put it through. We're already using Python and selling it as a "better Perl", but for me to push Smalltalk there has to be an obvious first-class implementation that we can afford and will obviously do what we need. Cincom is too expensive, and given their pricing policies I expect it to continue to be a language with limited public knowledge and acceptance.

I would think that astronomy in the state university system would count
as non-commercial use.  However, I'm not speaking for Cincom here,
rather as an ex British University lecturer.  I'd go ahead and use the
non-commercial system (is got everything in it the commercial one has)
and then see if Cincom sues.  Suing is only worthwhile if the party
being sued actually has enough money to cover costs, so that's going to
prevent Cincom suing a University astronomy department immediately,
right?

> 
> -- Russell

-- 
_______________,,,^..^,,,____________________________
Eliot Miranda              Smalltalk - Scene not herd


From glenfant at equod.com.nospam  Tue Dec  5 05:59:56 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Tue, 5 Dec 2000 11:59:56 +0100
Subject: Graphs from python
References: <90gse2$14uo$1@news.rchland.ibm.com>
Message-ID: <90ihje$hmb$1@reader1.imaginet.fr>

Hi,

You can use reportlab package to draw your graphs in a PDF file. It's use is
identical whatever OS you use (I did not test BeOS or Macintosh but it's
supposed to work too).
reportlab provides a low level API to the PDF document but making classes
for various charts should not be that difficult.
Get reportlab (free) from http://reportlab.com

"Larry Whitley" <ldw at us.ibm.com> a ?crit dans le message news:
90gse2$14uo$1 at news.rchland.ibm.com...
> The simulation models I write and run generate a lot of data and to make
> sense of it all I put the data into a set of graphs.  The graphs are the
> typical spread sheet fare, bar graphs, X/Y graphs, scatter plots, etc.
> Today, I read the data into spreadsheets from the spreadsheet and create
my
> graphs by hand and I'd like to automate the process.
>
> I'm looking at using win32com (by Mark Hammond) to drive excel or
> piddle+graphite to generage graphs directly from python.  While I'm sure
it
> will get the job done, the excel object model is a bit overwhelming.
Piddle
> by itself isn't at the right level.  Graphite appears get to the right
level
> but the pre-alpha rating makes me think that it's a bit early to jump in.
>
> Are there other approachs I should consider?  Whatever I end up with, I'd
> like it to work from either Unix or Windows.  If I can't have both, I can
> make do with just windows.  Of the two approaches above, is one
> significantly better than the other (by whatever measure you want to
> choose)?
>
> Thanks,
> Larry
>
>



From joconnor at mail.cybermesa.com  Mon Dec  4 18:51:04 2000
From: joconnor at mail.cybermesa.com (Jay O'Connor)
Date: Mon, 04 Dec 2000 23:51:04 GMT
Subject: Random string
Message-ID: <200012042351.QAA24052@mail.cybermesa.com>

An embedded and charset-unspecified text was scrubbed...
Name: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20001204/aca503ad/attachment.ksh>

From victor at prodigy.net  Thu Dec 28 11:07:07 2000
From: victor at prodigy.net (Victor Muslin)
Date: Thu, 28 Dec 2000 16:07:07 GMT
Subject: Python with Apache on WinNT newbie question
Message-ID: <3a4b6391.4061375@localhost>

To get Apache server to run a Python CGI script, the script needs to
have on its first line a directive telling Apache the file name of the
Python interpreter, like so:

	#!d:/apps/python/python.exe
	#
	# The rest of my cgi script follows
	# ....

The problem with this is deploying this script on different machines
that may have Python interpreter installed in different directories.
Surely others had run into the same issue. What is a typical solution?
Unfortunately, if this directive can take environmental variables I
could not figure out how to do it.

Thanks.


From mwh21 at cam.ac.uk  Thu Dec  7 05:09:46 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 07 Dec 2000 10:09:46 +0000
Subject: OT: _Fifth Business_ (was Re: Python Origins)
References: <90k3pq$50p$1@nnrp1.deja.com> <slrn92tu7t.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com> <90n1b0$4jr$1@panix3.panix.com> <5hGX5.68$E3.1197@uchinews> <90ncs0$mt5$1@panix2.panix.com>
Message-ID: <m3lmts4k5x.fsf@atrus.jesus.cam.ac.uk>

aahz at panix.com (Aahz Maruch) writes:

> In article <5hGX5.68$E3.1197 at uchinews>,
> Ben Wolfson <wolfson at midway.uchicago.edu> wrote:
> >In article <90n1b0$4jr$1 at panix3.panix.com>, Aahz Maruch <aahz at panix.com> wrote:
> >>
> >>What's _Fifth Business_?  (I did a quick search on Google and Amazon and
> >>didn't find anything related to you.)
> >
> >_Fifth Business_ is the first book in the Deptford trilogy by Robertson
> >Davies, a Canadian and a man with a very impressive beard (the other
> >books being _Manticore_ and _World of Wonders_).
> 
> <confused look>  But what does that have to do with Andrew?

http://www.amk.ca/davies/

maybe?

No, I'd never heard of him either.

Cheers,
M.

PS: I'd forgotten how one of c.l.p's anti-flame measures was to take
an even slightly inflammatory subject and start talking about
something else.  It's nice to be reminded occasionally :-)

-- 
  I would hereby duly point you at the website for the current pedal
  powered submarine world underwater speed record, except I've lost
  the URL.                                         -- Callas, cam.misc


From dale at out-think.NOSPAMco.uk  Mon Dec  4 08:47:53 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Mon, 04 Dec 2000 13:47:53 +0000
Subject: General default exception handler
References: <lj0n2to9vs8bmv3oo49mqgoc6h7esh8t37@4ax.com> <NZMW5.10915$lF4.165334@e420r-atl1.usenetserver.com>
Message-ID: <138n2tojuquau637pm7obq8c4gvtl7hutd@4ax.com>

"Steve Holden" <sholden at holdenweb.com> wrote:

>Dale Strickland-Clark <dale at out-think.NOSPAMco.uk> wrote in message
>news:lj0n2to9vs8bmv3oo49mqgoc6h7esh8t37 at 4ax.com...
>> We have a substantial set of COM objects in Python that run a Web site.
>>
>> Currently, if an exception is raised, the user gets an unfriendly message
>in their browser and we
>> may never know about it.
>>
>> I want to log all unhandled exceptions but I'd rather not have to put a
>try/except block around
>> every entry point - because there are dozens of them.
>>
>> Is there a general way to catch all unhandled exceptions so I can log them
>and pass back something
>> nice and soothing to the user?
>>
>> I'm planning to use the traceback module to record the exception info.
>>
>> Thanks
>> --
>> Dale Strickland-Clark
>> Out-Think Ltd
>> Business Technology Consultants
>>
>If you're prepared to accept URLs which use an extended path to indicate
>arguments, one poosibility is to direct all requests to a "framework page"
>which then calls the required Python functionality and traps any errors it
>raises.
>
>See, for example,
>
>    http://www.webtechniques.com/archives/1998/02/kuchling/
>
>in which Andrew Kuchling describes the basic technique.  This seems to give
>you what you want.
>
>regards
> Steve
>
>
Interesting idea but that won't work in this case. I have to trap the error inside the COM object or
much of the detail is lost.

--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants




From bvdpoel at uniserve.com  Mon Dec 25 13:09:28 2000
From: bvdpoel at uniserve.com (Bob van der Poel)
Date: Mon, 25 Dec 2000 11:09:28 -0700
Subject: adding fields to picked structures
Message-ID: <3A478D58.F1B2ABA@uniserve.com>

I've got a little database which I read/write using cpickle. Now, I want
to add a field to my database. The old one might be:

	class Mydata:
		def __init__(self, ...)
			foo=''
			bar=''
		...

Now, a list of Mydata's has been saved to a file.

Later, I decide that I need to add another field, say 'newdata'. So, I
modify the class definition and try to read my data.

So, far all is good...until I try to access 'newdata'. If it's a new
record (not read in by pickle) all is okay, but if I try to access
'newdata' in an existing record I get

	TypeError: attribute-less object (assign or del)

So, I suppose I could write a conversion utility...but, I'd like to
avoid that if possible. Am I missing something...or is this just the way
life should be?


-- 
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bvdpoel at uniserve.com
WWW:   http://users.uniserve.com/~bvdpoel



From grey at despair.rpglink.com  Mon Dec 18 18:27:59 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Mon, 18 Dec 2000 23:27:59 -0000
Subject: while true: !!!
References: <mailman.977180341.10206.python-list@python.org>
Message-ID: <slrn93t7bv.le.grey@teleute.rpglink.com>

On Mon, 18 Dec 2000 23:32:23 +0100 (MET), Carsten Geckeler <nospam at no.spam>
wrote:
>`while true' would be more readable and clear than `while 1'.  By the way,
>explain a newbie that -1 is true. ;)

    0 is true, all else is false.  And?  :P

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From henry_crun3583 at my-deja.com  Sun Dec 31 16:12:52 2000
From: henry_crun3583 at my-deja.com (henry_crun3583 at my-deja.com)
Date: Sun, 31 Dec 2000 21:12:52 GMT
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <t4t9qv73vvvfff@corp.supernews.com> <slrn94u2o4.fkv.qrczak@qrnik.zagroda>
Message-ID: <92o7gh$nd2$1@nnrp1.deja.com>

I can't really see a problem with any of the suggested indents.

> > do:
> >     R=Themistor.ReadResistance()
> >     T=R_to_Temperature(R)
> > until (T>=25)

Is neater but I don't have any trouble reading (b)

> > do:
> >     R=Themistor.ReadResistance()
> >     T=R_to_Temperature(R)
> >     until (T>=25)

if thats what the parser likes. But why can't the parser ingnore extra
indents, in the same way C,Pascal ignore blocks added purely for
clarity? If you want some extra indent, add it.

> > do:
> >       R=Themistor.ReadResistance()
> >       T=R_to_Temperature(R)
> >     until (T>=25)

Tim Peters raises the question of do: while vs do: until

Personally I think "do: until" more often reflects the algorithm I am
representing, however "do: while" leads to a single consistent test for
both ("while:" & "do:while") loops, and can reduce bugs, especially
when you decide to change a loop from one to the other and forget to
fix the terminate condition. (I am a VERY vague programmer)



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


From nospam at wanted.com  Wed Dec 20 05:51:37 2000
From: nospam at wanted.com (huber)
Date: Wed, 20 Dec 2000 11:51:37 +0100
Subject: real life examples
Message-ID: <u48q19.cth.ln@adsl.bapnt>

i try to teach python to my son (13 years old)!
but instead of stupid for loops, classes nobody ever needs and so on .....
are there some good real life examples a child could be interested in ?

thanks
udo 


From donn at u.washington.edu  Fri Dec 29 12:37:27 2000
From: donn at u.washington.edu (Donn Cave)
Date: 29 Dec 2000 17:37:27 GMT
Subject: Need help with string method
References: <92if8v$iu1$1@nnrp1.deja.com>
Message-ID: <92ii4n$j2q$1@nntp6.u.washington.edu>

Quoth jvanar1 at my-deja.com:
| I have a pretty dumb question here.  I can't seem to get any string
| methods working from *within* a script.  here is a super-simple example
| that yields an error
|
| >>import sys
| >>import string
| >>theString = 'This is a string'
| >>print theString.find('string')
|
| now this works fine from the python command line but gives an error
| when run as a script. Here is the error message:
|
| Traceback (innermost last):
|   File "./clean", line 18, in ?
|     print theString.find('string')
| AttributeError: 'string' object has no attribute 'find'
|
| Am I missing an 'import' somewhere.  Thanks for any help...

   $ python -c 's = "hi"; print s.find("i")'
   1
   $ python1.5 -c 's = "hi"; print s.find("i")'
   Traceback (innermost last):
     File "<string>", line 1, in ?
   AttributeError: 'string' object has no attribute 'find'
   $

According to my theory, you're getting two interpreters, 2.0 when
you simply invoke python and version 1.5 when you run a file.

I don't know why.  If you invoke your file directly, then the
interpreter specified in the #! top line is probably not the
first python found in your PATH list.

	Donn Cave, donn at u.washington.edu


From chris at araidesign.com  Wed Dec 20 19:38:15 2000
From: chris at araidesign.com (Chris Arai)
Date: Wed, 20 Dec 2000 16:38:15 -0800
Subject: Need Help sorting alphabetically.
References: <3A3F1A91.A7ED9182@araidesign.com> <90109F230duncanrcpcouk@194.238.50.13>
Message-ID: <3A4150F7.A0E4B08E@araidesign.com>

Hello All!

Thanks for all the efforts to read my mind.  Especially Alex who provided a
nice methodical approach, and Duncan who actually read my mind the best.

The purpose of all this was to be able to sort a directory and then be able
to use HTMLgen to create SeriesDocs.  I modified Duncan's approach slightly
thinking that it might optimize it ever so slightly (not really an issue in
my case.)  Here is what I actually ended up using:

import os,string,re

def make_key(astring,pattern=re.compile(r'(\d+)|([a-z]+)')):
    # Used [a-z]+ because \D+ matched "A_"
    # Put lower() here so I could use [a-z] instead of [a-zA-Z],
    # and only call lower() once
    vals = pattern.split(string.lower(astring))
    res = []
    for i in range(1, len(vals), 3):
        if vals[i]:
            res.append(int(vals[i]))
        else:
            res.append(vals[i+1])
    return res

def alphasort(list):
    l1, res = [], []
    for v in list: l1.append((make_key(v), v))
    l1.sort()
    for k, v in l1: res.append(v)
    return res


if __name__=='__main__':

    # My real goal was to sort a directory "alphabetically"
    # (whatever I meant by that ;-))
    m=os.listdir('.')

    s=alphasort(m)
    print "%20s   %20s"% ('Sorted','Unsorted')
    for i in range(len(s)):
        print "%20s   %20s"%(s[i],m[i])

----end snip----
ouput:

bash-2.02$ python alphasort.py
              Sorted               Unsorted
        alphasort.py           globtest.py~
       alphasort.py~          makethumb.py~
      #globtest.py~#           makethumb.py
        globtest.py~                outtest
        makethumb.py           Pg66_100.GIF
       makethumb.py~           Pg71_300.GIF
             outtest         Pg71_A_300.gif
            outtest2               outtest2
         Pg1_100.GIF         #globtest.py~#
         Pg2_100.GIF          alphasort.py~
         Pg9_100.GIF           alphasort.py
        Pg17_100.GIF           Pg90_100.GIF
        Pg20_100.GIF           Pg17_100.GIF
        Pg66_100.GIF            Pg2_100.GIF
        Pg71_300.GIF           Pg20_100.GIF
      Pg71_A_300.gif            Pg9_100.GIF
        Pg90_100.GIF            Pg1_100.GIF


Thanks again,

Chris

Duncan Booth wrote:

> chris at araidesign.com (Chris Arai) wrote in
> <3A3F1A91.A7ED9182 at araidesign.com>:
>
> >Hi,
> >
> >This seems like it should be obvious, but it is evading me!
> >
> >I would like to sort alphabetically (I'm not sure that is the correct
> >term) so that strings of alpha numerics sort alphabetically instead of
> >by ASCII order.
> >
> >eg:
> >>>> m=['Pg95_100.GIF',
> >'Pg9_100.GIF','Pg95A_100.gif','Pg9A_100.gif','Pg10_100.gif']
> >>>> m.sort()
> >>>> m
> >['Pg10_100.gif', 'Pg95A_100.gif', 'Pg95_100.GIF', 'Pg9A_100.gif',
> >'Pg9_100.GIF']
> >
> >
> >I would like the order to be
> >['Pg9_100.GIF', 'Pg9A_100.gif', 'Pg10_100.gif, 'Pg95_100.GIF',
> >Pg95A_100.gif' ]
> >
> >What is the obvious answer??
> >
> Try this:
> ---- begin cut ---
> import re
> from string import lower
>
> def make_key(value, pattern = re.compile('(\d+)|(\D+)')):
>     vals = pattern.split(value)
>     res = []
>     for i in range(1, len(vals), 3):
>         if vals[i]:
>             res.append(int(vals[i]))
>         else:
>             res.append(lower(vals[i+1]))
>     return res
>
> def cleversort(list):
>     l1, res = [], []
>     for v in list: l1.append((make_key(v), v))
>     l1.sort()
>     for k, v in l1: res.append(v)
>     return res
>
> m=['Pg95_100.GIF',
> 'Pg9_100.GIF','Pg95A_100.gif','Pg9A_100.gif','Pg10_100.gif']
>
> print cleversort(m)
>
> ----- end cut ----
> The output is:
> ['Pg9_100.GIF', 'Pg9A_100.gif', 'Pg10_100.gif', 'Pg95_100.GIF',
> 'Pg95A_100.gif'] as you requested.
>
> This sorts each sequence of digits as its numeric value, and each
> sequence of non digits as its lowercase equivalent (which I think is what
> you were asking for).

--
Chris Arai
Arai Design
chris at araidesign.com
http://www.araidesign.com
707 431 1299
FAX 431 1599

-------------- next part --------------
A non-text attachment was scrubbed...
Name: chris.vcf
Type: text/x-vcard
Size: 224 bytes
Desc: Card for Chris Arai
URL: <http://mail.python.org/pipermail/python-list/attachments/20001220/b5d42f26/attachment.vcf>

From gerrit.muller at philips.com  Wed Dec 13 05:42:30 2000
From: gerrit.muller at philips.com (Gerrit Muller)
Date: Wed, 13 Dec 2000 11:42:30 +0100
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de>
Message-ID: <3A375296.C60B3224@philips.com>

> : - I personally find not having to declare variables a negative (too
> : dangerous), though it's reasonable for simple scripting languages
> 
> Matter of taste, no? :-)
> 
There has been a suggestion that the IDE could warn for "new" variables
(undefined, never referenced), which could indicate typo's

Gerrit
-------------------------------------------------------------------------
Gerrit Muller                Philips Research Laboratories Eindhoven
Building:   WLp 1 17                    Prof. Holstlaan 4
Phone: +31 40 27 45242               5656 AA  Eindhoven
Fax:   +31 40 2743741          The Netherlands
mailto:gerrit.muller at philips.com
http://www.extra.research.philips.com/natlab/sysarch/index.html


From ivnowa at hvision.nl  Thu Dec 14 08:54:34 2000
From: ivnowa at hvision.nl (Hans Nowak)
Date: Thu, 14 Dec 2000 13:54:34 GMT
Subject: sorting date-strings of form: Wed, 05 Jan 2000 10:07:02
References: <3A38CC30.8DA734F4@psi.ch>
Message-ID: <3A38CF82.49A1@hvision.nl>

Jorn Verwey wrote:
> 
> Hello,
> 
> I have a document with a lot of sections which need to be ordered by date. The date is given in the
> format given in the subject: Wed, 05 Jan 2000 10:07:02. How would I go about? Being a novice I would
> rather not re-invent the wheel (it will end up square). I took notice of two packages which might
> help me (mxDateTime and normalDate), but it didn't seem straight forward. I welcome suggestions!
> TIA.

You can use the rfc822 module:

>>> import rfc822
>>> rfc822.parsedate('Wed, 05 Jan 2000 10:07:02')
(2000, 1, 5, 10, 7, 2, 0, 0, 0)

This returns a 9-tuple with time & date info. It can be used directly
for sorting. If necessary, the tuple could be processed further with the
time module.

HTH,

--Hans Nowak
Cheapo Vision Europe BV


From dsh8290 at rit.edu  Wed Dec 27 19:19:04 2000
From: dsh8290 at rit.edu (D-Man)
Date: Wed, 27 Dec 2000 19:19:04 -0500
Subject: Some Python 2.1 ideas
In-Reply-To: <3A4A50E6.11BBE0FD@uswest.net>; from pbjorn@uswest.net on Wed, Dec 27, 2000 at 01:28:22PM -0700
References: <3A452881.7686E717@uswest.net> <%vb16.187233$U46.5801214@news1.sttls1.wa.home.com> <3A464D5D.B541C04E@uswest.net> <mailman.977688488.25438.python-list@python.org> <20001225150250.6FA06A84F@darjeeling.zadka.site.co.il> <mailman.977760789.10224.python-list@python.org> <roy-EA8C3B.11361325122000@news.panix.com> <3A479C29.2631F27B@uswest.net> <mailman.977876470.23568.python-list@python.org> <3A4A50E6.11BBE0FD@uswest.net>
Message-ID: <20001227191904.A11725@westsidecnc.rh.rit.edu>

Yeah, it is a problem if you don't know where the file came from.  How
about creating a standard text file format?  All files could be opened
in binary mode on all platforms, then EOL is handled by each
application (or better yet make a standard module that does the
implementation correctly once).

Seriously, I think that redesigning text files would be best.  You
don't have a problem reading a jpeg or mp3, regardless of the system
it came from and the system it is being used on.  Why should plain
text files be any different?  Plain text files should be the simplest
of all, but alas they are not.

:-)
-D

On Wed, Dec 27, 2000 at 01:28:22PM -0700, Bjorn Pettersen wrote:
<snip> 
> Which would be wrong, since the file you're reading doesn't necessarily come
> from the platform you're running on... (otherwise os.linesep would work).
> 
> -- bjorn
> 



From dennis at codegrinder.org  Fri Dec 15 10:34:15 2000
From: dennis at codegrinder.org (Dennis Conrad)
Date: Fri, 15 Dec 2000 16:34:15 +0100
Subject: Accessing serial port on windows
Message-ID: <91dds1$b0i$05$1@news.t-online.com>

Hello!

I'm "ogling" to move from Perl to Python as my "main scripting language" and
wanted to learn a bit of it by writing my new project in Python. But I need
to read data from the serial port (on Windows) and can't find any
information on this on the net. Is there a possibility to do it? Or do I
have to use the WinAPI?

Thank you very much

Dennis





From phrxy at csv.warwick.ac.uk  Wed Dec 13 10:30:51 2000
From: phrxy at csv.warwick.ac.uk (John J. Lee)
Date: Wed, 13 Dec 2000 15:30:51 +0000
Subject: Trivial Distutils problem
Message-ID: <Pine.SOL.4.30.0012131525400.9329-100000@mimosa.csv.warwick.ac.uk>

With the very basic Distutils script below, I get the following error:

$ python setup.py build
running build
running build_py
error: package directory 'TableIO' does not exist

Why does it think I want to install a package?  It's just one module to go
in the root package.

I'm on windows (above output pasted from a cygwin window, but it does the
same in a DOS shell).


#! /usr/bin/env python

# Distutils setup.py for TableIO

from distutils.core import setup, Extension

setup(name = "TableIO",
      version = "1.4",
      description = "A module for reading ASCII tables into NumPy arrays (and "
       "lists).",
      author = "Michael A. Miller",
      author_email = "miller5 at uiuc.edu",
      py_modules = ["TableIO.py"]
      # should really be Extension("TableIO._tableio"... ?, but Tableio.py
      # does import _tableio rather than import TableIO._tableio
      #ext_modules = [Extension("_tableio", ["_tableio.c",])]
      )


John



From aleaxit at yahoo.com  Thu Dec 28 18:09:01 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 29 Dec 2000 00:09:01 +0100
Subject: Python vs. PHP (& Java?)
References: <69BC4C6AE83ED311BC9400805FA7B14AF84D2C@il93exp01.css.mot.com> <mailman.978018791.20357.python-list@python.org> <92gdp5$6fo$1@animus.fel.iae.nl>
Message-ID: <92ghfh02kug@news2.newsguy.com>

"Carel Fellinger" <cfelling at iae.nl> wrote in message
news:92gdp5$6fo$1 at animus.fel.iae.nl...
> Alex Martelli <aleaxit at yahoo.com> wrote:
> > "Kemp Randy-W18971" <Randy.L.Kemp at motorola.com> writes:
>
>
> >> Can we nip language comparisons in the bud?
> ...
> > (I'll take ""If thought is life//And strength & breath,//And the
> > want//Of thought is death,//Then am I//A happy fly,//If I live//
> > Or if I die." over either, but then, my Blakeian bias is well
> > known).
>
> been doing to much perl lately? Such a pearl deserves Python treatment:

No perl, but my posts' line counts are high enough as is, without
starting to quote poems as 8 lines when they'll fit into two (besides,
I'd have broken _this_ particular gem into just *four* verses, had
I perchance been the author thereof, rather than a Jacobin sympathizer
of an impecunious engraver, so...).


Alex





From ingoogni at home.nl  Wed Dec 27 11:55:08 2000
From: ingoogni at home.nl (ingo)
Date: Wed, 27 Dec 2000 16:55:08 GMT
Subject: ran out of memory...
Message-ID: <Xns9017B2F37seed7@news1.zwoll1.ov.home.nl>

In an attempt to create a 3d dataset, with Python, that can be renderd in 
povray (or, when using a different header, with vtk) I came up with the 
code below.
The initial idea was to use Numeric.fromfunction(), but that didn't work 
out because of the box size and resolution. Or is it possible and I just 
missed it?
In the current approach there is, besides the speed, one problem. When 
making big datasets (150x150x150) the machine runs out of memory while 
working on Dens=map(fnDensity, x, y, z). How can the memory use be 
reduced? I thought about putting the result of fnDensity in a cStringIO 
object. That would probably mean looping through x,y,z 'mannualy', wich 
seems to be no fun either.

Since I'm still very much a programming newbie, any comment on the code is 
appreciated.

---%<------------%<------------%<------------%<---

import math, array, Numeric
def fnx(ix): return (xMin+xLen*(ix/xRes))
def fny(iy): return (yMin+yLen*(iy/yRes))
def fnz(iz): return (zMin+zLen*(iz/zRes))
def fnDensity(x, y, z):
    return (math.cos(Api*z)*math.cos(Api*y) +
        math.cos(Api*x)*math.cos(Api*z) +
        math.cos(Api*x)*math.cos(Api*y))
def fnLimit(Density):
    return int(255*(1/(MMD)*(Density-MinD)))

#size of the box
xMin, yMin, zMin=-math.pi,-math.pi,-math.pi
xMax, yMax, zMax= math.pi, math.pi, math.pi
xLen=xMax-xMin; yLen=yMax-yMin; zLen=zMax-zMin

#x,y,z resolution of the box.
xRes, yRes, zRes = 10.0, 10.0, 10.0
xyRes=xRes*yRes; yzRes=yRes*zRes; xyzRes=xRes*yRes*zRes

'''
x=[];y=[];z=[]
for ix in range(1, xRes+1):
    for iy in range(1, yRes+1):
        for iz in range(1, zRes+1):
            x.append(xMin+xLen*(ix/xRes))
            y.append(yMin+yLen*(iy/yRes))
            z.append(zMin+zLen*(iz/zRes))
'''

#get 'x-axis points'.
Repx=Numeric.ones(int(xRes))*int(yzRes)
x=Numeric.array(map(fnx,range(1, xRes+1)))
x=Numeric.repeat(x,Repx)

#get 'y-axis points'.
Repy=Numeric.ones(int(yRes))*int(zRes)
y=Numeric.array(map(fny,range(1, yRes+1)))
y=Numeric.repeat(y,Repy).tolist()
y=y*int(zRes)

#get 'z-axis points'.
z=map(fnz,range(1, zRes+1))*int(xyRes)

#calculate densities for voxel xyz.
A=0.5; Api=A*math.pi
Dens=map(fnDensity, x, y, z)

#find minimum and maximum densities,
#scale them to fit between 0 - 255.
MaxD=max(Dens); MinD=min(Dens); MMD=MaxD-MinD
Dens=map(fnLimit,Dens)

#write to file.
FileName= "df3Test.df3"
f=open(FileName, 'wb')
f.write(chr(int(xRes)/256)); f.write(chr(int(xRes)%256))
f.write(chr(int(yRes)/256)); f.write(chr(int(yRes)%256))
f.write(chr(int(zRes)/256)); f.write(chr(int(zRes)%256))
array.array('B', Dens).tofile(f)
f.close()

---%<------------%<------------%<------------%<---

groeten,
Ingo

-- 
Photography: http://members.home.nl/ingoogni/
Pov-Ray    : http://members.home.nl/seed7/


From rodsenra at correionet.com.br  Wed Dec 13 12:32:54 2000
From: rodsenra at correionet.com.br (Rodrigo Senra)
Date: Wed, 13 Dec 2000 15:32:54 -0200
Subject: Calldll under Python 2.0
Message-ID: <3A37B2C6.1257E86D@correionet.com.br>

Hi folks!

By any chance have you used calldll.pyd with Python 2.0 over
Windows ? I have just given it a try but two crashes took place:
1) Inside IDLE, "import calldll" causes a Dialog with abend msg.
2) Inside Pythonwin, "import calldll" causes an endless beep :o)
   with no abend dialog.

Any different experiences ?
TIA
Rod
-- 
Rodrigo Senra         
Computer Engineer   (GPr Sistemas Ltda)  rodsenra at correionet.com.br 
MSc Student of Reflection (IC- UNICAMP) Rodrigo.Senra at ic.unicamp.br
http://www.ic.unicamp.br/~921234 (see also   http://www.gpr.com.br)





From MarkH at ActiveState.com  Fri Dec  8 19:06:48 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sat, 09 Dec 2000 00:06:48 GMT
Subject: Implementing IPersistStorage using PythonCOM
References: <t308m58pkecra3@corp.supernews.com> <3A307D8C.10909@ActiveState.com> <t325urdmtq1t07@corp.supernews.com>
Message-ID: <3A317651.5000109@ActiveState.com>

Edward Blakes wrote:

 
> My understanding was that this simply converts a string to a PyIID, not
> generates a brand new GUID. As you will see below, I directly placed the
> string per your suggestion, but to no avail.

My mistake - sorry.  Your code was fine.

> I made the changes suggested, got more output but with the same result.
> Given your response I believe I clipped important information that, in its
> absence, mislead you. So that I won't do that again, I am placing the entire
> source here so that you can run it directly. I hate the length but I want
> you to have all the information needed.

Can you mail it to me - all the lines are wrapped!

Mark.



From tim.one at home.com  Wed Dec 13 18:47:08 2000
From: tim.one at home.com (Tim Peters)
Date: Wed, 13 Dec 2000 18:47:08 -0500
Subject: Python 2.0 Windows installer hangs
In-Reply-To: <WUCY5.129444$U46.4375622@news1.sttls1.wa.home.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCOENAIDAA.tim.one@home.com>

[posted & mailed]

[North Krimsly]
> When I run the Python 2.0 Windows installer on Windows 2000,

>From where did you get it?  For example, did you get it from python.org, or
from ActiveState?  What's the name of the installer file?  What's its size
(in bytes)?

> I get to the second screen (where it asks which components I want to
> install).  Then I click on next and it just hangs, never to return.
> I have disabled my anti-virus utility as well as my software firewall.
> No other applications were running.  I'm stumped.

Same here, and I wrote the installer that comes from python.org <wink>.
That's file BeOpen-Python-2.0.exe, size 5,811,039 bytes.  Is that what
you're using?  "Second screen" doesn't make sense if so (component selection
is at earliest the third screen; and it may be the fourth).

> Any ideas on what might be causing the installer to hang?

None.  Yours is the only report on this I've seen.  But, as above, I don't
know which installer you're talking about.

> Sorry if this is the wrong venue for posting.  If so
> please point me to the right place to post.

This is as good as any.  You could also open a Python bug report on
SourceForge, if it's the python.org installer:

    http://sourceforge.net/bugs/?group_id=5470

Random things to try:

+ Try it on a different machine.

+ Try downloading a fresh copy of the installer (the "Windows installer"
link at http://www.python.org/2.0/#download).

+ If you tried installing when logged in with administrator privileges, try
it from an account without administrator privileges.  Or vice versa.

+ If you downloaded via IE, try using Netscape; or vice versa.  We've seen a
surprising number of corrupted downloads (meaning more than one <frown>),
but can't account for them.  None have displayed the symptom you describe,
though.

+ Ensure you have enough disk space.

+ Accept all defaults.

in-a-pinch-install-linux-instead<wink>-ly y'rs  - tim




From phrxy at csv.warwick.ac.uk  Mon Dec 11 17:21:58 2000
From: phrxy at csv.warwick.ac.uk (John J. Lee)
Date: Mon, 11 Dec 2000 22:21:58 +0000
Subject: C API mismatch with 2.0
In-Reply-To: <Pine.SOL.4.30.0012111210490.29266-100000@mimosa.csv.warwick.ac.uk>
References: <Pine.SOL.4.30.0012102209040.29674-100000@mimosa.csv.warwick.ac.uk> <7qTY5.3054$O%.484335@newsc.telia.net> <Pine.SOL.4.30.0012111210490.29266-100000@mimosa.csv.warwick.ac.uk>
Message-ID: <Pine.SOL.4.30.0012112220470.9610-100000@mimosa.csv.warwick.ac.uk>

On Mon, 11 Dec 2000, John J. Lee wrote:
[...]
> I presume the problem is that I compiled it with Paul Sokolovsky's patched
> 1.5.2 headers.  Is this still necessary to compile with gcc on windows,
> and if so, are there patched 2.0 headers around?  Paul doesn't have any on
> his page.

Okay, forget it, I guess it's pretty obvious that this *is* the problem.
Luckily the patch is small enough that it's applicable by hand to the new
headers.


John



From wware at world.std.com  Sat Dec 16 23:35:48 2000
From: wware at world.std.com (Will Ware)
Date: Sun, 17 Dec 2000 04:35:48 GMT
Subject: Is Stackless and uthread.py SMP friendly?
References: <91br9v$qa4$1@nnrp1.deja.com> <ku1yva5ocq.fsf@lasipalatsi.fi> <91dp4j$5fj$1@news.gte.com> <91dt39$fbk$1@nnrp1.deja.com>
Message-ID: <G5p3Fo.LH4@world.std.com>

jbalding at hotmail.com wrote:
> Okay. My next question: Is there a simple way of taking advantage of
> multiple processors using Python, or is this something that is going to
> be like pulling teeth? In other words, are there existing packages that
> allow relatively high-level, transparent access to multiple processors?

Here is something that looks like it might be a Python wrapper for MPI,
which is a communication protocol for doing the kind of clustering you
are interested in:
http://www.cpsc.ucalgary.ca/~mtigges/mpipython.html

There is also a relevant-looking mailing list here:
http://www.mpi.nd.edu/MailArchives/lam/

Here is something that also looks promising:
http://pypvm.sourceforge.net/
This uses PVM, which is an older protocol for cluster computing. My casual
first impression is that PyPVM looks like it might be better maintained
than mpipython, not based on any actual knowledge.
-- 
# - - - - - - - - - - - - - - - - - - - - - - - -
# Resistance is futile. Capacitance is efficacious.
# Will Ware	email:    wware @ world.std.com


From -$Paul$- at verence.demon.co.uk  Tue Dec 12 18:19:55 2000
From: -$Paul$- at verence.demon.co.uk (Paul Wright)
Date: 12 Dec 2000 23:19:55 -0000
Subject: Handling signals and performance issues.
References: <meRU5.211$Mx.12164@sjc-read.news.verio.net> <m9dgmAAe+tM6EwnY@gol.com> <roy-FD4A85.20535009122000@news.panix.com>
Message-ID: <916bqr$la$1@verence.demon.co.uk>

In article <roy-FD4A85.20535009122000 at news.panix.com>,
Roy Smith  <roy at panix.com> wrote:
>
>You might want to look at the source to the unix "tail" command (you 
>should be able to find it on any of the many GNU/Linux distribution 
>sites) and see how it implements the "-f" option, which is essentially 
>what you are trying to emulate.

It does it by repeatedly sleeping and stating the file: I've just run
strace on it.

-- 
----- Paul Wright ------| /dev/hda - 3.2Gb of spinning metal, all alone in
-paul.wright at pobox.com--| the night...
http://pobox.com/~pw201 | 


From glenfant at equod.com.nospam  Wed Dec  6 10:35:24 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Wed, 6 Dec 2000 16:35:24 +0100
Subject: update
References: <3A2D9A08.43F8FBE8@hypermath.org>
Message-ID: <90lm3u$mpe$1@reader1.imaginet.fr>

I run both python 1.5.2 and Python 2.0 on the same Windows NT box.
The only thing to do is to change the start menu entries for both idle so
they include the FULL path to the required files, otherwise, you'll get some
funny (?) behaviours.
Using the bash provided with Cygwin in place of the very limited windows
command interpreter will help to switch from a python version environment to
the other one ("alias" is available).

"leonrom" <leonrom at hypermath.org> a ?crit dans le message news:
3A2D9A08.43F8FBE8 at hypermath.org...
> I run v.1.5.2, do I have to uninstall it before installing v.2.0



From scarblac at pino.selwerd.nl  Thu Dec 21 18:58:41 2000
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 21 Dec 2000 23:58:41 GMT
Subject: Reality check PYZOMA = python / zope for  Mac ?
References: <t44t20nmcgena7@corp.supernews.com>
Message-ID: <slrn945648.96k.scarblac@pino.selwerd.nl>

Jason Cunliffe <jasonic at nomadicsltd.com> wrote in comp.lang.python:
> I am wondering what the prospects are for Python and Zope on Macintosh:

I don't know anything about this (I don't use Zope and I don't have a Mac)
but I had never expected anyone to want to run Zope on a Mac. Are you really
using a Mac as a web server?!

I thought Macs were pretty cool (and certainly MacOS X) but I thought Zope
was the kind of application that would be completely irrelevant for Macs.

-- 
Remco Gerlich


From goose at enteract.com  Tue Dec 12 02:35:34 2000
From: goose at enteract.com (Jerome Mrozak)
Date: Tue, 12 Dec 2000 01:35:34 -0600
Subject: [Q] Are Exceptions used that much in practice?
Message-ID: <3A35D546.A4731B5B@enteract.com>

I realize that Python allows me to catch exceptions.  But being a novice
I'm not sure if they are used much in the 'real world'.

In Java, my current bread-winning language, exceptions are caught mostly
because the compiler/language demands it.  I haven't yet designed a
system in Java such as:

while(1):
  try:
    read
  catch:
    break
  else:
    ...

Another way of putting it is, if the try-catch-else didn't exist would
anyone be terribly put out?  (If not then the construct isn't much
used...)

Thanks,
Jerome.
  
-- 
Jerome Mrozak          "Never buy a dog and bark for yourself"
goose at enteract.com     --"Slippery" Jim DiGriz
                         (the Stainless Steel Rat)


From glenfant at equod.com.nospam  Wed Dec 13 15:03:30 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Wed, 13 Dec 2000 21:03:30 +0100
Subject: time.strptime(...) Python 1.5.2/Win32 missing ?
Message-ID: <918ken$pju$1@reader1.imaginet.fr>

Where's this function???

<pythonsession>

>>> import time
>>> print time.__doc__
This module provides various functions to manipulate time values.
...
strftime() -- convert time tuple to string according to format specification
strptime() -- parse string to time tuple according to format specification
>>> atime = time.strptime('10:20:30', '%H:%M:%S')
Traceback (innermost last):
  File "<pyshell#71>", line 1, in ?
    atime = time.strptime('10:20:30', '%H:%M:%S')
AttributeError: strptime

</pythonsession>

Seems that there's no time.strptime function! Or did I miss something ?

Thanks



From jhauser at ifm.uni-kiel.de  Sun Dec 17 14:28:08 2000
From: jhauser at ifm.uni-kiel.de (Janko Hauser)
Date: 17 Dec 2000 20:28:08 +0100
Subject: Error evaluating numbers starting in zero
References: <3A3D07BD.85EFEB3D@wag.caltech.edu>
Message-ID: <87k88ysv93.fsf@lisboa.ifm.uni-kiel.de>

"Richard P. Muller" <rpm at wag.caltech.edu> writes:

> "01/08/99", python runs into an error when it tries to eval("08").
> 
The use of int("08") may help

HTH,
__Janko

-- 
  Institut fuer Meereskunde             phone: 49-431-597 3989
  Dept. Theoretical Oceanography        fax  : 49-431-565876
  Duesternbrooker Weg 20                email: jhauser at ifm.uni-kiel.de
  24105 Kiel, Germany


From jrepetto at mxmlab.com  Fri Dec  8 06:14:20 2000
From: jrepetto at mxmlab.com (Jean-Claude REPETTO)
Date: Fri, 08 Dec 2000 11:14:20 GMT
Subject: Looking for compile.py
Message-ID: <3a30c245.95478818@news.wanadoo.fr>

Hi,

I am looking for the David Ascher's compile.py script to create
Microsoft Visual C++ project files. 
The link I have found on
http://www.python.org/doc/current/ext/win-cookbook.html is :

 http://starship.python.net/crew/da/compile/

But it doesn't work ((111) Connection refused) .

Thanks for your help.

Jean-Claude Repetto






From teyc at bigfoot.com  Sat Dec  9 00:32:34 2000
From: teyc at bigfoot.com (Chui Tey)
Date: Sat, 9 Dec 2000 15:32:34 +1000
Subject: sgmlop binaries
Message-ID: <90sg8k$7cc$1@bunyip.cc.uq.edu.au>

Are there any procompiled binaries for sgmlop for python 1.52 on the net?
Thanks.

Chui





From aleaxit at yahoo.com  Mon Dec 18 06:43:35 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 18 Dec 2000 12:43:35 +0100
Subject: Error evaluating numbers starting in zero
References: <mailman.977137021.11074.python-list@python.org>
Message-ID: <91kta6027be@news1.newsguy.com>

"Mike Brenner" <mikeb at mitre.org> wrote in message
news:mailman.977137021.11074.python-list at python.org...
> >> I have a python checkbook program that evaluates date strings, such
> >> as "01/31/99". Most dates are fine, but if I have a date such as
> >> "01/08/99", python runs into an error when it tries to eval("08").
>
> > Use int() instead of eval().
>
>
> So, that may also explain the failure with
>
> i = 08
>
> and with reading in numbers,
> the python interpretor must be using EVAL instead of INT?

If it did otherwise, then
    i = 010
would wrongly give i a numerical value of ten, rather than
the value of eight it's *supposed* to.

It may or may not be wise to have the 0<digits> notation
for octal that the C language (AFAIK) pioneered; I, quite
personally, think it's not wise, since leading-zeros should
not affect a numeric literal meaning to such an extent, and
it's a hard feature to explain to complete beginners -- a
different syntax might have been a wiser choice right from
the start.  But compatibility constraints force Python to
keep it, whatever else one thinks of it, until perhaps the
mythical future 'Python 3000' that WILL be allowed to break
backwards compatibility.

Given that 0<digits> indicates an octal number, it's surely
best to diagnose digit-strings that are an error in that
context, rather than gobble them down with some halfway
sensible 'interpretation' (as some very old C compilers
used to do...).


Alex





From aahz at panix.com  Thu Dec 21 10:36:11 2000
From: aahz at panix.com (Aahz Maruch)
Date: 21 Dec 2000 07:36:11 -0800
Subject: Testing if a list exist
References: <3A421FA8.61A49D48@hotmail.com>
Message-ID: <91t81b$255$1@panix3.panix.com>

In article <3A421FA8.61A49D48 at hotmail.com>,
joonas  <keisari_ at hotmail.com> wrote:
>
>How can I test if a list exists or not.

Based on the code you're using, it seems more that you're trying to test
whether a variable of a specific name exists, rather than testing to see
whether that variable contains a list.  I'd recommend setting the value
of the variable to None before you want to start using it; later on, you
can just say

if foo:

and go on from there.  That's assuming foo will only contain a list; if
it might contain something other than a list, use the types module or code
like this:

if type(foo) == type([]):
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

I don't really mind a person having the last whine, but I do mind
someone else having the last self-righteous whine.  --Aahz


From pbjorn at uswest.net  Mon Dec 25 14:35:21 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Mon, 25 Dec 2000 12:35:21 -0700
Subject: os.exec?
References: <LXw16.8561$KY1.23035@news1.rivrw1.nsw.optushome.com.au> <mailman.977727428.19738.python-list@python.org> <jwbnews-82EF1A.10532625122000@news.olympus.net>
Message-ID: <3A47A179.64548C79@uswest.net>

"John W. Baxter" wrote:

> [snip]
> Or, moving to a higher level, there is the commands module, which wraps
> those things mentioned above.  I tend to be lazy and use that.

The commands module is hopelessly unix centric when it doesn't need to be so I
consider it pretty much useless...

-- bjorn




From rturpin at my-deja.com  Tue Dec 26 09:51:43 2000
From: rturpin at my-deja.com (rturpin at my-deja.com)
Date: Tue, 26 Dec 2000 14:51:43 GMT
Subject: One Python 2.1 idea
References: <mailman.977771949.17911.python-list@python.org> <9292tn$jka$1@nnrp1.deja.com> <929n9c013m3@news1.newsguy.com>
Message-ID: <92ab9u$dou$1@nnrp1.deja.com>

In article <929n9c013m3 at news1.newsguy.com>,
  "Alex Martelli" <aleaxit at yahoo.com> wrote:
> No implicit caching.

Has there been some discussion about caching? If done
well, the vast majority of references would reach
their object's attribute or method for the cost of a
couple of integer compares and array fetches. This
should be an order of magnitude faster than string
lookup in a dictionary. This especially benefits
method invocation, since that costs a failed lookup
to the instance dictionary, and a successful lookup
in the class dictionary. That's almost enough to
scare me away from methods!

Most work in a Python program involves object methods
or attributes. (At least, this is the case in my
programs. ;-) Increasing the speed of reference to
these should have appreciable benefit, across the
board. Yeah, it would make the code a bit more
complex to maintain. But this isn't a special case --
it is a general mechanism central to how Python works.

Russell





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


From tim.one at home.com  Thu Dec 28 15:31:49 2000
From: tim.one at home.com (Tim Peters)
Date: Thu, 28 Dec 2000 15:31:49 -0500
Subject: One Python 2.1 idea
In-Reply-To: <92f1it01at1@news1.newsguy.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEMEIFAA.tim.one@home.com>

[Tim]
> About x.foo = x.foo, I never saw anyone suggest that to save
> time, before this thread.  Where it did come up several times
> over the years was people trying to customize how a method works
> for a particular instance, rather than make a proper subclass.
> Since Python made instance.__class__ writable, I don't see any
> excuse for *that* form of trickery anymore.

[Alex Martelli]
> Might I ask for a clarification of this latter thought...?
>
> Say I have a class Ogre with a few methods:
>
> class Ogre(Bugaboo):
>     def fee(self):
>         print 'fee'
>     def fie(self):
>         print 'fie'
>     def foo(self):
>         print 'foo'
>     def fum(self):
>         print 'fum'
>
> and several instances thereof, one of which is referenced by
> variable x.
>
> Now, I find out that, for that one instance only, I want to
> override the normal Ogre.foo method with another function
> which I also happen to have around:
>
> def fun(self):
>     print 'have fun!'
>
>
> What are the 'good Pythonic ways' to do this, as opposed to
> 'dirty tricks'?
> ... [assorted trickery with the "new" module] ...

class UniqueOgreFoo(Ogre):
    foo = fun
x.__class__ = UniqueOgreFoo

x.foo() # now prints 'have fun!'

> ...
>     x.__class__ = new.classobj('NewOgre', (Ogre,), {'foo':fun})
>
> Are you saying the second approach is preferable?

I'm saying (1) it's clearer to use a subclass than to do the "x.foo = x.foo"
trick; (2) the subclass method does not create cycles; and, (3) using the
"new" module to spell a subclass is unnecessarily obscure.

> Or is there some even-simpler one that I'm missing?  I do realize
> a cycle of references is generated by the new.instancemethod, as
> discussed on this thread -- is that the only issue,

Very few Python programmers understand the "new" module, so I'd say the
obscurity of the code is a bigger issue in practice.  Not that mucking
around with __class__ is recommended to begin with!  That's obscure too,
just not as obscure as new.*.  Best would have been to make x an instance of
UniqueOgreFoo from the start.

> and does GC solve it (if no __del__'s are present),

I hesitate to answer that (and not just because I did before <wink>),
because the set of types gc will "chase" is very likely to change from
release to release, and people should be able to guess good answers to this
kind of question for their particular case of interest via just trying it.
For example,

>>> class C:
        def f(self): pass

>>> addresses = {}
>>> for i in range(1000):
        c = C()
        c.f = c.f
        addresses[id(c)] = 1

>>> len(addresses)
617
>>>

If gc had not cleaned up some of the C instances, every id(c) would have
been unique.  OTOH, that 617 unique addresses were seen strongly suggests
that pure refcounting wasn't enough to clean them up.  Two variations:

>>> addresses = {}
>>> for i in range(1000):
        c = C()
        addresses[id(c)] = 1

>>> len(addresses)
3
>>>

So in that case (no "c.f = c.f") C instances are getting reused very
quickly, so pure refcounting probably sufficed.  While in this case (after
adding a __del__) they're not getting recycled at all:

>>> C.__del__ = lambda self: 0
>>> addresses = {}
>>> for i in range(1000):
	c = C()
	c.f = c.f
	addresses[id(c)] = 1

>>> len(addresses)
1000
>>>

Those are all probabilistic clues, and platform-dependent, but good enough.
Running gc.collect() explicitly and looking at the length of gc.garbage
before and after can nail it if you need more assurance.

give-a-man-a-fish-and-you've-fed-him-for-a-day-but-teach-a-man-to-
    fish-and-he'll-come-by-every-day-asking-for-bait-lures-spare-
    lines-and-your-best-pole<wink>-ly y'rs  - tim




From greg at cosc.canterbury.ac.nz  Wed Dec 20 19:35:42 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Thu, 21 Dec 2000 13:35:42 +1300
Subject: while true: !!!
References: <3A4024E5.98843CFC@cosc.canterbury.ac.nz> <91qadb$1di$1@plutonium.compulink.co.uk>
Message-ID: <3A41505E.FA94F865@cosc.canterbury.ac.nz>

gbreed at cix.compulink.co.uk wrote:
> 
> Adding a boolean type wouldn't make anything simpler.

No, but I really meant having a boolean type *instead*
of the notion of being able to treat anything at all
as a boolean value.

>  It would only
> change the question to "What objects x will boolean(x) return false?"

There wouldn't be any such function. Or if there were,
it would be analagous to chr() and map (0, 1) --> (false, true)
and nothing else, and it would be used about as
rarely as chr().

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


From Norman_Shelley-RRDN60 at email.sps.mot.com  Thu Dec 21 18:08:23 2000
From: Norman_Shelley-RRDN60 at email.sps.mot.com (Norman Shelley)
Date: Thu, 21 Dec 2000 16:08:23 -0700
Subject: ?Why is 0e-6 invalid?
References: <3A428C7C.5CFBB181@email.sps.mot.com>
Message-ID: <3A428D67.CF2F2FDE@email.sps.mot.com>

Oh, and this is considered valid! ?
>>> 0.0e-6
0.0



Norman Shelley wrote:

> Why is 0e-6 invalid and 0*.000001 is valid?
>
> They are equivalent and a "normal" mathematically trained individual
> sees them for what they are immediately.
>
> (I have a trivial engineering notation to number convertor that takes
> "0u" or "5u" and converts them to "0e-6" and "5e-6" and then evaluates
> them. The later works fine, the former fails. Granted I could check for
> this condition, but it seems that the Python evaluator should be able to
> work on this boundary case as it still consistent.)
>
> >>> 0*.000001
> 0.0
> >>> 1e-6
> 1e-06
> >>> 0e-6
>   File "<stdin>", line 1
>     0e-6
>      ^
> SyntaxError: invalid syntax



From eblakes at drwtrading.com  Sun Dec 10 19:14:34 2000
From: eblakes at drwtrading.com (Edward Blakes)
Date: Sun, 10 Dec 2000 19:14:34 -0500
Subject: Implementing IPersistStorage using PythonCOM
References: <t308m58pkecra3@corp.supernews.com> <3A307D8C.10909@ActiveState.com> <t325urdmtq1t07@corp.supernews.com> <3A317651.5000109@ActiveState.com> <t339eb6pvurf67@corp.supernews.com> <3A32D22B.3050901@ActiveState.com>
Message-ID: <t3878bc92giqb5@corp.supernews.com>

I'm not building for sources. I'd appreciate your emailing me a new
pythoncom20.dll. I can wait for a later ActivePython distribution, since we
are currently in a "proof of concept" stage for this use of Python. If we
find we're ready to deploy and a new version is not available, I'll look
into building from sources.

Thanks for your help.

Ed

"Mark Hammond" <MarkH at ActiveState.com> wrote in message
news:3A32D22B.3050901 at ActiveState.com...
> IPersistStorage is pretty badly broken when trying to implement it from
> Python. A really obvious error in PyGPersistStorage.cpp.
>
> I have fixed it here, but not too sure how that helps you - if you are
> building from sources, then you can simply get a new CVS update and
> rebuild.  If you aren't I could try mailing you a new pythoncom20.dll
>
> Mark.
>




From red5sf at my-deja.com  Sat Dec  9 19:07:13 2000
From: red5sf at my-deja.com (red5sf at my-deja.com)
Date: Sun, 10 Dec 2000 00:07:13 GMT
Subject: Dumb Newbie Question
References: <3A327D01.9841EF98@drew.edu>
Message-ID: <90uhff$9ul$1@nnrp1.deja.com>

dude, i just had this same problem.  I tried editing the PYTHONPATH
variable in the PythonPath editor (via Pythonwin) but it didn't quite
work.

The simplest solution (if you're using PythonWin as your IDE) is to add
edit the pythonwin path.  Go to the Edit PythonPath from the Tools
pulldown menu and edit the pythonwin registry to include your script
folder.

hope that's clear,
christian

In article <3A327D01.9841EF98 at drew.edu>,
  Scott Weisel <sweisel at drew.edu> wrote:
> I'm trying to set up Python so I can import the programs that I
wrote.  I
> want to modify $PYTHONPATH so the directory with my programs is in
the path
> but I'm not sure how.
>
> I thought I might be able to modify sys.path like this
> sys.path = sys.path + "c:\\python16\\programs
> but I'm getting a TypeError stating "illegal argument type for built-
in
> operation."
>
> Any suggestions would be appreciated
> -Scott
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From jerry_spicklemire at my-deja.com  Thu Dec 28 13:29:43 2000
From: jerry_spicklemire at my-deja.com (jerry_spicklemire at my-deja.com)
Date: Thu, 28 Dec 2000 18:29:43 GMT
Subject: Reality check PYZOMA = python / zope for Mac ?
References: <t44t20nmcgena7@corp.supernews.com> <slrn945648.96k.scarblac@pino.selwerd.nl> <91uar5$fj0$2@newshost.accu.uu.nl>
Message-ID: <92g0qf$mgb$1@nnrp1.deja.com>

  Jason Cunliffe wrote:

>> I am wondering what the prospects are for Python and Zope on
>> Macintosh: Is the mac getting left behind or will the new MacOSX in
>> fact bring Python and Zope more easily into the fold?

to which Martijn Faassen replied:

> Since MacOS X is really a BSD Unix, it should run Zope just fine.
> I think I heard this mentioned in fact on some Zope mailing list.
> I don't know much about this either, though; check with the Zope
> mailing lists.

Yes, see

 http://www.zope.org/Members/jshell/buildingZopeOnMacOSXServer

 http://www.zope.org/Members/sspickle/MacOSXBHFS

 http://lists.zope.org/pipermail/zope/2000-December/037536.html

 http://lists.zope.org/pipermail/zope/2000-December/037541.html

 http://lists.zope.org/pipermail/zope/2000-December/037543.html


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


From scarblac at pino.selwerd.nl  Tue Dec 12 07:00:30 2000
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 12 Dec 2000 12:00:30 GMT
Subject: A question.
References: <9152a0$acb$1@panorama.wcss.wroc.pl>
Message-ID: <slrn93c4mc.78m.scarblac@pino.selwerd.nl>

DeHa <deha at poczta.onet.pl> wrote in comp.lang.python:
>     I am new to Python, and just try to write my first program. But I have a
> small problem.
> 
>     My program should find first numbers in the set from 1 to 100000. I made
> such a code:

You probably meant prime numbers.
> 
> first = range(100001)
> first = first[1:]

Instead, simply first = range(1,100001) would be easier. Actually,
the range should start at 2, since that is the first prime number
(and you don't want to delete numbers that are divided by 1).

> for i in first:
>     if i != []:
> #until end of the list

"for i in first" means that i will become all the elements of first in turn.
So i never becomes [], since first only contains integers. And it already
loops until the end of the list.

>         for j in first[first.index(i):]:                           #find all
> numbers divided by i

This also includes the number i itself, it shouldn't.

>             if j % i == 0:
>                 first[first.index(j): first.index(j)] = []    #and delete it
> from the list

You are deleting from a list while you are walking through it with 'for',
that's a bit tricky.

>     And it doesn't work, and I don't konw why? Could anyone help me? I try
> to find answer in the tutorial, but the information in it doesn't change
> anything.

It is easier to start with an empty list of primes, and add numbers that
can't be divided by the numbers in the list.

Something like this:

primes = [2]  # We need one prime to start with
for candidate in range(3, 100001, 2):    
    for prime in primes:
        if candidate % prime == 0:
            break
    else: 
        # The 'else' clause happens when the 'break' never happened
        primes.append(candidate)


This is not the fastest solution, but it works.

-- 
Remco Gerlich


From sreiling_ at _yahoo.dot.com  Thu Dec  7 15:31:46 2000
From: sreiling_ at _yahoo.dot.com (sreiling_ at _yahoo.dot.com)
Date: Thu, 07 Dec 2000 20:31:46 GMT
Subject: Graphs from python
References: <90gse2$14uo$1@news.rchland.ibm.com>
Message-ID: <3a2ff37f.485165821@news.tripos.com>

You could also check out Gnuplot. There is a Gnuplot.py package that
interfaces to it. I don't know if it has the features you need,
though.

	S.
On Mon, 4 Dec 2000 13:48:18 -0600, "Larry Whitley" <ldw at us.ibm.com>
wrote:

>The simulation models I write and run generate a lot of data and to make
>sense of it all I put the data into a set of graphs.  The graphs are the
>typical spread sheet fare, bar graphs, X/Y graphs, scatter plots, etc.
>Today, I read the data into spreadsheets from the spreadsheet and create my
>graphs by hand and I'd like to automate the process.
>
>I'm looking at using win32com (by Mark Hammond) to drive excel or
>piddle+graphite to generage graphs directly from python.  While I'm sure it
>will get the job done, the excel object model is a bit overwhelming.  Piddle
>by itself isn't at the right level.  Graphite appears get to the right level
>but the pre-alpha rating makes me think that it's a bit early to jump in.
>
>Are there other approachs I should consider?  Whatever I end up with, I'd
>like it to work from either Unix or Windows.  If I can't have both, I can
>make do with just windows.  Of the two approaches above, is one
>significantly better than the other (by whatever measure you want to
>choose)?
>
>Thanks,
>Larry
>
>



From wcuddy at lserv.ja10629.home  Sun Dec 10 11:41:11 2000
From: wcuddy at lserv.ja10629.home (Wayne Cuddy)
Date: Sun, 10 Dec 2000 16:41:11 GMT
Subject: using execfile() in class/instance methods
References: <slrn932qna.rtk.wcuddy@lserv.ja10629.home> <90tb1302s27@news2.newsguy.com> <slrn934t6r.s9l.wcuddy@lserv.ja10629.home> <90ueaf02n20@news2.newsguy.com>
Message-ID: <slrn937chv.s9l.wcuddy@lserv.ja10629.home>

In article <90ueaf02n20 at news2.newsguy.com>, Alex Martelli wrote:
>"Wayne Cuddy" <wcuddy at lserv.ja10629.home> wrote in message
>news:slrn934t6r.s9l.wcuddy at lserv.ja10629.home...
>    [snip]
>> Your Example Worked But I Want Multiple Instances Of This Class And Be
>> Able To Pass The Filename As An Arg.
>
>So you don't want a class-attribute but an instance-attribute, OK, then you
>do need to pass the appropriate dictionary to execfile (no need for a
>locals/globals distinction here):
>
>class Ba:
>    def __init__(self, filename):
>        execfile('ba.txt', self.__dict__)
>    def bu(self):
>        print self.bn['key']
>
>ba = Ba('ba.txt')

Good, this is a usable solution for me..
>
>ba.bu()
>
>OK so far?
>
>> I Also Would Prefer That Bn Be A
>> Local Variable Of The Function Containing The Execfile() Call.  This
>
>The locals() dictionary is not guaranteed to be usable this way -- it
>will surely work for reading, but modifications to it will not
>necessarily be reflected in your local-variables (there's an important
>optimization behind this restriction, so, suffer it...!).
>
>Just pass any old dictionary and use it...!
>
>> Would Allow Me To Either Return It Or Bind It To Any Variable Name Of
>> The Instance.  So I Want To Do Something Like This:
>>
>> Class Bla:
>>       ...
>>       ...
>>       Def F1(Self,Filename):
>>   Execfile(Filename)
>>   Self.Dictfromdisk = Bn
>>
>>       Def F2(Self):
>>   Print Bn
>
>Didn't you just say that you wanted bn to be LOCAL to the function
>doing execfile?  If you got your wish, then here it would be local
>to f1, thus completely unknown in f2.
My, mistake should F2 should have contained: print self.distfromdisk..

>
>The only way f2 can access bn just like that if it is a GLOBAL
>variable, which you can obtain by using
>    execfile(filename,globals())
>in f1, but is a perfectly horrid idea IMHO.
>
>Apart from this weirdness in f2, f1 can be implemented easily:
>
>    def f1(self, filename):
>        a={}
>        execfile(filename, a)
>        self.dictFromDisk = a['bn']
>
>for example.
>
>
>Alex
>
>
>
I like your second solution better anyway.
You solved my problems... Thanks Alex



From bernhard at abnoba.intevation.de  Thu Dec 14 05:40:35 2000
From: bernhard at abnoba.intevation.de (Bernhard Reiter)
Date: 14 Dec 2000 10:40:35 GMT
Subject: Python for REAL Projects ?
References: <3a380d1e.409238@bendigo.news.origin.net.au>
Message-ID: <91a833$k34$1@newsserver.rrzn.uni-hannover.de>

In article <3a380d1e.409238 at bendigo.news.origin.net.au>,
	dragon at hitech.net.au writes:

> I have been looking at, and trying out Python (under Win NT) and
> although Python SEEMS interesting can it "cut the mustad" on a real
> client project ??

Yes.  At least it does here. :)

> Can this type of project be sucessfully tackled with Python + Tkinter
> + MySQL ??

I prefer wxPython over Tkinter.
(Note that BoaConstructor, when it is getting stable aims to be a
delphi like IDE for wxPython. It is quite impressive already.)

As to your project, it is hard to tell if you are faster with
python. It certainly can be done and it can be done nicely, 
but it might need a little bit more eningeering on your part.
	Bernhard


-- 
Professional Service around Free Software                (intevation.net)  
The FreeGIS Project                                         (freegis.org)
Association for a Free Informational Infrastructure            (ffii.org)
FSF Europe                                            (www.fsfeurope.org)


From grisha at ispol.com  Fri Dec  8 17:15:55 2000
From: grisha at ispol.com (Gregory Trubetskoy)
Date: Fri, 08 Dec 2000 22:15:55 GMT
Subject: Documentation - underscores in \section
Message-ID: <Pine.LNX.4.21.0012081707130.29380-100000@saturn.ispol.com>

[Sorry this is perhaps somewhat off-topic]

For LaTeX gurus out there - 

Has anyone noticed that underscores do not work in \section, even if
preceeded by a \ (backslash). 

Mod_python, having an underscore in the title, is very much affected by
this little problem...

I am not a LaTeX expert, so I thought I was doing something wrong, but
then noticed that the orginal Python documentation has the same problem.

E.g. if you look at the PDF documentation for sunau module, AU_read,
section 14.4.1, you will see in the description something like:

AUprotect unbox voidb at x kern .06....

Is there any way around this?

--
  Gregory (Grisha) Trubetskoy
       grisha at ispol.com



From tanzer at swing.co.at  Sun Dec 17 04:33:39 2000
From: tanzer at swing.co.at (Christian Tanzer)
Date: Sun, 17 Dec 2000 10:33:39 +0100
Subject: Solution: Direct access to Printer I/O lines 
In-Reply-To: Your message of "16 Dec 2000 20:27:53 +0100."
             <91gfnp$7r4$1@animus.fel.iae.nl> 
Message-ID: <m147aCK-000wcEC@swing.co.at>

Carel Fellinger <cfelling at iae.nl> wrote:

> Alex Martelli <aleaxit at yahoo.com> wrote:
> ...
> > This gives me a good chance for one of my favourite rants,
> > '"Singleton" design-pattern considered harmful'!-)  But
> > 'Singleton' is not being ideally applied, so I'll skip that.
> 
> And in an other thread you wet my appetite and direct me here!
> So let us hear that rant, mister, and enlight us mere mortals
> who still think that there is legitimate use of this singleton
> pattern.

Whether something is a singleton is highly dependent on (the scope of)
the application. If you want to have an evolvable and/or reusable
solution it might be better not to use the singleton pattern. 

Often just a tiny shift in your (or your customer's) point of view
suffices to convert an obvious case of singleton-ness into a situation
where more than one object is needed. Even in the case of a shared
resource the singleton pattern fails if you discover the need of
supporting more than one instance of the resource in question (e.g.,
two instead of one printers).

A random example: some time ago I implemented a class Command_Line
tackling the same problem as the getopt module but providing a nicer
interface. It seems obvious that you cannot have more than one
instance of this class per application. Some time later I found it
too difficult to try to teach command line interfaces to Windows users
and implemented a class Script_GUI which uses the Command_Line
instance of a python script to provide a WIMP interface for the
script. Of course, Script_GUI needs a Command_Line instance of its
own.

In my experience, the number of possible instances of a class is not a
property of the class but of the application where the class is used.
Putting a restriction on the number of instances into the class
therefore decreases the reusability of the class and makes life very
unpleasant if the original application must be changed to accommodate a
different number of instances.

Although I've held this view for a number of years I occasionally
succumb to the lure of the singleton pattern and more often than not I
regret it later on.

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




From nospam at ktgroup.co.uk  Tue Dec 12 06:07:11 2000
From: nospam at ktgroup.co.uk (nospam at ktgroup.co.uk)
Date: 12 Dec 2000 11:07:11 +0000
Subject: HPUX install
References: <913ett$u0b$1@nnrp1.deja.com>
Message-ID: <ulmtl525c.fsf@ktgroup.co.uk>

Brad <blosh at tsbrass.com> writes:

>   cd Parser ; make OPT="-O" VERSION="2.0" \  prefix="/usr/local"
> exec_prefix="/usr/local" all  cc -Ae -O -I./../Include -I.. -DHAVE_CONFIG_H
> -c pgenmain.c (Bundled) cc: warning 480: The -A option is available only with
> the C/ANSI C pro duct; ignored. (Bundled) cc: warning 480: The -O option is

You need either the purchasable HP ANSI C compiler, not the bundled C compiler,
or gcc.

I've built Python with the HP ANSI C compiler with no problems. I've never
tried with gcc.


From fredrik at effbot.org  Tue Dec 19 10:46:35 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 19 Dec 2000 15:46:35 GMT
Subject: tkinter event binding to mouse double click
References: <91n7l6$7ij$1@news.nuri.net>
Message-ID: <vpL%5.459$AH6.36619@newsc.telia.net>

"June Kim"
> In the code following, the binding doesn't seem to work; it can't detect
> mouse double-clicking at all.

assuming you're really double-clicking in the frame widget,
there's nothing wrong with your code.

maybe you're not clicking fast enough?

> (Maybe a platform-specific problem, I reckon)

works on win95, win98, win NT 4, and win 2K.  maybe you
have some other kind of win32?

</F>




From thomas at xs4all.net  Tue Dec 12 12:14:42 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Tue, 12 Dec 2000 18:14:42 +0100
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
In-Reply-To: <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu>; from katz@Glue.umd.edu on Tue, Dec 12, 2000 at 11:46:48AM -0500
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu>
Message-ID: <20001212181442.L4396@xs4all.nl>

On Tue, Dec 12, 2000 at 11:46:48AM -0500, Roy Katz wrote:

> > (Hint: Everything is a reference to an object.  Variables are
> > named references, not actual objects.  Some objects can be
> > modified in place.  Some cannot be modified.  That's all)
> 
> Okay! tell me, which are the objects which cannot be modified in place?

The immutable ones. Those include the numeric types, tuples and strings, and
classes that define themselves as immutable.

> Pardon my choice of diction, but I am against inventing new terminiologies
> (modified-in-place??).

Mutable. Of the builtin types, that's lists, dictionaries, arrays, and most
classes.

> While I'm brainstorming with Python grammar changes, what about a real &
> operator.

What would the real & operator do ? I assume here you mean the unary & from
C, the 'address-of' operator ? The binary & already exists, and does exactly
what it does in C. If you want to know if an object you have now is the same
object as you had previously, either store a reference to it and compare it
using 'is', or use the 'id()' builtin. You can't convert the number returned
by 'id' into a reference, though, for a lot of reasons -- the most important
of which is "that isn't the Python way" :) The rest have to do with data
integrity, reference-counting isues, code sanity, and lots more.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From qrczak at knm.org.pl  Sun Dec 31 11:57:01 2000
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 31 Dec 2000 16:57:01 GMT
Subject: array of objects
References: <92n1s7$tq9$1@nnrp1.deja.com> <siE36.2779$Qb7.318673@newsb.telia.net> <FRI36.76$KD3.30489@typhoon.aracnet.com>
Message-ID: <slrn94upat.jlu.qrczak@qrnik.zagroda>

Sun, 31 Dec 2000 08:14:27 -0800, Daniel Klein <DanielK at aracnet.com> pisze:

> How then is list comprehension different (and/or better) than say:
> 
> obj = []
> for x in range(N): obj.append(test())

It is an expression. You don't have to bind the object to a name
(obj here).

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


From ggabor at sopron.hu  Wed Dec  6 07:22:10 2000
From: ggabor at sopron.hu (Gabor Gludovatz)
Date: Wed, 6 Dec 2000 13:22:10 +0100 (CET)
Subject: Tkinter question
Message-ID: <Pine.LNX.4.21.0012061321250.7564-100000@ns.sopron.hu>

Hello,

I'd like to control two Listboxes with one scrollbar, but I am unable to
write a lambda which would call the two seperate yview functions.

here is a little piece of my code:

        self.wordlist1 = Listbox(self.dictframe_left, yscrollcommand =
self.scrollbar.set)
        self.wordlist1.pack(fill = "both", expand = 1)

        self.wordlist2 = Listbox(self.dictframe_right, yscrollcommand =
self.scrollbar.set)
        self.wordlist2.pack(fill = "both", expand = 1)

        self.scrollbar['command'] = lambda
*largs: self.wordlist1.yview(largs), self.wordlist2.yview(largs)


on the last line you can see the lambda expression.. but it's wrong from
at least three aspects...
ad1: self.wordlist1.yview in the lambda is not accessible in the lambda's
namespace
ad2: largs is a tuple
ad3: it doesn't work :)

somebody please could help me?

thanks in advance

-- 
 Gabor Gludovatz <ggabor at sopron.hu> http://www.sopron.hu/~ggabor/






From rwklee at home.com  Mon Dec  4 19:39:44 2000
From: rwklee at home.com (Rick Lee)
Date: Tue, 05 Dec 2000 00:39:44 GMT
Subject: Researching Python IDEs, debuggers, code coverage, unit testing and 
 mem  management tools
References: <3A2A175C.D144ACA6@hotmail.com> <90fd6f$3vp$1@news1.xs4all.nl>
Message-ID: <3A2C3950.774E7C4A@home.com>

Thanks.  That was very helpful.  According to Archeoptyrx's web page, their
Wing IDE is now at commercial release, presumably that means post-beta.  Any
update on its stability?

Also, I would be interested in people's view on the importance of a 'project
manager' for projects involving multiple modules.

- Rick Lee

Boudewijn Rempt wrote:

> DataJockey at hotmail.com wrote:
> > Hello all,
>
> > I am new to Python, so I researching Python IDEs, debugging, code
> > coverage, unit testing and mem management tools.  I have found very few
> > sources of information.  So I figured I would ask the group to reply
> > with the tools that you know of in the above areas, wheather they are
> > freely available or commercial products.
>
> > I would like to compile a list of these tools, from the feed back that I
> > receive.  Also, feed back on these tools would be great.
>
> > I hope that everyone could learn something new from the results. =)
>
> I've just got an article published that compares Pythonworks,
> Pythonwin, Wing IDE and IDLE on InformIT: http://www.informit.com.
>
> --
>
> Boudewijn Rempt  | http://www.valdyas.org



From aleaxit at yahoo.com  Tue Dec 19 08:41:34 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 19 Dec 2000 14:41:34 +0100
Subject: while true: !!!
References: <mailman.977180344.10242.python-list@python.org>
Message-ID: <91noj60178k@news2.newsguy.com>

"Carsten Geckeler" <nospam at no.spam> wrote in message
news:mailman.977180344.10242.python-list at python.org...
> On Mon, 18 Dec 2000, Steve Lamb wrote:
>
> > On Mon, 18 Dec 2000 17:05:51 +0100, Alex Martelli <aleaxit at yahoo.com>
wrote:
> > >I think it would be inconsistent, un-Pythonic, to add a fundamental
type for
> > >'conceptual' reasons, without any practical, measurable benefit.  For
> > >example, Python does away with the distinction between 'a character'
and 'a
> > >string of length 1' -- a distinction that is religiously maintained in
C,
>
> Not fair!  C is a lowlevel language (relatively to interpreter languages
> as Python etc.).  In C you can say

C is a low-level language, Java a bit less so, Icon even less, and
Haskell could hardly be accused of being very low-level -- they all
distinguish a character from a stringth of length 1 as fundamentally
different datatypes, as do many, indeed most, other languages.

My claim is that Python is simpler, clearer, and MORE effective,
for NOT trying to draw this level of distinction.


> s[3] = char
> Do you know anything faster than this?  It just writes the byte char into
> the memory space pointer str + 3.  In most cases strings do consist of
> chains of bytes in memory.

Not sure what you mean by 'chain' here -- Haskell is the only one
of the languages I've listed that models strings as _lists_ (aka
linked lists, aka chains) of characters, most others (excluding
some Lisp dialects & variants) normally preferring to model them
as _arrays_ (aka blocks, &c) of characters (what _Python_ calls
'lists' are actually flexible arrays, which further confuses the
terminology).

Modeling character as a fundamental datatype, distinct from a
1-character string, is clearly NOT something that languages
do just for performance reasons -- in Java, strings are not
mutable, so the idiom you quote does not translate, yet it still
models character as a separate datatype; Haskell's strings-as-
chains is anything BUT well-performing... yet, it STILL models
character as a separate datatype.

So it's clearly not a sacrifice that language designers endure
just because it will enhance their performance... some must
believe it's a GOOD distinction to draw.  Python (and Perl
and Tcl, just to name a couple others) disagree -- and I
think they're right.


> Please, don't be so mean to C.  It's such a nice programming language!
> Remember, Python itself is programmed in C.  ;)

I program in C when I must (though I'd much rather take a
step up to C++, which I can in most situations but *NOT* when
I want my Python extensions to be taken seriously by the
Python community, alas...).  That doesn't mean I have to
_like_ it:-).


> But to your point:  If you are sure that your string is always shorter
> than 79 characters, it'n no faster way than doing it as C.  No overhead
> due to range checking and reallocating.  But of course, from the everyday
> user's point, interpreter languages as Python or Perl do a nicer job.  But
> every language has it's field of appliance.  Just imagine programming a OS
> kernel in Python... ;)

I can at least imagine programming it in C++ (BeOS) or even, if need
be, in Objective C (Next) -- in both cases, with a C-like subset
available for the very lowest portions of the job, but a semi-decent
higher-level approach for most of it.


Alex





From not.this at seebelow.org  Sat Dec  9 07:09:10 2000
From: not.this at seebelow.org (Grant Griffin)
Date: Sat, 09 Dec 2000 12:09:10 +0000
Subject: exec'ing an import
References: <3A315C1A.90CB8D98@seebelow.org> <p6qpuj1amr0.fsf@informatik.hu-berlin.de>
Message-ID: <3A3220E6.4D893842@seebelow.org>

Martin von Loewis wrote:
> 
> Grant Griffin <not.this at seebelow.org> writes:
> 
> > Does anybody have an explanation?
> 
> Perhaps script.py is not on sys.path, so import won't find it.

Well, I had thought that too, so I went through a lot of path-related
contortions (even going so far as to explicitly include the absolute
path in the exec'd import statement), but that wasn't it.

The thing that cured it was executing "python script.py" rather than
just "script"; the script itself was unchanged.

=g2
-- 
_____________________________________________________________________

Grant R. Griffin                                       g2 at dspguru.com
Publisher of dspGuru                           http://www.dspguru.com
Iowegian International Corporation	      http://www.iowegian.com


From MarkH at ActiveState.com  Fri Dec  8 01:26:44 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Fri, 08 Dec 2000 06:26:44 GMT
Subject: win32evtlogutil: How does it works ?
References: <3A2D556F.6240C661@club-internet.fr> <3A2D8223.4070408@ActiveState.com> <3A3026BF.CB70C2E2@club-internet.fr>
Message-ID: <3A307DDD.60309@ActiveState.com>

Patrice Garachon wrote:

> 
> Today, I have tried to use PythonService.exe as Dll message.
> If you use the 1001  or 1002 Id number, it doesn't works but if you  use
> 0x4000100 1 or 2 It works ?

Yes - I believe this is the "severity" bit - ie, indicating if it is 
error, warning or info number 1001.

Mark.



From aleaxit at yahoo.com  Sun Dec 24 17:26:07 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 24 Dec 2000 23:26:07 +0100
Subject: One Python 2.1 idea
References: <mailman.977590932.19335.python-list@python.org> <9237md$orm$1@nnrp1.deja.com> <924bjt0e3n@news1.newsguy.com> <925baq$5sj$1@nnrp1.deja.com>
Message-ID: <925tak01tpu@news1.newsguy.com>

<rturpin at my-deja.com> wrote in message news:925baq$5sj$1 at nnrp1.deja.com...
> Let me frame, up front, where I think we disagree. We
> both love Python for its level of abstraction, its
> flexibility, its portability, and its ability to serve
> as glue. The issue is NOT Python vs. something else. It

For me, it's Python *plus* something else -- but I
think that much is clear.

> is about how Python should evolve AT THIS POINT in its
> lifecycle. My view is that Python, from a language
> viewpoint, is quite mature. There are not a lot of big
> abstractions yet-to-be implemented. Or if there are, I
> have not seen them discussed. Most of the enhancements

The 'interfaces' strawman-proposal (which doesn't seem
to have made it into a PEP, yet) is the most significant
one I can think of.  Generators, micro-threads, &c, are
also of some importance... but, yes, maturity, and thus
stability, are surely there.

> now proposed are small and incremental. Python has
> reached the point where the greatest gains come from
> improving its underlying infrastructure: more speed,
> running on more platforms, supporting larger-scale
> programming. This is GOOD. It is a sign of success.

Nolo contendere regarding the importance of
infrastructure -- but I wouldn't single out speed as
a particularly crucial aspect.


> Stackless Python is a small but exciting step in the
> direction of improving infrastructure. I wish, instead

Sure, but it's not particularly oriented to improving
speed, is it?  Rather, allowing good exposure of new
abstractions (generators, coroutines, micro-threads)
seems to be what it's all about.


> the abstractions it implements. For 10% penalty, I'll use
> list comprehensions, with the expectation that in a year
> or two, I'll gain an X% bonus. (And if we don't? What
> good does it do to have list comprehensions if programmers
> routinely use loops instead, because they know it will
> get them an easy 10% performance boost? The goal is not
> just to provide nice abstractions, but to seem them used!)

I don't think a 10% difference matters substantially, either
way!  _This_ may be the crux of our disagreement.  Sure,
if the infrastructure can be enhanced to get a few percents
here and there, why not.  But that's not going to change my
programming style, nor make much of a difference to the
effectiveness of the applications I write.  And if some coder
_in a higher-level language_ is letting "few-percent issues"
sway his or her style away from clarity and simplicity, then
I think the solution is _education_... explaining exactly how
and why this is a losing trade-off.

*Orders of magnitude* in performance DO make a huge
difference -- but I don't think this is what we're talking
about.  Well, we _might_, I guess, when a list is being
built up by .append calls, with frequent reallocations of the
list -- *that* is something to watch out for, O(N*N) cost
where O(N) is easily achievable.  But if the infrastructure
was somehow enhanced to detect the pattern of 'many
appends in a loop' and switch the internal structure to
something more performing, this would optimize the
'lower-level' (item-by-item) approach, NOT the higher
abstraction level one.  (Actually, I suspect that adding a
couple of new types, a stringbuffer and a listbuffer, that
ARE optimized for speedy in-place modification, to be
used during 'construction phases' then transformed back
into basetypes -- strings and lists -- is an easy approach
for that; however, this moves *away* from abstraction, so
I doubt it would meet with your approval...:-).


> One BIG benefit to abstractions like JOIN and list
> comprehension is that performance improves without the
> application programmers doing anything. The next release
> of the programming environment makes everything run

But why shouldn't said 'next release' be able to optimize
the 'lower-abstraction' approach, too?  As long as the
language allows and encourages the coder to express
the code's intent exactly and precisely, optimizations are
feasible anyway.  In other words, I do NOT consider
potential optimizations a 'BIG' benefit of higher-level
abstractions over lower-level ones.

> faster. Or so one hopes. This is well understood in the
> DBMS world. We teach beginning SQL programmers to use
> the constructs, even if they think they can beat the
> current implementation. That's what we also ought to
> teach Python programmers, right?

Right, but NOT because of performance considerations!
*Clarity and simplicity*, and the resulting gain in coding
and maintenance productivity, is what we should be
focusing on (hopefully removing *order-of-magnitude*
performance bottlenecks that can result when coding is
*naive* rather than just *simple, direct and clear*!-).


> > Actually, for all-out speed in a higher-level language, ..
>
> Why restrict ourselves to higher-level languages? For all
> out speed with portability, we could program in C.

Yes, but at LARGE productivity costs, if we used C for
_all_ of our coding, rather than just a very few critical
hot-spots.  The productivity costs of using Scheme or
O'Caml instead of Python aren't anywhere close to those
C would impose, in a single-language setting; which
is *why* it makes most sense to 'restrict ourselves to
higher-level languages', for most of our coding.


> > I prefer Python because I do NOT prize speed in a
> > higher-level language above other factors -- great
> > practicality at "playing well with others", clarity,
> > cleanness of syntax, great productivity.
>
> I would explicitly list "working at a higher level of
> abstraction."

When I compare working in Python with working in
Haskell, say, or perl, or, I _think_, ML dialects (not
enough practical real-world experience to be sure!),
it does not seem to me that I'm working at a higher
level of abstraction; the abstraction-level seems quite
comparable.  (Actually, this goes for C++, as well,
most of the time -- but, admittedly, lower-level issues
_do_ keep interfering often enough in that case).

Rather, I find Python's key advantages to be the ones I
listed above.  Wrt Haskell, for example, I notice much
more ease of integration with other parts of my world,
and higher productivity thanks to latent typing; wrt
perl, higher clarity and cleanness, and much higher
productivity in maintenance; and so on.

> Many things improve productivity, some
> of which Python lacks. There is a lot to be said for a
> highly integrated programming environment, as one finds
> in Delphi. But the tie to a specific GUI framework
> limits portability. I am glad that Python made the
> choice of simplicity and abstraction.

Me too!  Highly-integrated environments can always come
later -- simplicity and clarity remain.  'Abstraction' and
'higher-level' are basically synonyms, and I don't find
Python too different from, say, Scheme, in these terms.


> > If you do have another scale of values, Python may not
> > be the best choice for you at this time; specifically,
> > if for some reason the solutions MUST be single-language
> > ones (you can't or won't use a lower-level, faster
> > language for that 10% or so of your code that bottlenecks
> > its speed).  I do not have any such constraint, so Python
> > + some tiny amount of C++ at the right places .. is
> > exactly right for my needs...
>
> This introduces friction that undermines precisely Python's
> greatest selling point: increase in productivity. The
> problem is NOT just programming in C the parts of an
> application that are bottlenecks, but first identifying the
> bottlenecks. That's not too hard for those of us with
> experience at it; more difficult for many. And keep in mind
> that the solution may not be programming in C, but rewriting
> in longer-winded Python, which thereby becomes less readable
> and understandable, now undermining maintainability.

'Longer-winded Python' will not buy you order-of-magnitude
performance enhancements (unless some 'naive', O(N*N),
approach, is being turned into O(N) or at least O(N logN)
code -- and in that case, it would be hard to argue much
against 'long-windedness':-).

I don't think the art, or science, of identifying bottlenecks
and coding them away will disappear in my lifetime.  Nor
will a few percents' worth of optimization make much
difference in this regard.


> My bottom line is this: I don't think we should make people
> choose either-or. Either take the level of abstraction
> provided by Python (and solve performance problems after
> the fact) OR use something else. Python CAN BE faster. There

Sure -- by a few percents here and there.  Worthwhile.  Hardly
major, I think.

> is nothing in the language that forces list comprehensions
> to run slower the simple lists.

I think it boils down to allocation issues; as list comprehension
are now implemented in terms of a loop of appends to an
initially-empty list, that is going to cost compared to having
it allocated all at once.  Having a temporary mode, for lists
being built up, where more memory is invested (exponential
allocation, with constant amortized cost per item, rather than
linear allocation, with O(N) amortized cost per item) -- or the
list is at first built up in discontiguous memory, then the compact
block actually needed is obtained once only at the end, when
it's clear exactly how much is needed -- might be big wins in
performance terms (the 'listbuffer' ideas I mentioned above).

But if Python had such a 'listbuffer' object, and used it internally
for list-comprehensions, I'd _also_ like to see it exposed for
direct programmer use... why not?  I don't necessarily want to
have to *distort* my code into a list-comprehension, in cases
where some other approach might be clearer, as the only way
to get performance!

Say I want to build up and return a list of Fibonacci numbers,
however many are needed to get up to the first one that is
larger than an argument N.  The simplest approach:

def listFib(N):
    if N<1: return [1]
    result = [1, 1]
    next = 1
    while next <= N
        next = result[-1]+result[-2]
        result.append(next)
    return result

Making this into a list-comprehension would contort things... it
usually does, when each item in the list is defined in terms of
items before it.  Rather, if listbuffer objects existed, I might
get some substantial performance gain through them, without
(IMHO) affecting readability much, if at all:

def listFib(N):
    if N<1: return [1]
    result = listBuffer.listBuffer(1, 1)
    next = 1
    while next <= N
        next = result[-1]+result[-2]
        result.append(next)
    return list(result)

I'm all in favor of abstraction, of course (and motherhood and
apple pie, too!), but I don't want to force-fit it into my code...
as I've seen done too often in higher-level languages that
thought they could *force* programmers to use abstraction
by NOT providing necessary lower-level primitives!-).

> Living with this either-or
> makes sense early in a technology's lifecycle, when the
> abstractions are new and getting them out is more important
> than getting them fast. But they're out. It is time to shift
> the emphasis a bit, from features to infrastructure.

I won't _mind_ infrastructure-enhancements... but they won't
take the place of identifying and recoding key hot-spots, or,
at least, such is my current prediction.


> I might be persuaded otherwise if you would point out the
> BIG feature enhancements that are just around the corner.
> Right now, I don't see them. So what do I want next, more
> than anything else? More speed. And availability on all
> the palm platforms.

What *I* want (no idea if it's coming), is more and more ways
to express in my code all that I *know* about the semantics
involved.  "This object here satisfies the XYZ interface", "this
sequence here is sorted", "this function here is a total
ordering on yonder domain", etc.  Such 'rich assertions' may,
or may not, help debugging my code OR make it faster -- I'd
like a compilation mode where the compiler inserts checks for
all assertions it can't prove at compile-time (to help me ensure
I do indeed know what I'm talking about), and another where
it _optimizes_, *trusting* my assertions, and taking advantage
of anything it can infer from them.  But I wouldn't want the
rich-assertion-sublanguage constrained to stuff the compiler is
actually able to use (or check) _now_ -- I'd rather be able to
'code' existential and universal qualifiers, which are what I do
often need to 'express all that I know', and worry about having
effective checks or use of them (the infrastructure:-) later on.

This would get me more clarity and expressiveness now, and a
_possibility_ of enhanced checking and/or higher optimization at
some future time.  And wouldn't that be great?-)

As far as infrastructure goes -- better development environments
(e.g., the planned 'nanny'/'lint-ish' facility; more thorough checks
of coverage in unit-test frameworks; perhaps an ability to ask the
runtime [maybe in a special version] to selectively 'cause' certain
exceptional conditions -- simulate no-memory, disk-full, etc -- so
I can more easily and thoroughly check my exception-handling),
and standard-library enhancements, still seem to me to need to
be given priority over sheer-speed issues.  Not that I'd spit at a
few percents here, a few percents there, mind you -- the greatest
advantages of such small enhancements, however, would be to
give greater inducement to upgrade-to-newest-versions to shops
that let themselves be guided by such minor issues as those:-).


Alex





From root at [127.0.0.1]  Mon Dec 25 21:22:57 2000
From: root at [127.0.0.1] (deadmeat)
Date: Tue, 26 Dec 2000 02:22:57 GMT
Subject: variable scope?
References: <c5S16.8616$KY1.23318@news1.rivrw1.nsw.optushome.com.au> <928o67$9qj$1@paradise.nirvananet>
Message-ID: <5iT16.8620$KY1.23139@news1.rivrw1.nsw.optushome.com.au>

> it's all in the manual

which is hopeless (well, the online docs that is)..  maybe i'm just used to
professional doumentation. time to fork out $60 for another dead-tree
O'Reilly novel :)

>     global year

ah, duh. i should've guessed that one :)





From cnelson9 at gte.net  Wed Dec 20 17:30:15 2000
From: cnelson9 at gte.net (Clifford J. Nelson)
Date: Wed, 20 Dec 2000 22:30:15 GMT
Subject: Python in "math through programming" curriculum
References: <2g2t3t06r2agrrjmfnr41eu4lc2u39a8qi@4ax.com> <3A3EFF47.67DE69CF@gte.net> <kj624toonmgugtrgrpkea9b01a56hgogam@4ax.com>
Message-ID: <3A413A14.54DF58D1@gte.net>


Kirby Urner wrote:
> 
> "Clifford J. Nelson" <cnelson9 at gte.net> wrote:
> 
> >I don't think Python should be a standard at all, because of
> >my experience with it on the Mac. Keep the options open for
> >computer programming languages.
> >
> >Here are two:
> >
> >Clean at: http://www.cs.kun.nl/~clean/
> 
> I went to this website.  It doesn't work for the Mac at all --
> Windows only.  So what's your beef with Python on the Mac?
> At least it exists for the Mac.

That's why I don't attach ANY weight to what you write!
Clean used to be a "Mac first product", and Version 1.3.x
soon to be updated to Version 2.0 runs on the Mac. You would
know that it runs on the Mac if you could just read.

> Also:  "Allthough Clean is not easy to learn (it's very
> different from imperative languages, such as Pascal and C),
> once you can program in Clean, you can easily create games
> with the Clean Game Library."
> 
> Do you think this emphasis on game programming is better
> than my "math through programming" approach?  It seems to
> me that writing games is (a) a rather specialized goal
> (vs. learning generic math ideas) and (b) a domain in
> which it's difficult to compete with already-existing
> commercial products (i.e. if your goal is to become a
> game programmer, vs. competent in basic math, you have a
> long road ahead of you).
> 
[snip]
I take it personally when Bucky Fuller's Synergetics (see 1)
is made to seem like some far out religion, by you Kirby
Urner, or anyone else, or when computer programming in a
language (Python) that can't apparently help itself well
enough to use it to make a good interactive programming
environment for the Mac, is promoted over much easier
languages, like Clean (see 2) or Object Caml 3.0 (see 3) or
even the hard to learn J programming language (see 4).

All of those languages (including Python) are garbage
compared to Mathematica of course, because programming (OOP)
is all just really pattern matching and replacement (see 5),
but they're free and they work on both the Mac and Windows,
so they are worth looking at.

1) http://www.rwgrayprojects.com/synergetics/synergetics.html
2) http://www.deene.ufu.br/clean/clean1.html
3) http://pauillac.inria.fr/caml/
4) http://www.jsoftware.com/
5) http://www.stephenwolfram.com/publications/talks/IMS/imstalk.html

> 
> My guess is you've done approximately zero using any of
> the languages under discussion (Python, Clean, Caml).  If
> I'm wrong, I hope you will show me otherwise.  I don't
> have a lot of time for completely uninformed discussion.
> 
> Kirby

You could inform yourself, if you could only read.

  Cliff Nelson


From emile at fenx.com  Fri Dec 22 09:37:10 2000
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 22 Dec 2000 06:37:10 -0800
Subject: while (thread.alive()):
References: <mailman.977283904.26696.python-list@python.org> <91tgvm$3q8$2@taliesin.netcom.net.uk> <usnngy396.fsf@roundpoint.com>
Message-ID: <91votl$5ie90$1@ID-11957.news.dfncis.de>


"Ben Hutchings" <ben.hutchings at roundpoint.com> wrote in
message news:usnngy396.fsf at roundpoint.com...
> "Garry Knight" <garryknight at bigfoot.com> writes:
>
> > In article
<mailman.977283904.26696.python-list at python.org> "Jay
> > O'Connor" <joconnor at cybermesa.com> wrote:
> >
> > > I'm continuously amused that people try to take
language design
> > > forward and still drag obsolete ideas along for the
ride.
> >
> > Like having a print command that doesn't, by default,
send output
> > to the printer...  :o)
>
> I think that bug was introduced by the shift to glass
terminals.
> Would you like to use a Teletype instead?


Well, then at least return would mean something.  ;-)

But-I'd-prefer-a-TI710-hardcopy-terminal-ly y'rs

--

Emile van Sebille
emile at fenx.com
-------------------





From dnew at san.rr.com  Wed Dec 20 18:57:14 2000
From: dnew at san.rr.com (Darren New)
Date: Wed, 20 Dec 2000 23:57:14 GMT
Subject: How to make a GUID with plain python
References: <3a413a17.3416402@192.168.0.1>
Message-ID: <3A41475A.5DC4ABD8@san.rr.com>

> A plain python algorithm itself would be very useful, since one could
> use it for RPC / COM as well as for creating indexes for distributed
> databases.

Well, concatenating one of the IP addresses assigned to the machine to the
clock would give you something that should be globally unique, if you had an
IP address that wasn't unroutable. Generate a random number to tack on the
end, and you're set?

-- 
Darren New / Senior MTS & Free Radical / Invisible Worlds Inc.
San Diego, CA, USA (PST).  Cryptokeys on demand.
Personal malapropism generator free with purchase!
Steganography: The manual is hidden in the source code.
Gold mining is only profitable because of the gold found.


From rayvd at nospam.firetail.org  Wed Dec  6 03:03:05 2000
From: rayvd at nospam.firetail.org (Ray Van Dolson)
Date: Wed, 06 Dec 2000 08:03:05 GMT
Subject: SWIG and int main()
References: <_3QV5.397$T02.19944@sjc-read.news.verio.net> <t2g2etldugf1ac@corp.supernews.com> <8cXV5.530$T02.23604@sjc-read.news.verio.net> <slrn92i3tf.8tq.jepler@potty.housenet>
Message-ID: <ZomX5.1331$T02.60321@sjc-read.news.verio.net>

jepler at inetnebr.com (Jeff Epler) wrote in
<slrn92i3tf.8tq.jepler at potty.housenet>: 

>I suspect that you may be calling Python's main(), rather than your
>main(). 
>
>potty% ARGV0=./hi python hi
>./hi: can't open file 'hi'
>
>... and then python calls the C exit() which will terminate without
>letting Python catch an error or anything.
>
>You will need to rename your main() function, perhaps using the C
>preprocessor to do the trick:
>     cc -Dmain=module_main test.c -o test.o
>will rename main to module_main .. just change the corresponding name in
>your swig .i file.
>
>Jeff

Thanks Jeff, this did indeed fix the problem!  Probably wouldn't have 
figured that one out on my own for a while. :)

Ray


From dsh8290 at rit.edu  Thu Dec 21 23:55:08 2000
From: dsh8290 at rit.edu (D-Man)
Date: Thu, 21 Dec 2000 23:55:08 -0500
Subject: Speed of Python 1.5.x vs 2.0
In-Reply-To: <4vj54t4v0hvjsnstfmtmf36q9irotp59sj@4ax.com>; from tyler@tylereaves.com on Thu, Dec 21, 2000 at 10:53:32PM -0500
References: <4vj54t4v0hvjsnstfmtmf36q9irotp59sj@4ax.com>
Message-ID: <20001221235508.C20586@westsidecnc.rh.rit.edu>

RH7 huh?  Well, that's not a good sign.  I don't know anything about
the speed comparison other than python 2 isn't ridiculously slow.  But
RH decided to ship an unstable compiler with their distro so you can
have extra fun building things.  

To build python 2.0 on my RH7 system (on a Intel PII) I had to first
hand-fix /usr/include/bits/xopen_lim.h to set the LONG_BIT and
WORD_BIT macros properly.

HTH,
-D

On Thu, Dec 21, 2000 at 10:53:32PM -0500, Tyler Eaves wrote:
> 	I'm soon going to attemp to talk my webhost into upgrading to
> 2.0, but I figure the only way I'm going to be able to make that
> happen is if I can show results that show 2.0 being faster / more
> effcient than 1.5.x. Their servers run on Red Hat 7.
> 
> Thanks for any help.
> 
> ---
> Tyler Eaves



From pbjorn at uswest.net  Thu Dec 21 21:23:11 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Thu, 21 Dec 2000 19:23:11 -0700
Subject: reading numbers from file
References: <slrn9457uq.9tl.jp@localhost.localdomain>
Message-ID: <3A42BB0E.11699DDF@uswest.net>

Here's a start. It only deals with comments when reading a line at a time, and I
haven't tested the edge cases, but it should at least get you going...

-- bjorn

import string

class NumberReader:
 def __init__(self, file):
  self.fp = file

 def readNumberLine(self):
  line = ''
  while 1:
   line = self.fp.readline()
   if not line:
    break
   if line.strip() == "":
    continue # skip empty lines
   if line.strip().startswith('#'):
    continue # skip comment lines
   return map(int, line.split())

 def readNumbers(self, n=5):
  res = []

  for i in range(n):
   word = ''

   # skip whitespace
   char = self.fp.read(1)
   while char in string.whitespace:
    char = self.fp.read(1)

   # read next word (number)
   while char not in string.whitespace:
    word += char
    char = self.fp.read(1)

   res.append(int(word))

  return res

def test():
 import StringIO
 f = StringIO.StringIO("""
 23 41 24 0 2 42 532 53 234
 34 234 53 23 4 323 53 345
 0 32 64 128 256 512 1024
 """)
 nr = NumberReader(f)
 a = nr.readNumberLine()
 print a
 print nr.readNumbers(3)
 print nr.readNumbers(5)

if __name__=="__main__":
 test()


Jacek Pop?awski wrote:

> I have file with numbers:
>
> 23 41 24 0 2 42 532 53 234
> 34 234 53 23 4 323 53 (etc...)
>
> 1.How to read one line of this file and write it to list of numbers?
>   (like: a=[23,41,24...])
> 2.How to read only 5 numbers, write it to list, then another 5, and write
>   it to second list?
> 3.What will change if I want to add comments (#) and empty lines to my file?
>
> --
> Don't care which God you follow
> Whose promises you swallow
> Time and again
> We must meet at the end              "Lord of the Last Day" - Ronnie James Dio



From cfelling at iae.nl  Fri Dec 15 10:13:57 2000
From: cfelling at iae.nl (Carel Fellinger)
Date: 15 Dec 2000 16:13:57 +0100
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com>   <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <tcmay-FCA22F.14073713122000@news.got.net> <918v7a02scn@news1.newsguy.com> <tcmay-04E80C.16181313122000@news.got.net> <91a1ed02a4j@news2.newsguy.com> <3A389B1A.4F33FFD4@cfmu.eurocontrol.be> <91ae8s02llf@news2.newsguy.com> <3A397778.2E6D84D3@cosc.canterbury.ac.nz> <M0f_5.27169$x6.14244092@news2.rdc2.tx.home.com>
Message-ID: <91dcfl$39n$1@animus.fel.iae.nl>

Rainer Deyke <root at rainerdeyke.com> wrote:

> (where Singleton is an instance of the (meta-)class that implements
> singletons, not listed here because I haven't written or found it yet)

I once tried the mixin approach, but was left with to many loose ends
so I would go for the factory approach like:

'''Singleton Pattern

Using the singleton factory each class keeps track of its own singleton,
unless some base class had the magic attribute "__singleton" set to [].
In that case all derived classes share their singleton; so all instances
in use better be in a sole inheritance line from that base class up and
you better take care to instantiate the last class in that particular
derivation chain first!
'''

def singleton(Class):
    if hasattr(Class, "__singleton"):
        a = Class.__singleton
        if a == []:
            a.append(Class())
        return a[0]
    else:
        name = '_%s__singleton' % Class.__name__
        if not Class.__dict__.has_key(name):
            Class.__dict__[name] = Class()
        return Class.__dict__[name]


#singleton=instance

if __name__ == "__main__":
    class A: pass
    class AA(A): pass

    setattr(A, "__singleton", [])
    assert singleton(AA) is singleton(A)
    assert singleton(A) is singleton(A)
    assert singleton(AA) is singleton(AA)

    del A.__singleton
    assert not singleton(AA) is singleton(A)
    assert singleton(A) is singleton(A)
    assert singleton(AA) is singleton(AA)
-- 
groetjes, carel


From tim.one at home.com  Sun Dec 31 16:43:22 2000
From: tim.one at home.com (Tim Peters)
Date: Sun, 31 Dec 2000 16:43:22 -0500
Subject: Is 100,000 entries big for a dictionary?
In-Reply-To: <92iug4$vqb$1@nnrp1.deja.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCECDIGAA.tim.one@home.com>

[posted & mailed]

[rturpin at my-deja.com]
> I'm in the process of architecting an application with medium
> sized data sets. Python dictionaries are tempting as part of
> the representation mechanism. How do they perform when they
> have 100,000 entries?

Provided you have sufficient RAM, the theoretical performance (for lookup,
insert and delete) remains O(1) expected case, O(N) worst case, regardless
of dict size.  This is also what I've seen in practice:  Python ensures that
at least 1/3 of the internal hash table entries are unused (it uses an
extreme form of open addressing rather than buckets of linked lists), so
collisions are very rarely a problem.  It's possible to contrive keys that
will cause collisions systematically, but not easy -- unlikely to happen by
accident in 2.0.

If you don't have enough RAM, and collisions are frequent, performance can
be atrocious.  Curiously, no way is provided for you to find out whether you
*are* suffering frequent collisions!  If you suspect you are, (a) you need
to instrument dictobject.c; and, (b) you're probably wrong <wink>.

if-you-do-have-frequent-collisions-i'll-consider-it-a-bug-ly y'rs  - tim




From scarblac at pino.selwerd.nl  Fri Dec 15 16:57:20 2000
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 15 Dec 2000 21:57:20 GMT
Subject: __repr__ policy problem: should module be prefixed to output?
References: <3A3A3A6D.2C3B1A91@sbc.su.se> <91dfl0014sa@news1.newsguy.com>
Message-ID: <slrn93l4p4.v6.scarblac@pino.selwerd.nl>

Alex Martelli <aleaxit at yahoo.com> wrote in comp.lang.python:
> "Per Kraulis" <per at sbc.su.se> wrote in message
> news:3A3A3A6D.2C3B1A91 at sbc.su.se...
> > However, I have realized that there is a policy problem. Should __repr__
> > add the module name as prefix to the output string, or not? Or is there
> > some other solution?
> 
> I have exactly the same problem in gmpy (I do supply a portable
> binary picklable string, but also a repr output that is eval'uable
> to restore something 'close to' the original).

Personally I think that this problem isn't that important. Pick one,
be consistent, and it'll be ok. If you start thinking about this sort of
thing a lot, nothing will get done because there are important choices
everywhere!

I went with the module name in __repr__, because of the tiny chance that
I have two classes with the same name in different modules and I want to
be able to tell them apart in the interpreter. But then, without the
module name, the strings are shorter. It doesn't matter much.

-- 
Remco Gerlich


From dale at out-think.NOSPAMco.uk  Mon Dec 11 14:08:40 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Mon, 11 Dec 2000 19:08:40 +0000
Subject: reloading modules question
References: <90scdf$hpe$1@troll.powertech.no> <ouk43t494isd9m8fk92udjjbg2pb6crq7q@4ax.com> <910om3$2vk$1@troll.powertech.no>
Message-ID: <kg9a3t0s6ah2c7t3pf94o3vl4ls2hespuc@4ax.com>

"Syver Enstad" <syver at NOSPAMcyberwatcher.com> wrote:

>> I would suggest you are better off just stopping and starting the affected
>web sites. That's what we
>> do and it only takes a second.
>
>You mean something like pushing the unload button in the applications
>property sheet?
>Starting and stopping the default web site in inetmgr is not enough.
>
>> You can automate the procedure using WMI or ADSI. I'm sure I used to have
>an example script but I'm
>> buggered if I can find it right now.
>I think I saw some documentation on it in the documentation for IIS. I'll
>look it up.
>
>
Starting and stopping the web site has always worked for us. (Windows 2000 Server)

--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants


From root at rainerdeyke.com  Thu Dec  7 16:59:11 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Thu, 07 Dec 2000 21:59:11 GMT
Subject: Help/example of bit testing
References: <90llq7$8fc@dispatch.concentric.net> <B6556369.1DF7A%fgranger@teleprosoft.com>
Message-ID: <PKTX5.346$x6.531284@news2.rdc2.tx.home.com>

"Fran?ois Granger" <fgranger at teleprosoft.com> wrote in message
news:B6556369.1DF7A%fgranger at teleprosoft.com...
> genMasq = [1, 2, 4, 8, 16, 32, 64, 128, 256, 512, 1024, 2048, 4096, 8192,
> 16384, 32768]
>
> data = 198
>
> for i in range(7):
>     if data & genMask[i]:
>         print "bit" + str(i) + "is on"

for i in range(7):
  if data & (1 << i):
    print "bit" + str(i) + "is on"


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From haircut9 at my-deja.com  Thu Dec  7 11:58:10 2000
From: haircut9 at my-deja.com (haircut9 at my-deja.com)
Date: Thu, 07 Dec 2000 16:58:10 GMT
Subject: help with ncurses - if anyone can give me a simple ncurses program
References: <3a14053a.0@news1.mweb.co.za> <3dsnoriu62.fsf@kronos.cnri.reston.va.us>
Message-ID: <90ofiv$l8s$1@nnrp1.deja.com>

In article <3dsnoriu62.fsf at kronos.cnri.reston.va.us>,
  Andrew Kuchling <akuchlin at mems-exchange.org> wrote:
> "Hanna" <juliaj79 at hotmail.com> writes:
> > find out more about structure. I am not sure how to go about it. Can
anybody
> > who can program in ncurses send me example programs?
>
> You did look at http://www.python.org/doc/howto/curses/curses.html,
right?
>
> I'd like to add a Demo/curses directory to the Python CVS tree.  If
> anyone has reasonably small programs that use the curses module,
> please let me know.
>
> --amk
>

I'm working on an automatic uploader using ncurses. Actually it's my
first project in Python but I was just curious how it worked. But might
give some insights too.
The Python2 docs helped a lot, though


Arthur
--
+-+
"Ahh, tireur isol?. Obtenez la baguette magique! Utilisez la baguette
magique!"
+-+


Sent via Deja.com http://www.deja.com/
Before you buy.


From paul.moore at uk.origin-it.com  Thu Dec 21 11:42:44 2000
From: paul.moore at uk.origin-it.com (Paul Moore)
Date: Thu, 21 Dec 2000 17:42:44 +0100
Subject: Events in win32com
References: <Y1g=Op6IfWSoJn1ISFtOyuvqwBb3@4ax.com> <FY0=OlZFLy71DtnK62u5TEPlhOts@4ax.com> <3A40060B.10504@ActiveState.com> <cYxAOmxi9FgNERgBo29zuCl5c2h1@4ax.com> <3A418B5B.5010001@ActiveState.com>
Message-ID: <7DJCOoKRJvE3IIYAXXveM7omaTC6@4ax.com>

On Thu, 21 Dec 2000 05:52:29 +0100, Mark Hammond
<MarkH at ActiveState.com> wrote:

>Paul Moore wrote:
>
>> Am I doing something wrong, or is this not expected to work?
>
>It is expected to work, and I can't see what you are doing wrong.
>
>If you can work out a very simple case that fails (ie, suiltable for 
>inclusion in the test suite) I would look at it, but without knowing the 
>exact signatures in question, or knowing if Python is finding the sigs 
>in the makepy file etc, I really have no idea...

I'll see what I can do. The problem is that my small test case
currently relies on a non-standard control. I'll see if I can get a
test in IE or Excel or something...

Paul.



From donn at u.washington.edu  Tue Dec 19 12:56:16 2000
From: donn at u.washington.edu (Donn Cave)
Date: 19 Dec 2000 17:56:16 GMT
Subject: __repr__ policy problem: should module be prefixed to output?
References: <3A3A3A6D.2C3B1A91@sbc.su.se> <91dfl0014sa@news1.newsguy.com> <Pine.LNX.4.10.10012152140170.10103-100000@kev.ioc.ee> <Pine.LNX.4.10.10012160011400.14411-100000@kev.ioc.ee> <slrn93tjv5.h7.amk@207-172-146-173.s173.tnt3.ann.va.dialup.rcn.com> <slrn93ud1g.5ot.scarblac@pino.selwerd.nl>
Message-ID: <91o7g0$dec4$1@nntp6.u.washington.edu>

Quoth scarblac at pino.selwerd.nl (Remco Gerlich):
| A.M. Kuchling <amk at mira.erols.com> wrote in comp.lang.python:
...
|> Frankly, the only place I ever use repr() is in debugging print
|> statements and log files, where I need to be very clear about what an
|> object is and avoid confusing the integer 1 with the string '1'.  I
|> therefore write __repr__ on classes to be as useful for debugging
|> purposes as possible; it doesn't recurse into subobjects, but does
|> include the class name, the id() of the object, and any other unique
|> identifier for the object (an ID number, user name, whatever).
|> 
|> To put my money where my mouth is, I'd like to file a documentation
|> bug about this.  What bits of the Python documentation promote this
|> view of repr()?  One is section 3 of the reference guide, in its
|> description of the __repr__ special method name; any others?
|
| What is the bug, precisely? You never use it?

The bug is that some documentation promotes a meaning for repr()
that is neither generally valid nor practical.

| The language reference has it when in the description of __repr__, and
| the library reference in the description of repr().
|
| This is only really an issue with instances, since you get to make your
| own __repr__ for those. And in my experience, it is usually quite possible
| to return a string that could be used to build the object again (for instance,
| by giving the right arguments to the constructor). Of course, if you
| have objects with lots of subobjects and so on, it may not be feasible
| and you don't use it.
|
| But I still like the rule. And, well, it's what repr does.

That's the problem, though, it is not what repr does, generally.  What
use is a rule that has never been and will not be generally observed?
Python class instances in the library, C objects like compiled regexps.  
>From the implementations that come with the distribution, it's pretty
clear to me that repr serves the purpose described above, a programmer's
printout, and that's an essential function too.

	Donn Cave, donn at u.washington.edu


From knotwell at ix.netcom.com  Thu Dec 28 12:43:58 2000
From: knotwell at ix.netcom.com (Brad Knotwell)
Date: 28 Dec 2000 09:43:58 -0800
Subject: Erlang (was Re: Good introduction to functional programming with Python?)
References: <mailman.977984412.30211.python-list@python.org> <92f0il01a5u@news1.newsguy.com>
Message-ID: <86wvckqw4h.fsf@localhost.my.domain>

"Alex Martelli" <aleaxit at yahoo.com> writes:
> It sure looks unique to me for its emphasis on concurrency (if there
> are other FP's that are similarly concurrency-oriented, I have not seen
> them yet).  But Scheme (albeit impure) and Erlang (albeit 'culturally
> not highly HOF-prone') together surely make up for a substantial share
> of functional-programming use, and just as surely form a solid case for
> static-typing NOT being central to FP (just as, say, Smalltalk and
> Python together form a solid case for static-typing not being central
> to OO, no matter how much Eiffel'ers may rant otherwise:-).

John Reppy's concurrent extensions to SML/NJ (CML) are delivered as part of 
the standard distribution.  Furthermore, his book "Concurrent Programming in 
ML" covers these extensions in detail.  What I don't know is 

FWIW, I believe the Glasgow Haskell compiler has concurrent extensions as well,
but I've no idea if they're delivered with the runtime environment or not.

Interestingly enough, finishing Reppy's book caused me to redesign a 
sequential Python program into a bunch of threads.  Overall, the design
became much clearer, but time will tell if it's as robust as the previous
program.

> Alex

I'm now wondering if Stackless Python's uthreads are cooperatively or
pre-emptively multi-tasked.

--Brad



From e_d_k at my-deja.com  Tue Dec 26 16:42:15 2000
From: e_d_k at my-deja.com (EdK)
Date: Tue, 26 Dec 2000 21:42:15 GMT
Subject: Can't install any packages
References: <92acdv$eid$1@nnrp1.deja.com> <92ap0f02b6r@news1.newsguy.com>
Message-ID: <92b3bn$dq$1@nnrp1.deja.com>

> A .PYD file is actually a DLL, renamed to avoid confusing it with
> "normal" DLL's.  I suspect the packages you download support
> Python 2.0, which came out so soon after 1.6; upgrading to 2.0
> might solve it (unless some very old packages only support the
> 1.5 version).  1.6 ruled the roost for such a short time (its
> final was out the same day as 2.0's beta!) that few or no pre-built
> binaries are likely to support it.

Thank you. I will upgrade.

   -EdK


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


From auto103690 at hushmail.com  Sat Dec 30 23:41:06 2000
From: auto103690 at hushmail.com (Jeff)
Date: Sun, 31 Dec 2000 04:41:06 GMT
Subject: Running Boa Contructor with the latest versions of Python & wxPython
References: <92ld65$4un$1@news.nuri.net>
Message-ID: <CNy36.80153$df5.2861912@news1.crdva1.bc.home.com>

A non-text attachment was scrubbed...
Name: not available
Type: text/plain  charset=iso-8859-1
Size: 1393 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20001231/44c1e3af/attachment.bin>

From fxital at aol.comnospam  Wed Dec  6 10:30:56 2000
From: fxital at aol.comnospam (FxItAL)
Date: 06 Dec 2000 15:30:56 GMT
Subject: Help Finding Time Module
Message-ID: <20001206103056.01096.00002956@ng-cv1.aol.com>

Hello All,

      I'm having trouble creating an executable file from my script and need to
find the time module. If I search for "time.*" there are no files found but, if
I use the command "from time import *" in IDLE there are no errors and
"localtime(time())" returns an accurate value. Pease help me if you can.

Python 2.0 on Windows ME

Thanks Al


From dsh8290 at rit.edu  Thu Dec 21 14:26:36 2000
From: dsh8290 at rit.edu (D-Man)
Date: Thu, 21 Dec 2000 14:26:36 -0500
Subject: A Suggestion for Python Colon Syntax
In-Reply-To: <91tfg1$282$1@hecate.umd.edu>; from billtj@y.glue.umd.edu on Thu, Dec 21, 2000 at 05:43:29PM +0000
References: <91tc8h$2pp$1@hecate.umd.edu> <W2r06.1076$AH6.86249@newsc.telia.net> <91tfg1$282$1@hecate.umd.edu>
Message-ID: <20001221142636.B2861@westsidecnc.rh.rit.edu>

On Thu, Dec 21, 2000 at 05:43:29PM +0000, William Djaja Tjokroaminata wrote:
> Fredrik Lundh <fredrik at effbot.org> wrote:
> : "William Djaja Tjokroaminata" wrote:
> :> However, a lot of time when I ran my Python program, I got syntax
> :> error, because I forgot to put colon (:) after my if (or for, etc.) clause.
> :> My question is, is the colon really necessary, syntactically, in Python?
> 
> : Short answer:  The colon isn't there for the computer, it's
> : there for human readers.  Just like the colons in your post:
> 
> Well, the colons in my post is because I am writing in English.  But
> Python does not have semicolon etc, just to make it "pretty", but they do

Umm, actually you can put a semicolon at the end of each line (or only
some lines) and it will behave like the semicolon in C/C++/Java/Perl.

It is really useful (needed) if you wanted to write 2 statements on 1
line.  I don't know why that would be needed, but . . . 


-D



From ian at cfmu.eurocontrol.be  Thu Dec 14 05:02:37 2000
From: ian at cfmu.eurocontrol.be (Ian Wild)
Date: Thu, 14 Dec 2000 10:02:37 GMT
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <tcmay-FCA22F.14073713122000@news.got.net> <918v7a02scn@news1.newsguy.com> <tcmay-04E80C.16181313122000@news.got.net> <91a1ed02a4j@news2.newsguy.com>
Message-ID: <3A389B1A.4F33FFD4@cfmu.eurocontrol.be>

Alex Martelli wrote:
>  
> I can't find a source any more, but one of the most interesting
> arguments I've heard for "only inherit from abstract classes"
[...otherwise...]
> it would break the idea that your concrete
> classes partition the universe of discourse, if it were.

Where does this partitioning requirement come from?
 
> This is based on the idea of subclassing as asserting IS-A.

?

It seems to me you'd need an "(e) None of the above"
class very frequently if your world doesn't happen to have
any genuine structural divisions.

Dragging in that old favourite, "Squares and Rectangles",
your partitioning clause means I'd have to invent a new class,
"NonSquareRectangle", adding no interesting features, just so
I can have aSquare behave like aRectangle.  This seems like
unnecessary work, just to keep some C++ guru happy.


From MarkH at ActiveState.com  Thu Dec 28 19:51:47 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Fri, 29 Dec 2000 00:51:47 GMT
Subject: automating excel with python ... getting excel to go away
References: <baP26.99236$yR4.2642456@news1.rdc1.tx.home.com>
Message-ID: <3A4BDEFD.8000603@ActiveState.com>

Scott Hathaway wrote:

>  myExcel.Quit

Quit is a method - try:

myExcel.Quit()

Mark.



From lists at TheWebsons.com  Sat Dec 23 16:21:10 2000
From: lists at TheWebsons.com (Ben Ocean)
Date: Sat, 23 Dec 2000 13:21:10 -0800
Subject: Why Won't This Work?
In-Reply-To: <00122409193304.00852@localhost.localdomain>
Message-ID: <LOBBJMHOIFOLFIPFANHJOEGBCFAA.lists@TheWebsons.com>

YAY!!! Thanks, Matt!
BenO

-----Original Message-----
From: matt [mailto:matt at virtualspectator.com]
Sent: Saturday, December 23, 2000 12:19 PM
To: Ben Ocean
Subject: Re: Why Won't This Work?


did you also do a 
make install  
(as root)

On Sun, 24 Dec 2000, you wrote:
> Hi;
> Because I'm building Zope, I'm trying to install Python-1.5.2. I've
> successfully run:
> ./configure --with-thread
> mmake
> and everything appears to be in order. But when I type
> python
> at a command prompt, I get a bash error. What have I not done?
> TIA,
> BenO
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list
-- 




From canimal at my-deja.com  Fri Dec 15 15:30:21 2000
From: canimal at my-deja.com (canimal at my-deja.com)
Date: Fri, 15 Dec 2000 20:30:21 GMT
Subject: stewpid newbie tricks
References: <3A39D4F9.648D8DBA@mediaone.net>
Message-ID: <91dv0t$h3a$1@nnrp1.deja.com>

In article <3A39D4F9.648D8DBA at mediaone.net>,
  Phil Shinn <philshinn at mediaone.net> wrote:

<snip leadup>

> My problem is when I try to schedule this to run on a daily basis by
> using the NT command line AT:
>
> at 16:00 /Every:m,t,w,th,f,s,su "d:program
> files\python\mine\spider\runSpider.bat"
>
> nothing seems to happen, i.e., scheduler does not seem to kick in, no
> file gets created, not sure what happened.
>
> Clues for the clueless?

Maybe Satheesh has the whole answer here, not sure.  But I think batch
files need a distinct instance of the cmd interpreter.  Don't remember
why, but it's somewhere in the deja archives for the msdos.batch.nt
newsgroup if you want to know more.  Try:

at 16:00 /Every:m,t,w,th,f,s,su "d:\program
files\python\mine\spider\cmd.exe /c runSpider.bat"

(the /c terminates the instance after the batch file runs.)

Now here is a newbie question.  Can't you just call your  .py program
directly?  I thought the .py suffix was associated with the python
interpreter.

Matt


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


From tim.one at home.com  Sun Dec 24 14:02:30 2000
From: tim.one at home.com (Tim Peters)
Date: Sun, 24 Dec 2000 14:02:30 -0500
Subject: One Python 2.1 idea
In-Reply-To: <m38zp6nk8k.fsf@localhost.localdomain>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEGEIFAA.tim.one@home.com>

[Lieven Marchand]
> I think Common Lisp shows that getting very good speed in Python is
> quite feasable. The original CMUCL implementers weren't such a large
> group. Their highly optimising compiler (which incidentally is also
> called Python) has on occasion beaten FORTRAN at numerics. I don't
> know why some people in the Python community think compiling Python is
> such a problem.

The reasons vary, in large part depending on whether or not they've ever
worked on a compiler <wink>.  Whatever, the fact is that there isn't a
Python compiler, nor a credible plan for such a beast to come into
existence.  If the CMUCL compiler group contained at least one person, it
was infinitely larger than the group of people working on a Python compiler
(Vyper notwithstanding).

> Practically all the problems have been tackled and
> solved 20 years ago in the Lisp community.

Excellent!  Then we would greatly appreciate it if you took one of their
compilers, plugged in a Python parser, and released the result.  Sounds like
it shouldn't take you more than a week or two, given that all the problems
were solved so long ago.

reality-check-ly y'rs  - tim





From aleaxit at yahoo.com  Thu Dec 21 17:09:43 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 21 Dec 2000 23:09:43 +0100
Subject: 2 question from a python newbie!
References: <91tl43$rko$1@neptunium.btinternet.com> <slrn944q0t.919.scarblac@pino.selwerd.nl> <91tq58$cq6$1@uranium.btinternet.com>
Message-ID: <91tvmk07qs@news2.newsguy.com>

"G. Willoughby" <sab at NOSPAM.freeuk.com> wrote in message
news:91tq58$cq6$1 at uranium.btinternet.com...
    [snip]
> P.S. so python is really used as a scripting language and nothing more?

Oh, no!  What gives you this idea?  Zope, for example, is a major
framework for developing and deploying web applications; it is
developed in Python.  It has been called "Python's killer app", in
fact (although not everybody would agree with giving it such an
exhalted status, it's a defensible assessment); Python's "core team"
now works for the firm which created and develops Zope.  In the
ActiveState release (or by adding the win32all extensions to the
standard release), Python can be used to write NT 'services', COM
servers, etc.

"A scripting language and nothing more" suggests a language
that only lives for limited customization/automation purposes "on
board" of a 'real' application; although this may be an outdated
view of higher-level (aka 'scripting':-) languages, the 'nothing
more' strongly suggests that this is the speaker's view.

The idea that a 'real application' is packaged as 'a stand-alone
EXE file' is rather dated.  Most "real apps" today come as Windows-
Installer packages (or other kinds of installer-generated self-
unpacking archives), and their installation involves placing several
files of many kinds in various strategic places, updating the Win
registry, etc.  Python applications are just as suitable for this kind
of deployment as those written in any other language; Gordon
McMillan's 'Installer' is a good start at getting such an application
into self-unpacking installer-file form, for example.


Alex





From ian at cfmu.eurocontrol.be  Thu Dec 14 08:45:06 2000
From: ian at cfmu.eurocontrol.be (Ian Wild)
Date: Thu, 14 Dec 2000 13:45:06 GMT
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <tcmay-FCA22F.14073713122000@news.got.net> <918v7a02scn@news1.newsguy.com> <tcmay-04E80C.16181313122000@news.got.net> <91a1ed02a4j@news2.newsguy.com> <3A389B1A.4F33FFD4@cfmu.eurocontrol.be> <91ae8s02llf@news2.newsguy.com>
Message-ID: <3A38CF3E.8BBD2216@cfmu.eurocontrol.be>

Alex Martelli wrote:
> 
> Forget C++ -- we're talking OO _analysis_.  Assume you're
> trying to perform OOA on the universe of discourse "immutable
> planar 2D geometric polygons", for the sake of definiteness.
> 
> If Rectangle and Square are both concrete (instantiable)
> classes, with *overlapping* requirements for membership,
> how do you classify a given rectangular object which
> happens to have equal sides?  Toss a coin?

Both.

Unless you're using classes purely as an implementation
shortcut, why need an object be a member of precisely *one*
class?  What if your analysis is later implemented in
a language without classes (eg Self?), or one where
classes exist only by convention (Perl)?


From cjw at connection.com  Mon Dec  4 09:50:44 2000
From: cjw at connection.com (Colin J. Williams)
Date: Mon, 04 Dec 2000 09:50:44 -0500
Subject: PythonWin
Message-ID: <3A2BAF44.178AC60@connection.com>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20001204/cc33d248/attachment.html>

From tismer at tismer.com  Fri Dec 15 05:48:45 2000
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 15 Dec 2000 12:48:45 +0200
Subject: Is Stackless and uthread.py SMP friendly?
References: <91br9v$qa4$1@nnrp1.deja.com>
Message-ID: <3A39F70D.AA5B1DB8@tismer.com>


jbalding at hotmail.com wrote:
> 
> I read with interest a recent thread asking about threading, the global
> interpreter lock, and symmetric multiprocessing. Nobody mentioned, as
> far as I could see, Stackless Python and microthreads. Having just
> picked up a dual PIII machine, will Stackless and uthread.py make my SMP
> experience with Python a happy one?

Stackless will work just fine, but there is no support for
the second processor. If Python should support having
a real thread for every real processor in some future,
you can of course run microthreads on top of each.
Probably not what you are looking for :-)

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From ben.hutchings at roundpoint.com  Mon Dec 18 09:08:34 2000
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 18 Dec 2000 14:08:34 +0000
Subject: Changing IDLE's stdout color at runtime
References: <91isi7$hup$1@news.hccnet.nl>
Message-ID: <u8zpdzusd.fsf@roundpoint.com>

anton at vredegoor.doge.nl (Anton Vredegoor) writes:

> Hello,
> 
> I am looking for something like:
> 
> def SetIdleStdoutColor(color):
> 	pass  # this function needs defining
> 
> print 'hello world'
> SetIdleStdoutColor('green') 
> print 'hello again'
> SetIdleStdoutColor('defaultcolor')
> 
> Has anybody got an idea how to do this?

You'd need to work out whether standard output is connected to a
terminal or terminal emulation; if so, what kind of terminal that is;
whether that terminal supports colour; and if so, what the appropriate
control codes are for changing the colour.  You *could* assume that
any terminal supports the ANSI escape codes for colour (ESC [30n, ESC
[31n, etc, if I remember correctly), but that would really annoy
anyone using a terminal that doesn't.

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From jkndeja at my-deja.com  Fri Dec 15 04:20:38 2000
From: jkndeja at my-deja.com (jkndeja at my-deja.com)
Date: Fri, 15 Dec 2000 09:20:38 GMT
Subject: Solution: Direct access to Printer I/O lines
References: <917k9b$9i9$1@nnrp1.deja.com> <slrn93hrjn.2vp.ssthapa@localhost.localdomain>
Message-ID: <91cnp3$g59$1@nnrp1.deja.com>

OK, fired by my own enthusiasm I have started work on a 'directlpt'
class extension for Python under Linux, to allow one to get easy access
to the actual data lines.

I have the following thoughts about the interface, and would be
interested in any comments:

Since (in the general case) I want to restrict port access to only the
LPT ports, and since I will need to know when to use ioperms to
allow/disallow port access, I suggest:

- The module implements a class, say directlpt
- The module only allows three 'Singleton' instances of the class,
corresponding to LPT1--3. This is set up in the constructor:

import directlpt

l1 = directlpt.directlpt(1) # gives access to LPT1:

anotherl1 = directlpt.directlpt(1) # will raise an exception - only one
instance allowed

l2 = directlpt.directlpt(2) # gives access to LPT2:

l_bad = directlpt.directlpt(15) # raises an exception - out of range

and then you can write simething like

l2.inb(directlpt.STATUS) # to read the status port.


Comments?

    Thanks
    Jon N



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


From DanielK at aracnet.com  Mon Dec  4 10:35:09 2000
From: DanielK at aracnet.com (Daniel Klein)
Date: Mon, 4 Dec 2000 07:35:09 -0800
Subject: cursors.echo()
Message-ID: <kLOW5.73$%c4.27797@typhoon.aracnet.com>

Hi,

I'm trying to envoke the echo() function but the cursors module does not
seem to be available on Windows. Is there an equivalent Windows function to
turn echo off/on?

Thanks,
Daniel Klein





From gerson.kurz at t-online.de  Wed Dec 27 13:14:46 2000
From: gerson.kurz at t-online.de (Gerson Kurz)
Date: Wed, 27 Dec 2000 18:14:46 GMT
Subject: Reading Unicode Files
Message-ID: <3a4a3186.17549945@news.t-online.de>

The Windows 2000 registry editor creates .REG files in UNICODE by
default. Is there an easy way to use something like readlines() for
it, that is, a list of u-strings ?


From junaftnoon at nospamplzyahoo.com  Tue Dec 26 15:24:45 2000
From: junaftnoon at nospamplzyahoo.com (June Kim)
Date: Wed, 27 Dec 2000 05:24:45 +0900
Subject: sniffing btw my web browser and a server
Message-ID: <92aush$rhc$1@news.nuri.net>

I'm doing some sort of web client programming, and have to observe the
messages that my browser and web-servers transmit, being tired of guessing
what the messages'd be and looking up the references and the specifications.

Anyone has done it in Python or knows how it could be done?


Best regards,

June

ps. using TCP/IP packet sniffers made for mere hacking seems too primitive
or low-level in this case.



From donn at u.washington.edu  Thu Dec 28 14:38:59 2000
From: donn at u.washington.edu (Donn Cave)
Date: 28 Dec 2000 19:38:59 GMT
Subject: Some Python 2.1 ideas
References: <3A452881.7686E717@uswest.net> <%vb16.187233$U46.5801214@news1.sttls1.wa.home.com> <3A464D5D.B541C04E@uswest.net> <mailman.977688488.25438.python-list@python.org> <20001225150250.6FA06A84F@darjeeling.zadka.site.co.il> <mailman.977760789.10224.python-list@python.org> <roy-EA8C3B.11361325122000@news.panix.com> <3A479C29.2631F27B@uswest.net> <mailman.977876470.23568.python-list@python.org> <3A4A50E6.11BBE0FD@uswest.net> <mailman.977962871.16241.python-list@python.org> <3A4AB05C.D41F4022@uswest.net> <92ftif$17fc$1@nntp6.u.washington.edu> <3A4B8865.86712FA1@uswest.net>
Message-ID: <92g4sj$c62i$1@nntp6.u.washington.edu>

Quoth Bjorn Pettersen <pbjorn at uswest.net>:
| Donn Cave wrote:
|> Quoth Bjorn Pettersen <pbjorn at uswest.net>:
|> D-Man wrote:
|>|> Yeah, it is a problem if you don't know where the file came from.  How
|>|> about creating a standard text file format?  All files could be opened
|>|> in binary mode on all platforms, then EOL is handled by each
|>|> application (or better yet make a standard module that does the
|>|> implementation correctly once).
|>|
|>| Feeling like fighting windmills today, eh <wink>? I think the current wish is to
|>| do it from the other end, i.e. instead of relying on the text file format to be
|>| "correct", implement a method on string object that is smart enough to handle all
|>| text file format line endings intelligently.
|>
|> Without the context of the I/O stream, to decide whether '\r' is
|> data or separator, isn't that string function going to be the opposite
|> of smart?  How does a string handle '\r' intelligently?
|
| By assuming that if it's the last character in the string, it came from a mac file and
| is the line terminator. Similarly for '\r\n' (dos) and '\n' (unix) when at the end of
| the string.
|
| It might be cute to have a MacTextFile class etc. that ensures that the line endings
| are consistent, but personally I think that is overkill (and is likely to be very
| slow).

I guess a function like that would often be the right thing.  But
it doesn't strike me as what I would have called intelligent handling,
just aggressive in its assumptions.  In a situation where the application
calls for an aggressive normalization of the data, it's obviously the
right thing.  Otherwise, it's useless, where a per I/O stream solution
might still be useful, and I guess cute too.

	Donn Cave, donn at u.washington.edu


From nas at arctrix.com  Wed Dec 13 13:43:16 2000
From: nas at arctrix.com (Neil Schemenauer)
Date: Wed, 13 Dec 2000 10:43:16 -0800
Subject: xreadlines (was Re: while true: !!!)
In-Reply-To: <slrn93g37o.9tj.jepler@potty.housenet>; from jepler@inetnebr.com on Wed, Dec 13, 2000 at 11:57:45PM +0000
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <slrn93g37o.9tj.jepler@potty.housenet>
Message-ID: <20001213104315.A17726@glacier.fnational.com>

On Wed, Dec 13, 2000 at 11:57:45PM +0000, Jeff Epler wrote:
> In Python, there is an existing precedent for creating a
> special 'iteratable' object where the entire object would not
> fit in memory, the 'xrange' function and its associated object
> type.
> 
> I take this cue to suggest the idea of an 'xreadlines' object.

Its been done before (by me and by other people).  The best
"solution" I came up with was "repeatmodule":

    http://arctrix.com/nas/python/repeatmodule-1.0.tar.gz

Here is an example of its use:

    import sys
    from repeat import repeat
    for line in repeat(sys.stdin.readline):
            ...

and a little documentation on the function:

    repeat.repeat(<function>, [<until>]) -> RepeatObject

    RepeatObjects are sequences that only support the method
    __getitem__.  <function> is called on __getitem__.  If
    <until> is not specified a false value from <function> raises
    an IndexError.  If <until> is specified, if the value is
    equal to <until> then IndexError is raised.

I never did end up using it.  Usually speed doesn't matter that
much.  If would be interesting to see how quickly Python can chew
up a file with my repeat hack and AMK's recent glibc getline
optimization.  Another day perhaps.

  Neil



From shindich at my-deja.com  Fri Dec  8 11:55:50 2000
From: shindich at my-deja.com (shindich at my-deja.com)
Date: Fri, 08 Dec 2000 16:55:50 GMT
Subject: A "C" version of Weak Reference extension
Message-ID: <90r3qi$ole$1@nnrp1.deja.com>

I am happy to announce that Weak Reference Extension project has been
fully converted to "C". Weak Reference extension is
 now completely CPython compatible.
 Check out Beta 2 at Source Forge
 (http://pyweakreference.sourceforge.net/ ;
 http://sourceforge.net/projects/pyweakreference/)

 Enjoy!

 Alex Shindich
 Visit http://www.shindich.com/


Sent via Deja.com http://www.deja.com/
Before you buy.


From dsh8290 at rit.edu  Tue Dec 26 17:48:21 2000
From: dsh8290 at rit.edu (D-Man)
Date: Tue, 26 Dec 2000 17:48:21 -0500
Subject: A Suggestion for Python Dictionary/Class
In-Reply-To: <920fe9$c20$1@hecate.umd.edu>; from billtj@y.glue.umd.edu on Fri, Dec 22, 2000 at 09:00:57PM +0000
References: <920fe9$c20$1@hecate.umd.edu>
Message-ID: <20001226174821.A8309@westsidecnc.rh.rit.edu>

I suggested a similar thing a while back, but many people on the list
didn't like it.  I thought it would improve readability, help catch
these errors, and be consistent to introduce a "local" keyword and
require local variables to be declared.  It wouldn't lose any of the
dynamicity of python by declaring the type of the variable as in
C/C++/Java.

The purpose of my 'local' keyword would be to restrict the namespace
to the names specified by 'local' (and global variables which already
must be declared with "global").

-D




From jh at web.de  Wed Dec  6 15:23:32 2000
From: jh at web.de (Jürgen Hermann)
Date: Wed, 6 Dec 2000 21:23:32 +0100
Subject: [ANN] MoinMoin Release 0.7 - SECURITY UPDATE
Message-ID: <90m77s$2e9$05$1@news.t-online.com>

A WikiWikiWeb is a collaborative hypertext environment, with an emphasis on
easy access to and modification of information. MoinMoin is a Python
WikiClone that allows you to easily set up your own wiki, only requiring a
Web server and a Python installation.

This is a security update, which explains the short release cycle. It
replaces some exec() calls by __import__(), which is much safer (or
actually,
safe in contrast to totally unsafe).

IF YOU HAVE INSTALLED RELEASE 0.5 OR 0.6, UPGRADE NOW!

Thanks to Richard Jones (richard at bizarsoftware.com.au) for spotting this.


Homepage:
    http://moin.sourceforge.net/

Download:
    http://download.sourceforge.net/moin/

Mailing lists:
    http://lists.dragon-ware.com/mailman/listinfo/moin-users
    http://lists.dragon-ware.com/mailman/listinfo/moin-dev


New features:
    * RecentChanges bookmarking

Bugfixes:
    * SECURITY FIX
    * Non-greedy extended WikiNames





From deskcode at my-deja.com  Sat Dec 16 20:20:13 2000
From: deskcode at my-deja.com (deskcode at my-deja.com)
Date: Sun, 17 Dec 2000 01:20:13 GMT
Subject: www,deskcode.com hot new coding resource python
Message-ID: <91h4ca$qat$1@nnrp1.deja.com>

www.DeskCode.com is a user-based web community for programmers and
developers.  Offers resources to 15 different programming languages.

Please come and visit us, share your knowledge by submitting articles,
tutorials, code snippets, or whole programs.  You can upload on to our
server or link to a homepage of your choice.

We really need the boost in our python resources. Deskcode.com was just
luanched on the 13th of Decemeber. Come and give us a view.

Thank You
www.deskcode.com


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


From schaffer at optonline.net  Tue Dec 19 00:28:35 2000
From: schaffer at optonline.net (Les Schaffer)
Date: 19 Dec 2000 00:28:35 -0500
Subject: PyOpenGL and Nvidia GL?
References: <tmi%5.60958$df5.2143169@news1.crdva1.bc.home.com>
Message-ID: <87k88xkmik.fsf@optonline.net>

Jeff said:
> Is this Nvidia's fault for just not supplying that function in their
> GL layer?

well, i've got good news and i've got bad news.

the good news is: 

glPolygonOffsetEXT is defined in nvidia's libglx.so.XXX in the
extensions directory:

   /usr/X11R6/lib/modules/extensions/libglx.so.1.0.5

and if you add this library to the gl_libs ("glx") in config.py in the
PyOpenGL top directory (you need to make sure the library is visible
when linking too), then you no longer get the undefined symbol error
for glPolygonOffsetEXT.

Now the bad news.

When you link with libglx, you get a whole slew of new undefined
symbols. some of them i believe are very old symbols from early
versions of X11.

i dont know how the geniuses at NVidia managed to do this, but if i
showed you the dumb hack that i made to _openglmodule.c, for example:

void WriteToClient(void){
       return;
}
void xf86memcpy(void){
       return;
}

...

to see how deep the undefined symbols are, thats just some of the
2-handfuls i found so far of undefined symbols, you will see why i
will never buy another video card from NVidia so long as their drivers
are closed source.

les 'live-and-learn' schaffer


From cjensen at bioeng.ucsd.edu  Thu Dec  7 14:34:08 2000
From: cjensen at bioeng.ucsd.edu (Curtis Jensen)
Date: Thu, 07 Dec 2000 11:34:08 -0800
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu> <slrn92tu7t.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com>
Message-ID: <3A2FE630.5C113AB@bioeng.ucsd.edu>

"A.M. Kuchling" wrote:
> 
> On Wed, 06 Dec 2000 17:06:55 -0800, Nathan Gray <nospam at caltech.edu> wrote:
> >It's really upsetting how often I've seen this sort of response to a polite
> >question on this newsgroup lately.  Tell me, what's so terrible with asking a
> >group of experts for pointers when researching a subject?  It's a perfectly
> >legitimate and highly effective way to learn about something.  There's plenty
> 
> I've noticed this trend, too, though T.C. May, the person you're
> replying to, has always had a low tolerance for cluelessness, as I
> noted when lurking on the cypherpunks mailing list in the early 90s;
> May's posts were often informative and entertaining, but not often
> placid.  (I was pleased to turn up here, though I'm not sure why he's
> taken up Python.)

A while back, someone wrote in and asked what a text editor was.  The
answers to this were polite and informative, and non-demeaning.  I'm
curious why this one and why others get ripped on, and others don't. 
Luck of the draw?

-- 
Curtis Jensen
cjensen at bioeng.ucsd.edu
http://www-bioeng.ucsd.edu/~cjensen/
FAX (425) 740-1451


From tim.one at home.com  Tue Dec 26 13:16:49 2000
From: tim.one at home.com (Tim Peters)
Date: Tue, 26 Dec 2000 13:16:49 -0500
Subject: One Python 2.1 idea
In-Reply-To: <20001226140032.C25820@xs4all.nl>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEIFIFAA.tim.one@home.com>

[Thomas Wouters]
> ...
> Actually, it's even worse :) Methods are never stored in an
> instance's dict, but are created 'on the fly' from the class's
> unbound method. The binding of the method isn't too expensive,
> but it isn't totally free either (if anything, it's a dict 'miss'
> and an aditional dict lookup in the class's dict. More than one
> if you use deep or multiple inheritance (or both.))
>
> You *can* speed it up, by doing 'x.foo = x.foo' somewhere early
> in your call chain :) ...

Note:

>>> class Oops:
	def foo(self):
		pass

>>> x = Oops()
>>> x.foo = x.foo
>>> x.foo.im_self is x
1
>>>

Oops!  That is, x is now involved in a reference cycle (x's dict contains a
bound method that points back to x).  Before 2.0, that's a memory leak.  In
2.0, it's a leak only if the class (unlike Oops above) has a __del__ method
(or you've disabled cyclic gc).

gotta-love-the-way-"x.foo=x.foo"-looks-though<wink>-ly y'rs  - tim




From tdelaney at avaya.com  Mon Dec 18 18:35:43 2000
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Tue, 19 Dec 2000 10:35:43 +1100
Subject: while true: !!!
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B0FBCC5@natasha.auslabs.avaya.com>

> >The idiom is actually
> >
> >while 1:
> >
> >	do something
> >
> >	if condition:
> >		break
> >
> >	do something else

> This is exactly what a python program should be NOT doing. "while 1" 
> is one of the most cryptic statements in Computer Science. other 
> languages are known for this stuff, but I do not know why python 
> doesn't auto define true and false, so that we don't have to write 
> unreadable stuff.

Personally, I agree with you - I was merely explaining *what* the idiom was
to someone who doesn't know.

I go well out of my way to avoid this idiom, as I feel you should always
have the termination condition at one end of the loop, preferably the top
(except in *Exception*al circumstances). For this reason, I would usually
use a Stream heirarchy of classes for handling files.

Tim Delaney
Avaya Australia



From alexander_prigozhin at my-deja.com  Thu Dec 14 14:22:24 2000
From: alexander_prigozhin at my-deja.com (alexander_prigozhin at my-deja.com)
Date: Thu, 14 Dec 2000 19:22:24 GMT
Subject: .NET in One Day Seminar - 2001 Schedule
Message-ID: <91b6l9$952$1@nnrp1.deja.com>

.NET IN ONE DAY: The Multi-Language Platform for the Age of the Internet
A one-day course by Bertrand Meyer

NEW LOCATIONS
=============

US:
---

Santa Barbara, CA: January 29
Pittsburgh, PA: February 5
Raleigh, NC: February 6
Detroit, MI: March 26
New York, NY: March 27
Boston, MA: March 29
Washington, DC: March 30
San Francisco, CA: April 6

CANADA:
-------

Montreal, Canada: April 2
Toronto, Canada: April 3
Vancouver, Canada: April 5

PACIFIC RIM:
------------

Singapore, Singapore: February 19
Kuala Lumpur, Malaysia: February 20
Bangkok, Thailand: February 22

EUROPE:
-------

Milan, Italy: March 2
Amsterdam, The Netherlands: March 5
Copenhagen, Denmark: March 6
Frankfurt, Germany: March 8
London, United Kingdom: March 9


**If your company is interested in an in-house presentation of
.NET in One Day please contact training at dotnetexperts.com**

INTRODUCTION
============

In July of 2000 Microsoft introduced the .NET framework, the most
important development since the introduction of Windows in 1991. The
result of a $2 billion investment, .NET is a revolutionary multi-
language platform integrating all aspects of application development
and closely integrating the Web at every step.

This information-packed one-day course covers the essentials of .NET,
including both the "big picture" and a review of all major aspects of
the technology. Presented by Dr. Bertrand Meyer, one of the pioneers of
modern software technology, it is based on more than one year's advance
exposure to .NET prior to the technology's official release. Dr.
Meyer's team at Interactive Software Engineering and Monash University
worked with Microsoft to integrate ISE's technology with .NET,
culminating in a joint appearance with Bill Gates at the Microsoft
Professional Development Conference that first introduced .NET.

COURSE OUTLINE
===============

1. Overview
-----------

* .NET in 15 minutes: an overview of the technology and its
contributions.

2. The challenges: if .NET is the answer, what are the questions?
-----------------------------------------------------------------

* Background: The Internet and the evolution of software development.
  E-commerce and its demands. Requirements of mission-critical
applications. Exposing the business model.
* Object Technology: contribution and challenges. O-O languages, GUI
tools, databases.
* Component-based development: COM, CORBA, Enterprise Java Beans. The
  notion of Interface Description Language. Contributions and
limitations of pre-.NET component approaches.
* Programming for the Web and e-commerce: CGI scripts, Active Server
Pages, Java Server pages. Advantages and drawbacks.
* The state of multi-language interoperability. Approaches to portable
  application development. Graphics, database issues.

3 .NET: The Vision and the Platform
-----------------------------------

* The .NET architecture: runtime, framework, platform, web services.
* The .NET runtime: architecture and goals. Comparison with the Java
  Virtual Machine.
* MSIL: the intermediate language. Security issues and the concept of
  verifiability. How critical is it to produce verifiable code?
* Organizing and extending your components: assemblies and metadata. An
  application: equipping components with contracts.

4. The .NET object model and type system
----------------------------------------

* Classes, methods, fields, properties and events
* .NET types: reference and value types, array types, arrays
* Inheritance concepts: multiple interface inheritance, novariance
* Encapsulating behavior: delegates
* C#: a language for programming .NET
* C# versus Java
* .NET mechanisms and the dominant languages: commonalities
  and mismatches. How easy is it to map an existing language
  into the .NET model?

5. Language interoperability
----------------------------

* Available languages and degree of interoperability. Cross-language
  inheritance; cross-language debugging
* The Common Language System: both a consumer and an extender be
  Levels of compliance
* Advantages and challenges of CLS compliance
* Examples: combining components from various languages
* From a common runtime to a common development environment:
  Visual Studio.NET and the concept of multi-language, pluggable
  environment. GUI, browsing, debugging

6. Frameworks and applications
------------------------------

* Web and Win Forms
* Remoting and threading capabilities
* ASP+: Active Server Pages +. Building advanced Web sites for e-
commerce
* Web services, SOAP and Building Block Services
* Database access and manipulation: ADO+

7. Summary and perspective
--------------------------

* .NET and the competition
* The significance of .NET
* Future developments
* Corporate strategies: getting ready for .NET


COURSE MATERIAL
================

The material distributed to participants includes more than 150 slides,
as well as supporting articles.

WHO SHOULD ATTEND
=================

This course presents a compact, in-depth survey of the .NET
technology over one day. It is intended for both managers
(VPs of technology, CTOs, project leaders, Web content
managers, e-commerce strategy leaders) and for software
developers who want to know about the most important technology
offering in many years. The .NET technology will affect everyone;
no one can afford to miss it.


ABOUT THE SPEAKER
==================

Dr. Bertrand Meyer has played a major role in developing modern software
technology and bringing it to software engineers working in production
environments. Bertrand is best known for his best-seller "Object-
Oriented Software Construction" (2nd edition, Prentice Hall, 1997),
recipient of the Jolt Award and one of the seminal works on modern
software technology. He is the author of 8 other books on software
technology, programming languages and object-oriented development,
including "Reusable Software", "Eiffel: The Language" and "Object
Success", and of numerous widely cited articles. He is a columnist for
Software Development magazine, the Journal of Object-Oriented
Programming and was Department Editor for IEEE Computer. He is the
series chair for the international TOOLS conferences (USA, Europe,
Pacific), devoted to component and object technology.
He is also the editor of the Prentice Hall Object and Component
Technology Series.
As head of ISE's development team, he has led the design of tools and
environments used routinely by major corporations worldwide for their
mission-critical applications.

FEE
===

USA: USD 695.00 or USD 645.00*
Singapore: 995 SGD or 895 SGD*
Malaysia: 1995 MYR or 1795 MYR*
Italy: 1,624,445 ITL or 1,507,695 ITL*
The Netherlands: 1795 NLG or 1695 NLG*
Denmark: 6,195 DKK or 5,775 DKK*
Germany: 1595 DEM or 1495 DEM*
United Kingdom: 495 GBP or 455 GBP*
Canada: 895 CAD or 795 CAD*

* if paid three weeks prior to the session.

A 10% discount will be applied to registrations
for 3 or more people from the same company registering at the same time.

A 20% discount will be applied to registrations
for 5 or more people from the same company registering at the same time.

REGISTRATION FORM
==================
The registration is available on-line at http:www.dotnetexperts.com,
by phone at 805-685-1006, or using the form below by fax at 805-685-
6869
or by email at training at dotnet.experts.com:

I am registering for:

___Santa Barbara, CA: January 29
___Pittsburgh, PA: February  5
___Raleigh, NC: February 6
___Singapore, Singapore: February 19
___Kuala Lumpur, Malaysia: February 20
___Bangkok, Thailand: February 22
___Milan, Italy: March 2
___Amsterdam, The Netherlands: March 5
___Copenhagen, Denmark: March 6
___Frankfurt, Germany: March 8
___London, United Kingdom: March 9
___Detroit, MI: March 26
___New York, NY: March 27
___Boston, MA - March 29
___Washington, DC: March 30
___Montreal, Canada: April 2
___Toronto, Canada: April 3
___Vancouver, Canada: April 5
___San Francisco, CA: April 6


***ALL PAYMENTS (Credit Card, International Money Order and
Bank Transfer) MUST BE MADE IN US DOLLARS.***

Address:
Company__________________________________________________

Department or Division __________________________________

Street Address___________________________________________

City _____________________Zip Code ______________________

State _____ _____________________________________________

Country ____________________________

Telephone ______________________ Fax ____________________

Email ___________________________________________________

Names of Participants:

______________________________________________

______________________________________________

______________________________________________

______________________________________________

______________________________________________

Method of Payment (check one):
_____Check enclosed (amount) ___________________________
(make check payable to Interactive Software Engineering, Inc.)
_____Credit Card
Number____________________________Exp. Date__________

Authorized Signature____________________________________

Substitutions will be accepted at any time. Any cancellation
received three weeks before the seminar will be liable to a
50% service fee. Cancellations received after this date will
be liable for the entire fee.


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


From tim.one at home.com  Sat Dec 23 02:49:55 2000
From: tim.one at home.com (Tim Peters)
Date: Sat, 23 Dec 2000 02:49:55 -0500
Subject: Python's scope rules
In-Reply-To: <91srpu0262q@news2.newsguy.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEDLIFAA.tim.one@home.com>

[Alex Martelli]
> ...
> I think you can say 'will' rather than 'would', as I see no
> reason why PEP 227 should not make it into Python 2.1.

Me neither, but that doesn't matter.  What's important is that Guido doesn't
have strong objections.

> I do hope that Python keeps the simplicity of avoiding the
> distinction between 'let' and 'set!' (creation of a new
> binding in this scope, vs modification of existing binding
> in some nested scope), variable-declarations, or other
> complexities to allow fancy stuff such as rebinding of
> lexically-nested-scoped variables.

Guido's (*possibly* wishful) hope is that 90% of scope complaints are due to
"the lambda problem", i.e. the need to abuse the default argument mechanism
to "import" local names into a nested lambda's scope.  Since lambda syntax
allows only an expression (and not a stmt), and all binding constructs are
stmts, there's no possibility of rebinding within the body of a lambda
anyway.

So, the hope is that 90% of all scope complaints will go away, and then the
remaining 10% are 10x easier to ignore <wink>.  The reluctant possibility of
extending this someday remains open, though.

> I know of no language that allows that without constraints --
> say that your scope is currently nested 5 levels deep and
> there's a binding for 'x' in each nested scope, how do you
> specify 'the x that's bound exactly 3 levels up from this one'?

def f():
    i = 1
    def g():
        i = 2
        def h():
            global f.i
            # i now refers to the i local to f

is one backward-compatible syntax suggested in the past.  There were others.
I didn't much like any of them.  But that doesn't matter <wink>:  Guido
didn't like them either.

> Java specifically disallows homonym variables in nested
> scopes, and maybe that should be considered as a part of
> PEP 227 -- such homonimy frequently causes confusion in
> languages which allow it, I think, and apparently Java's
> designers agree.

Guido will never go for any scheme that prohibits shadowing of global or
builtin names.  Given that, it would be a wart to make local names "more
special" than those.  Java, of course, doesn't have global or builtin names
to contend with, so doesn't have this problem (although Java has a related
problem:  it allows locals to shadow field names, else you couldn't change
the fields in base classes without possibly needing to rewrite bodies of
subclasses too).

Python 2.1 should have a new warning mechanism (PEP 230), and I hope we can
warn about shadowing at compile time.

cattle-prod-vs-electric-chair-is-a-hard-choice-ly y'rs  - tim




From sbrunning at bigfoot.com  Tue Dec 12 04:27:49 2000
From: sbrunning at bigfoot.com (Simon B.)
Date: Tue, 12 Dec 2000 09:27:49 GMT
Subject: Python2 distutils question: how to best "autoconfig"...?
References: <9130e60obc@news2.newsguy.com> <82aZ5.3458$Kd1.395204@newsb.telia.net> <913br4012r8@news2.newsguy.com> <LbbZ5.3172$O%.511511@newsc.telia.net>
Message-ID: <914r2g$126$1@nnrp1.deja.com>

In article <LbbZ5.3172$O%.511511 at newsc.telia.net>,
  "Fredrik Lundh" <fredrik at effbot.org> wrote:
> interesting.  I thought Tim was on a vacation, but his
> posts show up in the python.org archives.  maybe my
> ISP has installed a timbot filter?

Probably he posted via email - the mailing list hasn't moved over to
Digital Creations yet, and the usenet/mailing list bridge still isn't
working properly. Postings to usenet do make it across to the list, but
email postings don't seem to be getting across to the newsgroup.

--
Simon Brunning
When I was a kid, I used to pray every night for a new bicycle. Then I
realized that the Lord, in his wisdom, didn't work that way. So I just
stole one and asked him to forgive me. - Emo Philips


Sent via Deja.com http://www.deja.com/
Before you buy.


From glenfant at equod.com.nospam  Wed Dec 13 11:14:21 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Wed, 13 Dec 2000 17:14:21 +0100
Subject: Source Code Management
References: <fb4f3tsgffnmvme64aol1cn397te7ln349@4ax.com>
Message-ID: <918711$meh$1@reader1.imaginet.fr>

Have a look to cvs at http://www.gnu.org/software/cvs/cvs.html
http://www.wincvs.org
http://www.cvshome.org/

The best of the open source version control tool.
There are python wrappers to CVS somewhere, I don't remember where.

"Dale Strickland-Clark" <dale at out-think.NOSPAMco.uk> a ?crit dans le message
news: fb4f3tsgffnmvme64aol1cn397te7ln349 at 4ax.com...
> I'm looking for a source code management system for looking after Python,
Java, Javascript and
> possibly other languages too.
>
> Does anyone know of such a system that matches or comes close?
>
> Essential features:
> * Runs on Windows 2000
> * Works over the Internet (ftp would be good)
> * Secure (ftp level security would pobably suffice)
> * Module level version control with deltas or compressed archives
> * Package level (multiple modules) sign in/sign out
> * System level release control. (ie, this is now release x, note all
module versions for this
> release.)
>
> Desirable features:
> * Extendible - preferably scriptable with Python
> * Manage arbitrary files including documentation
> * GUI or command line interfaces
> * Low maintenance
> * Library stored as plain files and not in some odd database format
>
> That's not too much to ask, is it!  :-)
>
> Thanks
> --
> Dale Strickland-Clark
> Out-Think Ltd
> Business Technology Consultants



From mskott at image.dk  Wed Dec 27 19:56:33 2000
From: mskott at image.dk (Martin Skøtt)
Date: 28 Dec 2000 01:56:33 +0100
Subject: removing directories
References: <Pine.A41.3.96.1001227175036.37924A-100000@aix2.uottawa.ca>
Message-ID: <wvclgy7y.fsf@jumbo.skott.dk>

<abouf066 at aix2.uottawa.ca> writes:

> Greetings,
> 
> 	Is there a way to force removal of directories even if they
> contain files and/or directories. I just tried 
> 
> os.rmdir( path ) and removedirs ( path)
> 
> both functions complain about existence of files !?


I just tested shutil.rmtree on my Linux box with Python 1.5.1 and it
deletes a directory with files and subdirectories without
complaining. I didn't put in any other options than the path.
It is my experience that shutil is a lot easier to use than os, but
also more lethal (easier to make mistakes and do to much).
-- 
Med venlig hilsen
Martin Sk?tt
mskott at image.dk

Believe nothing no matter where you read it or who has said it, not even if I
have said it, unless it agrees with your own reason and your own common sense. 
--Buddha


From grey at despair.rpglink.com  Fri Dec 15 13:33:06 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Fri, 15 Dec 2000 18:33:06 -0000
Subject: comments of python by a perl user
References: <91dlar$7s3$1@nnrp1.deja.com>
Message-ID: <slrn93kov2.2nj.grey@teleute.rpglink.com>

On Fri, 15 Dec 2000 17:45:00 GMT, Ben Park <benpark at my-deja.com> wrote:
>A comparison of Perl and Python by a perl user:
>http://www.perl.com/pub/language/versus/python.html

    More like a level-headed comparison by Tom Christiansen.  

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From support at internetdiscovery.com  Mon Dec 18 13:56:53 2000
From: support at internetdiscovery.com (Mike Clarkson)
Date: Mon, 18 Dec 2000 18:56:53 GMT
Subject: freeze.py -s options under Windows
Message-ID: <3a3d0aa7.13061249@24.0.228.33>

Does anyone have any extra documentation of the -s command to
freeze.py under Windows?

I have sucessfully used freeze.py under Windows98 to freeze a Tkinter
application. I build from sources, and then moved things freeze needed
to a config directory, and put the .pyd and .dll files the exe needs
in the same directory as the exe. Things work fine if I freeze without
the -s option.

I want to use -s windows to get rid of the console on startup under
Windows, but when I build with -s, the application just dies without
giving me any Tkinter root window. Any ideas on how to freeze a GUI
application under Windows?

(It seems curious to me that the Makefiles generated with and without
-s windows are the same.)

Are there any rules about writing to stderr/stdout with -s windows?

Many thanks in advance,

Mike.



From parker at gol.com  Sat Dec  9 20:43:26 2000
From: parker at gol.com (Ian Parker)
Date: Sun, 10 Dec 2000 01:43:26 GMT
Subject: Handling signals and performance issues.
References: <meRU5.211$Mx.12164@sjc-read.news.verio.net>
Message-ID: <m9dgmAAe+tM6EwnY@gol.com>

In article <meRU5.211$Mx.12164 at sjc-read.news.verio.net>, Ray Van Dolson
<rayvd at nospam.firetail.org> writes
>I've written a small Python script that watches a logfile (kinda like tail -
>f) and splits it into certain other files in real time.  I'm running into 
>two issues however that I haven't quite resolved yet to get this script 
>running as I'd like it.  Firstly, the program uses up a lot of CPU time.  
>This is almost certainly because of the way I'm watching the logfiles for 
>changes.  Basically it's like this:
>
>while 1:
>       ln=inputLog.readline()
>       if ln:
>               processLine(ln)
>       else:
>               pass
>
>Much of the time of course there have been no changes to the logfile and so 
>it keeps looping, passing and waiting for a change.  While the system 
>actually remains fast, watching 'top' shows that python is using up all the 
>cpu when it can.  Is there a better way to watch a file for changes that 
>doesn' use such a cpu consuming while loop?
>
>My second problem is that while the program handles a CTRL-C, etc well--
>cleaning up and the like, if it is sent a kill -9 or kill -HUP by root or 
>the user spawning the process it dies a horrible death--ie, it doesn't clean 
>up and flush its buffers, or run the 'cleanup' functions I've defined with a 
>try: except: block.  How can I watch for a kill -9 or kill -HUP and have my 
>script respond accordingly before shutting down?
>
>Thanks much for any help provided.
>Ray Van Dolson

How about putting a "time.sleep(n)" in place of the "pass" statement?
That way you'd pause for a while when there was no new data in the log
file, but continue to loop quickly if multiple lines had been appended
during the previous n seconds.

-- 
Ian Parker


From thaiwey at my-deja.com  Wed Dec  6 17:01:10 2000
From: thaiwey at my-deja.com (thaiwey at my-deja.com)
Date: Wed, 06 Dec 2000 22:01:10 GMT
Subject: SSL socket in 2.0
References: <90lpjm$edp$1@nnrp1.deja.com>
Message-ID: <90mcv1$9c$1@nnrp1.deja.com>

In article <90lpjm$edp$1 at nnrp1.deja.com>,
  thaiwey at my-deja.com wrote:
> hi,
>
> i have been trying to compile the socket module with SSL support on my
> nt without much success. i was using the pcbuild.dsw in MSVC 6.0. can
> anyone advice me as to how i can set the necessary flags in the
project
> options/settings to turn on SSL.
>
> i have already compiled openssl-0.9.6 for my machine. i have also
added
> "USE_SSL" in the preprocessor definitions, paths to openssl in the
> include directories, library path to openssl, and "libeay32.lib
> ssleay32.lib" to the Object\library modules section. when compiling, i
> keep getting the following errors:
>
> Compiling...
> socketmodule.c
> G:\Python-2.0\Modules\socketmodule.c(2112) : error C2099: initializer
> is not a constant
> G:\Python-2.0\Modules\socketmodule.c(2114) : warning C4047:
> 'initializing' : 'int ' differs in levels of indirection from 'char
[4]'
> G:\Python-2.0\Modules\socketmodule.c(2115) : warning C4047:
> 'initializing' : 'char *' differs in levels of indirection from
> 'unsigned int '
> G:\Python-2.0\Modules\socketmodule.c(2118) : warning C4047:
> 'initializing' : 'int ' differs in levels of indirection from 'void
> (__cdecl *)(struct _object *)'
> G:\Python-2.0\Modules\socketmodule.c(2120) : warning C4047:
> 'initializing' : 'int (__cdecl *)(struct _object *,struct _iobuf *,int
> )' differs in levels of indirection from 'struct _object *(__cdecl *)
> (struct _object *,char *)'
> Error executing cl.exe.
>
> thanks.
>
> tw
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
>

actually, i found out that i can change some code in socketmodule.c so
that it compiles. i found the answer in faq
(http://www.python.org/doc/FAQ.html). but after that, when i try to use
it, i keep getting this error:

Traceback (most recent call last):
  File "httplib.py", line 819, in ?
    test()
  File "httplib.py", line 805, in test
    hs.connect(host)
  File "httplib.py", line 660, in connect
    self._conn.connect()
  File "httplib.py", line 616, in connect
    ssl = socket.ssl(sock, self.key_file, self.cert_file)
TypeError: ssl, argument 1: expected socket, instance found

does anyone know why?

tw


Sent via Deja.com http://www.deja.com/
Before you buy.


From thelocust at my-deja.com  Thu Dec  7 01:57:34 2000
From: thelocust at my-deja.com (thelocust at my-deja.com)
Date: Thu, 07 Dec 2000 06:57:34 GMT
Subject: can i make sys.stdin.read() not wait?
Message-ID: <90ncct$p58$1@nnrp1.deja.com>

I'm working on a program that requires me to read in input (usually one
character) if it is input.  For example, i have a while loop that looks
like this:

while 1:
      x = sys.stdin.read(1)
      if x == 'q':
           break
      test.handlestuff()

Well, my problem is is that it seems that read MUST have input to move
on.  Is there any way that it just attempts to get whatever is on the
TTY?  or maybe the last thing that was put to it?


Sent via Deja.com http://www.deja.com/
Before you buy.


From ennisr at my-deja.com  Mon Dec 18 15:36:56 2000
From: ennisr at my-deja.com (ennisr at my-deja.com)
Date: Mon, 18 Dec 2000 20:36:56 GMT
Subject: missing HtmlBuilder
Message-ID: <91lsh8$9p7$1@nnrp1.deja.com>

I am a relative newbie to Python, using Python 2.0 under Windows.
4DOM from Fourthought, Inc. has also been installed.

I have run into a problem trying to run the DOM examples from David
Mertz article "Charming Python: The dynamics of DOM"
(http://gnosis.cx/publish/programming/charming_python_2.txt).

Modules referenced in the following import statements seem to be
missing.

    from xml.dom import core
    from xml.dom.html_builder import HtmlBuilder
    from xml.dom.utils import FileReader

Do I have the wrong software loaded?  Where can I find the missing
modules?

Thanks for the help.





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


From tony at lsl.co.uk  Wed Dec 13 06:43:33 2000
From: tony at lsl.co.uk (Tony J Ibbs (Tibs))
Date: Wed, 13 Dec 2000 11:43:33 -0000
Subject: Thank you, Fredrik. (was Re: Label-Value)
In-Reply-To: <Pine.GSO.4.21.0012121550530.19682-100000@y.glue.umd.edu>
Message-ID: <002a01c064f9$ebd67a70$f05aa8c0@lslp7o.int.lsl.co.uk>

After saying thank-you nicely to Fredrik (always a nice thing to do),
Roy Katz wrote:
> just to clarify something about '@identifier':
> I'm playing devil's advocate. So don't see this
> as an attack.

And there's the problem.

When communicating in text, one does not have visual and auditory cues
as to what is intended. Just the text. ("just the text, ma'am")

Someone who decides to play devil's advocate and *doesn't say so*
"sounds" just the same (in text) as someone who really means what they
are saying. So that person shouldn't be offended if people respond as if
they meant it.

But don't see this as an attack. No, please don't. Really. I'm just a
friendly pedant, who likes the language just the way G*d* indented it.

[Mandatory exception: heh, Alex, I don't think "print >>" is anywhere
near as bad as 012 being 10 and not 12 - now *that's* a wart on the
language!]

Tibs
--
Tony J Ibbs (Tibs)      http://www.tibsnjoan.co.uk/
My views! Mine! Mine! (Unless Laser-Scan ask nicely to borrow them.)
[I've read it twice. I've thought it over. I'm sending it anyway.]




From aleaxit at yahoo.com  Wed Dec  6 06:45:57 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 6 Dec 2000 12:45:57 +0100
Subject: Calculating factorial
References: <3a2a9496.647995@uutiset.nic.fi> <3a2c8c2a_3@corp.newsfeeds.com> <90ib2e0216k@news2.newsguy.com> <vr8X5.51298$15.10971944@news1.rdc1.az.home.com>
Message-ID: <90l8t402a1d@news2.newsguy.com>

"Tim Hochberg" <tim.hochberg at ieee.org> wrote in message
news:vr8X5.51298$15.10971944 at news1.rdc1.az.home.com...
    [snip]
> > Since range(N) takes up memory space growing as O(N), for a
> > large enough N this memory consumption becomes irrelevant
> > compared to the memory which N! itself takes up.
>
> This is not true. If range supported long integers, for _really_ large N,
> range(N) would take up O(sum(log i))=O(log(N!)) memory space. So it would

Good (theoretical) point!  Oh well, at least I have good company in
this particular blind-spot (who was the eminent computer scientist
who claimed O(N) behavior for keysort -- forgetting that for large
enough N, the logN number-of-digits factor comes in, making the
whole O(N logN) again...?-).


Alex





From rturpin at my-deja.com  Sun Dec 31 17:26:20 2000
From: rturpin at my-deja.com (rturpin at my-deja.com)
Date: Sun, 31 Dec 2000 22:26:20 GMT
Subject: Is 100,000 entries big for a dictionary?
References: <mailman.978298994.20463.python-list@python.org>
Message-ID: <92obq8$qf5$1@nnrp1.deja.com>

In article <mailman.978298994.20463.python-list at python.org>,
  "Tim Peters" <tim.one at home.com> wrote:
> Provided you have sufficient RAM, the theoretical performance
> (for lookup, insert and delete) remains O(1) expected case,
> O(N) worst case, regardless of dict size.  This is also what
> I've seen in practice:  Python ensures that at least 1/3 of
> the internal hash table entries are unused (it uses an
> extreme form of open addressing ..

Thank you for the comments. I assume it does the usual
extensible hashing mechanism, extending the hash value
by one bit, and the table by a factor of two, when it
is time to grow? I'm glad to read that it uses open
probing. I think this makes more sense for tables that
reside in memory, and that have no strong granularity.

> if-you-do-have-frequent-collisions-i'll-consider-it-a-bug-ly y'rs

What hash function does it use?

Russell



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


From sbrunning at bigfoot.com  Tue Dec  5 11:37:43 2000
From: sbrunning at bigfoot.com (Simon B.)
Date: Tue, 05 Dec 2000 16:37:43 GMT
Subject: Turning a flat file into a hierarchical structure
Message-ID: <90j5km$9sv$1@nnrp1.deja.com>

I have a tuple which contains a representation of a heirarcical
structure, looking something like this:

testdata = ((1, 'egg '), (2, 'bacon'), (2, 'sausage'), (3, 'spam'),
(4, 'tomato'), (2, 'beans'))

(In real world, this tuple is populated from a text file)

I want to turn this into a tree structure, looking something like this:

[(1, 'egg ', [(2, 'bacon', []), (2, 'sausage', [(3, 'spam',
[(4, 'tomato', [])])]), (2, 'beans')])]

I'm think that I'm nearly there, but I'm stuck on how to deal with
climbing up the tree, that is to say, dealing with the transition from
a lower to a higher level. My existing code looks like this:

mylist = []
appendlist = mylist
lastlevel = testdata[0][0]

for level, text in testdata:
    if level == lastlevel:
        appendlist.append((level, text, []))
    elif level > lastlevel:
        appendlist = appendlist[-1][-1]
        appendlist.append((level, text, []))
    if level < lastlevel:
        break # What here?
    lastlevel = level

print mylist

Can anyone help me out?

--
Simon Brunning
The sooner you fall behind, the more time you'll have to catch up. -
Steven Wright


Sent via Deja.com http://www.deja.com/
Before you buy.


From sholden at holdenweb.com  Mon Dec 18 09:51:55 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 18 Dec 2000 09:51:55 -0500
Subject: Alarms in threads - how to get/mimic?
References: <Pine.LNX.4.21.0012161749380.7913-100000@traal.eecs.berkeley.edu> <ltd7erk8xd.fsf@asfast.com>
Message-ID: <kzp%5.39167$n51.702446@e420r-atl1.usenetserver.com>

Lloyd Zusman <ljz at asfast.com> wrote in message
news:ltd7erk8xd.fsf at asfast.com...
> Luca de Alfaro <dealfaro at eecs.berkeley.edu> writes:
>
> > [ ... ]
> >
> > What is not clear to me is the following.  When there is a timeout,
> > the exception Timeout is not caught by httplib.  Hence, if the thread
> > that called httplib just continues its job giving up on the
> > connection, httplib will not have a chance to do any clean-up, such as
> > for example releasing/closing the socket.
> >
> > Hence, it seems to me that in order to abort a web page load cleanly
> > upon timeout, I also have to modify the httplib module to handle the
> > exception Timeout, and do clean-up work if that happens.
> >
> > Is my understanding correct?
>
> Not entirely.  When using timeoutsocket, all you have to do is start
> catching timeout exceptions when you make the various method calls
> within httplib
[snip code]

Basically, if the code you make use of traps the errors raised by
timeoutsocket and re-raises them after it's done whatever cleanup it needs
to then there should be no problem.  However, there MAY be problems if the
timeoutsocket error is not caught by the higher-level protocol code, and
makes it to the outer code which called it.  In that case there is a
possibility that necessary clean-up may not be performed.
>
> It's recommended in Python that exception handling be done whenever
> it's appropriate; therefore, by wrapping your httplib calls in this
> same manner, you'll remain true to the Pythonic Way, and you won't
> have to re-write any httplib code when using timeoutsocket.
>
> Note that the timeoutsocket.setDefaultSocketTimeout() method call
> should be made before making any httplib calls.  If you want to use
> different timeout values when making different httplib calls (or other
> calls socket-based methods), you'll have to re-invoke the
> setDefaultSocketTimeout() method before each call that uses a changed
> timeout value.
>
Note also that you can set the timeout on each socket individually.
setDefaultSocketTimeout is usually used to ensure that connect() attempts in
upper-layer protocols (which happen before you have a socket object in your
application to manipulate) don't hang forever.

regards
 Steve







From jcm at bigskytel.com  Thu Dec 21 03:07:53 2000
From: jcm at bigskytel.com (David Porter)
Date: Thu, 21 Dec 2000 01:07:53 -0700
Subject: Replay commands
In-Reply-To: <20001221003637.A17989@bigskytel.com>; from jcm@bigskytel.com on Thu, Dec 21, 2000 at 12:36:38AM -0700
References: <91rep9$tev$1@nnrp1.deja.com> <20001221003637.A17989@bigskytel.com>
Message-ID: <20001221010753.C17989@bigskytel.com>

* David Porter <jcm at bigskytel.com>:
> * pinhong at eecs.berkeley.edu <pinhong at eecs.berkeley.edu>:
> 
> > When using Python interactively, I need to save my command history
> > for editing and reuse. Is that possible?
> 
> Yes.
> 
> An example of it being done can be found in IPP, the Interactive Python

<lot's of crap snipped>

> Prompt at <http://starship.python.net/crew/jhauser/miscellany.html>.
> 
> 
> David

I'm sorry about that!



From loredo at spacenet.tn.cornell.edu  Thu Dec  7 17:00:23 2000
From: loredo at spacenet.tn.cornell.edu (Tom Loredo)
Date: Thu, 07 Dec 2000 17:00:23 -0500
Subject: Calculating factorial
References: <3a2a9496.647995@uutiset.nic.fi> <90ghoh0lh@news2.newsguy.com> <9tkp2tsfc0hjq3dr7vc69a6hdf0at3q0rk@4ax.com>
Message-ID: <3A300877.E6A4C515@spacenet.tn.cornell.edu>

Howdy-

The functions below are Python versions of functions appearing in the
*Numerical Recipes* books by Press, Teukolsky, etc..  factrl(n)
returns n! as a float; factln(n) returns the natural logarithm
of n! as a float.  Both store values calculated for lowish n to
speed subsequent calls.  They are useful for calculations in
statistics, but probably not what you'd want to use for combinatorics.
Just wrote them a couple days ago; use at your own risk!

This is pure Python for portability; for speed you might consider
using gamma function calls from the cephes library.

Peace,
Tom Loredo

import Numeric as N
from math import log, exp

gammln_cof = N.array([76.18009173, -86.50532033, 24.01409822,
   -1.231739516e0, 0.120858003e-2, -0.536382e-5])
gammln_stp = 2.50662827465

def gammln(xx):
	"""Logarithm of the gamma function."""
	global gammln_cof, gammln_stp
	x = xx - 1.
	tmp = x + 5.5
	tmp = (x + 0.5)*log(tmp) - tmp
	ser = 1.
	for j in range(6):
		x = x + 1.
		ser = ser + gammln_cof[j]/x
	return tmp + log(gammln_stp*ser)
	
def factrl(n, ntop=0, prev=N.ones((33),N.Float)):
    """Factorial of n.
    The first 33 values are stored as they are calculated to
    speed up subsequent calculations."""
    if n < 0:
        raise ValueError, 'Negative argument!'
    elif n <= ntop:
        return prev[n]
    elif n <= 32:
        for j in range(ntop+1, n+1):
            prev[j] = j * prev[j-1]
        ntop = n
        return prev[n]
    else:
        return exp(gammln(n+1.))

def factln(n, prev=N.array(101*(-1.,))):
    """Log factorial of n.
    Values for n=0 to 100 are stored as they are calculated to
    speed up subsequent calls."""
    if n < 0:
        raise ValueError, 'Negative argument!'
    elif n <= 100:
        if prev[n] < 0:
            prev[n] = gammln(n+1.)
        return prev[n]
    else:
        return gammln(n+1.)


From aleaxit at yahoo.com  Tue Dec 26 19:21:49 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 27 Dec 2000 01:21:49 +0100
Subject: How to question
References: <mailman.977863690.15082.python-list@python.org>
Message-ID: <92bcrj030ge@news1.newsguy.com>

"Daniel Kinnaer" <Daniel.Kinnaer at AdValvas.be> wrote in message
news:mailman.977863690.15082.python-list at python.org...
> How can one 'clear' or 'cls' the interactive window from PythonWin or form
> IDLE? That is : I'd like to clear/cls this window so that all previous
> output in the interactive window is erased and only the last 'run' is
shown
> in the interactive window...  Thanks

In PythonWin, Ctrl-A (Select-All) then hit Delete (delete-selection).

In IDLE, I'm not sure.  Alt-A does the Select-All, but then I don't
know how to delete-selection.


Alex





From moshez at zadka.site.co.il  Mon Dec 18 22:54:23 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Tue, 19 Dec 2000 05:54:23 +0200 (IST)
Subject: while true: !!!
In-Reply-To: <20001218085229.A10917@westsidecnc.rh.rit.edu>
References: <20001218085229.A10917@westsidecnc.rh.rit.edu>, <B43D149A9AB2D411971300B0D03D7E8B0FBCBB@natasha.auslabs.avaya.com> <v04220801b663769e1800@[192.168.11.33]>
Message-ID: <20001219035423.1093FA84F@darjeeling.zadka.site.co.il>

On Mon, 18 Dec 2000 08:52:29 -0500, D-Man <dsh8290 at rit.edu> wrote:

> Thus while true: is equal to while 1: except that a boolean, not an
> integer is used for the test.  I, for one, would like to abolish the
> concept that an integer can denote true or false.  How many functions
> return 0 for success?  (just take a look at the standard C library).

Luckily, Python functions denote failure by exceptions, and success
is implicit. Returning a number (1/0) as a boolean is an accepted practice,
and is usually followed in the Python library. Exceptions are things
like os.system, which are thin wrappers to C calls. 
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From amk at mira.erols.com  Tue Dec 12 10:00:51 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 12 Dec 2000 15:00:51 GMT
Subject: Backslash escape in regular expressions
References: <jon.976507791@bezek.dstc.monash.edu.au> <3A34D69C.BAC68E27@engcorp.com> <jon.976591746@bezek.dstc.monash.edu.au>
Message-ID: <slrn93cg5s.j9.amk@207-172-113-220.s220.tnt5.ann.va.dialup.rcn.com>

On 12 Dec 2000 14:55:12 +1100, 
	Jonathan Giddy <jon at bezek.dstc.monash.edu.au> wrote:
>but instead matches the Python string '(hello)\n'.  I prefer the latter,
>but think the documentation should indicate that \a, \f, \n, \t, \v, and
>\x are also "special sequences".

It probably should mention that.  Care to submit a patch to libre.tex?

--amk 


From fredrik at effbot.org  Fri Dec 29 16:31:29 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 29 Dec 2000 21:31:29 GMT
Subject: 2001 Enchancement Wishlist
References: <3A4CEACB.220998A0@javanet.com>
Message-ID: <Ro736.2510$Qb7.273065@newsb.telia.net>

Raymond Hettinger wrote:
> 2. There is already a PEP to extend the eval() function which is currently
>     limited to dictionaries.  The goal is to accept any object defining
>     __getitem__() so that any mapping object can be used to specify the
>    globals and locals.   I recommend that we extend the same courtesy to
>    the % string formatting operator.  For example:
>
>              print "Spam with %s, eggs, and spam" % aUserDictionary

this already works -- if you use dictionary formatting
operators:

class D:
    def __getitem__(self, name):
        return name

d = D()

print "%(spam)s %(egg(10))s" % d

# prints spam egg(10)

</F>




From tim.one at home.com  Thu Dec  7 02:36:41 2000
From: tim.one at home.com (Tim Peters)
Date: Thu, 7 Dec 2000 02:36:41 -0500
Subject: Bug: Readline getting stuck on Linux and Solaris
In-Reply-To: <kuk89cbtdw.fsf@lasipalatsi.fi>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEMPICAA.tim.one@home.com>

[posted & mailed]

[Erno Kuusela]
> ..
> ps. anyone know if there's anything new wrt the license story?  last i
> heard the respective lawyers were still out on the subject. ...

The lawyers are still out, but are still talking.  There's nothing
definitive to show for it over the last few months, except that they may--
or may not --be making progress.  That they're all still talking tells me
that all sides still *want* to resolve it.

or-maybe-just-that-they-enjoy-each-other's-company<wink>-ly y'rs  - tim




From ngps at madcap.dyndns.org  Sat Dec  9 11:11:44 2000
From: ngps at madcap.dyndns.org (Ng Pheng Siong)
Date: 9 Dec 2000 16:11:44 GMT
Subject: SSL how to?
References: <yZ3Y5.3016$9B.89007@news1.oke.nextra.no> <p6qu28eiqig.fsf@informatik.hu-berlin.de> <90s40v$f6d$1@troll.powertech.no> <p6q1yvhc1vr.fsf@informatik.hu-berlin.de>
Message-ID: <90tlk0$pps$1@violet.singnet.com.sg>

According to Martin von Loewis  <loewis at informatik.hu-berlin.de>:
> The support is in the source code, both for Unix and Windows. It is
> not included in any of the binary distributions. For Unix, it is
> relatively simple to recompile the socket module so that it uses
> OpenSSL; for Windows (including NT) people have reported problems
> doing so. In any case, you need a working copy of OpenSSL for your
> target system, first.

<plug>
You can also use M2Crypto, which is compatible with both Python 1.5.2
and Python 2.0. Win32 binaries for Python 2.0 are available.

Here: http://www.post1.com/home/ngps/m2
</plug>


Cheers.


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



From blosh at tsbrass.com  Wed Dec 13 15:32:56 2000
From: blosh at tsbrass.com (Brad)
Date: Wed, 13 Dec 2000 20:32:56 GMT
Subject: smtplib
Message-ID: <918mdo$7d9$1@nnrp1.deja.com>

anyone know what this means...I am running Python 1.6 on HPUX10.20

smtplib.SMTPHeloError: (504, 'DNS verification of sending machine failed: no
mail will be accepted')


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


From aahz at panix.com  Mon Dec 18 14:58:19 2000
From: aahz at panix.com (Aahz Maruch)
Date: 18 Dec 2000 11:58:19 -0800
Subject: Timers in Python?
References: <3a3e42ef.16144323@news.skynet.be> <91li8l$33h$1@reader1.imaginet.fr> <3a3e61cf.24049354@news.skynet.be>
Message-ID: <91lq8r$gu2$1@panix6.panix.com>

In article <3a3e61cf.24049354 at news.skynet.be>,
Daniel <Daniel.Kinnaer at AdValvas.be> wrote:
>
>class TimeScheduled(threading.Thread):
>    def __init__(self, function, delay):
>        self.function = function
>        self.delay = delay

Gilles made a boo-boo.  Here's the correct code:

class TimeScheduled(threading.Thread):
    def __init__(self, function, delay):
        threading.Thread.__init__(self)
        self.function = function
        self.delay = delay

For more info on threading, take a look at 
http://starship.python.net/crew/aahz/
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

This is a signature anti-virus. 
Please stop the spread of signature viruses!


From sh at ttsoftware.co.uk  Mon Dec 18 06:30:03 2000
From: sh at ttsoftware.co.uk (Steve Horne)
Date: Mon, 18 Dec 2000 11:30:03 +0000
Subject: what is a closure?
References: <3a39033b@news.xtml.co.uk> <S18_5.3929$Kd1.455131@newsb.telia.net> <8nck3tocsg8kp1u2g1du6nd3lt9gfeckr2@4ax.com> <91dkvu$7mk$1@nnrp1.deja.com>
Message-ID: <ulsr3tctvq7m03hhsofogrl3r6cn3e9f8q@4ax.com>

On Fri, 15 Dec 2000 17:39:11 GMT, Jeremy Hylton <jeremy at alum.mit.edu>
wrote:

>A closure provides a way to resolve free variables in a function body;
>it is created when the function is defined. Currying is a way to create
>a binding for a formal parameter; it is used when a function is called

OK, so if I could write...

  def Test (a) :
    return lambda p : p + a

in Python, then this would be a closure.

However, what if I write...

  def Test :
    a = 1
    b = lambda p : p + a

    a = 2
    c = lambda p : p + a

    print b (1)
    print c (1)

Would the result be...

  2
  3

Or would it be...

  3
  3

In short, for the standard behaviour of a closure, would the value of
a be bound at the time the lambda is defined, or is some kind of
reference created.

-- 
Steve Horne
Home : steve at lurking.demon.co.uk
Work : sh at ttsoftware.co.uk


From Bill.Scherer at VerizonWireless.com  Tue Dec 12 14:53:41 2000
From: Bill.Scherer at VerizonWireless.com (Bill Scherer)
Date: Tue, 12 Dec 2000 14:53:41 -0500
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <r0pzoi11lhz.fsf@black.cs.washington.edu>
Message-ID: <3A368245.B4FC337@VerizonWireless.com>

Daniel Wood wrote:

> Here is what I wanted to do.  I have some random parameters, lambda

Beware: lambda is a Python keyword...



From tim.one at home.com  Sat Dec 23 02:49:55 2000
From: tim.one at home.com (Tim Peters)
Date: Sat, 23 Dec 2000 02:49:55 -0500
Subject: Implementation of diff algorithms in python?
In-Reply-To: <921ac8$lug$1@bunyip.cc.uq.edu.au>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEDLIFAA.tim.one@home.com>

[Chui Tey]
> Are there any implementations of diff algorithms in python?

Probably <wink>.  You can find a novel one in Tools/scripts/ndiff.py.  But,
offhand, I'm not aware of others.




From scarblac at pino.selwerd.nl  Mon Dec  4 21:02:12 2000
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 5 Dec 2000 02:02:12 GMT
Subject: int multiplication overflow and the future of python
References: <ghaelzoc9lp.fsf@synopsys.com>
Message-ID: <slrn92oivs.duu.scarblac@pino.selwerd.nl>

Ram Bhamidipaty <ramb at synopsys.com> wrote in comp.lang.python:
> On red hat 7.0 with python 1.6 I am running into overflow problems.
> 
> From looking at the bug database at sourceforge I can see
> that other people have run into this, and it looks like
> there is a fix....but I have not been able to find it. Is there
> a patch that I should apply to upgrade my 1.6 source to the
> latest version?

You should probably upgrade to 2.0. What sort of overflow problems
are you running into, btw?

> This leads to a larger question: Which version is being maintained?
> Is active developement taking place for both 1.6 and 2.0? I've been
> using 1.6 because the features in 2.0 seem like they involve bigger
> changes to the code base and I thought I should wait until the
> initial bugs are worked out. But if all new work is only going
> into the 2.x versions then I should probably stop using 1.6...
> 
> Any thoughts?

The only reason 1.6 ever came out was that CNRI, the old copyright holders,
wanted a version with a license they regarded as good. When it came out,
development had already progressed much further at BeOpen.

2.0 is the current Python version. There might be some bugs in it, but
I'd expect there to be a lot more in 1.6. Version 1.5.2 was really stable,
and 2.0 is good.

-- 
Remco Gerlich


From phd at phd.pp.ru  Thu Dec 14 08:06:52 2000
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 14 Dec 2000 16:06:52 +0300 (MSK)
Subject: split weirds
In-Reply-To: <QI2l1FAXBMO6EwD6@jessikat.demon.co.uk>
Message-ID: <Pine.LNX.4.30.0012141605520.14664-100000@fep132.fep.ru>

On Thu, 14 Dec 2000, Robin Becker wrote:
> Can somebody explain why strip behaves differently with and without an

   "split", not "strip" :)

> argument?
[skip]
> with the default whitespace arg seems as though a run of whitespace is
> being treated as a single character.

   It is documented behaviour: "If sep is not specified or None, any
whitespace string is a separator."

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




From donn at u.washington.edu  Tue Dec 19 14:20:29 2000
From: donn at u.washington.edu (Donn Cave)
Date: 19 Dec 2000 19:20:29 GMT
Subject: FIFO in python?
References: <3A3FAB7A.DC63610C@NOraqiaSPAMMY.com>
Message-ID: <91ocdt$80qg$1@nntp6.u.washington.edu>

Quoth David Lees <DavidL at NOraqiaSPAMMY.com>:
| I am trying to learn how to use FIFO and am unable to open a fifo under
| Linux RedHat 6.2.  I hang on the open (open(fn,'rw'), even though I have
| done a successful check for file existence (os.path.exists(fn)).  The
| same code works under NT.  Any thoughts on what is going on?

>From what I can tell, "rw" does not have any defined meaning in
open().  Try "r+" for read/write.

Read-only open blocks on a fifo, waiting for a write.  I'm kind
of surprised myself, would have expect the read() to block.

	Donn Cave, donn at u.washington.edu


From Daniel.Kinnaer at AdValvas.be  Mon Dec 18 09:06:33 2000
From: Daniel.Kinnaer at AdValvas.be (Daniel)
Date: Mon, 18 Dec 2000 14:06:33 GMT
Subject: search for pattern in StringList
Message-ID: <3a3e1935.5461298@news.skynet.be>

As a Python newbie, I was trying a very simple exercise, but got
stuck. Is there someone who can help me complete this code, please?

Thanks.
Daniel


MyList = ['an','been','call','door']

print "Scanning a StringList"

for i in range(0, len(MyList)):
    print i, MyList[i]

#try to show every item in MyList which contains an 'a'
#that is 'an' and 'call'.   In Pascal :
#for i:=0 to MyList.Items.Count -1 do
#    if pos('a',MyList.Items[i])>0 then writeln(MyList.Items[i]) 
#in Python : ???




From Lutz.Schroeer at tu-clausthal.de  Wed Dec 13 14:34:34 2000
From: Lutz.Schroeer at tu-clausthal.de (Lutz Schroeer)
Date: 13 Dec 2000 19:34:34 GMT
Subject: Tutorial for emacs python-mode?
Message-ID: <Xns9009D2CA0Latzikatz@139.174.2.56>

Hi,

after noticing that tere's a python-mode for Emacs I decided to switch to 
that Editor but unfortunately I can't find any tutorial for using this mode  
:-(

Any hints, links,...?

Lutz


From ommadawn at club-internet.fr  Wed Dec 13 08:21:26 2000
From: ommadawn at club-internet.fr (Christophe Prevost)
Date: Wed, 13 Dec 2000 13:21:26 -0000
Subject: Probleme with locale
Message-ID: <917pjb$ii2$1@reader1.imaginet.fr>

Hi,

I need locale for my re in french. And there's something strange...

import locale
this work ->
>>>locale.setlocale(locale.LC_ALL, "")
>>>locale.getdefaultlocale()
('fr_FR', 'cp1252')

this doesn't work ->
>>>locale.setlocale(locale.LC_ALL, "fr_FR")

Tell me Why?







From Daniel.Kinnaer at AdValvas.be  Sun Dec 24 05:34:00 2000
From: Daniel.Kinnaer at AdValvas.be (Daniel)
Date: Sun, 24 Dec 2000 10:34:00 GMT
Subject: About WebBrowser
References: <91vbr9$eo1q2@SGI3651ef0.iddeo.es> <9235bj$6e6$1@troll.powertech.no> <9236f3$6rb$1@troll.powertech.no> <3a45260d.52012638@news.skynet.be> <924a8b0ckb@news1.newsguy.com>
Message-ID: <3a45d0bc.1317426@news.skynet.be>

On Sun, 24 Dec 2000 08:55:22 +0100, "Alex Martelli"
<aleaxit at yahoo.com> wrote:


>
>You probably have not installed the win32all extensions from ActiveState
>(or the ActiveState distribution of Python, that comes with them); or else
>something has gone awry in the installation process (seems unlikely).
>
>
>Download the win32all extensions from ActiveState and install them.
>http://www.activestate.com/Products/ActivePython/win32all.html
>If you have Python 2, you need to get build 135 of win32all specifically
>(a single URL, possibly split by news-handling software...):
>http://www.activestate.com/download/ActivePython/windows/win32all/win32all-1
>35.exe

Thanks guys.  I've just downloaded ActivePython and will give it a go.

Happy Holidays!

Daniel


From tim.one at home.com  Tue Dec 26 16:26:47 2000
From: tim.one at home.com (Tim Peters)
Date: Tue, 26 Dec 2000 16:26:47 -0500
Subject: One Python 2.1 idea
In-Reply-To: <9292tn$jka$1@nnrp1.deja.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEIJIFAA.tim.one@home.com>

[Russell Turpin]
> ...
> First, I hope Alex Martelli does look at how list
> comprehension is implemented. Even though I am the one
> who has argued that the language features of Python are
> quite mature, I think list comprehension is just the
> right kind of language enhancement, raising the level
> of abstraction in a way that extends seamlessly the
> existing notation.

Same here!  listcomps were a fine addition.

> Second, a question. Consider the invocation of a method,
> such as:
>
>     x.foo()
>
> Does this really cause the lookup of the string "foo"
> in x's namespace dictionary? The first time? Every time?
> Is this also the case when data members are accessed?

First and every time, but that was covered in other msgs.

It's not quite as simple as that, though:  Python interns strings in source
code that "look like" they might be attribute names.  As a result, "foo" is
stored uniquely, and in most cases the lookup of "foo" in the class dict
will succeed via pointer equality:  in most cases no string comparison will
actually be done at runtime, nor will runtime code recompute the hash code
for "foo" (the hash code is cached in the interned string object).  This is
true (no string compare) even if you do

    getattr(x, "foo")

It's *not* true, though, if you do

    getattr(x, "f" + "oo")

Strings synthesized (computed) at runtime are not interned by magic, so the
lookup code ends up dealing with a distinct copy of "foo" and needs to do a
string compare to determine equality.  OTOH,

    getattr(x, intern("f" + "oo"))

again succeeds via cheap pointer equality (which is no real gain, since the
string compare is hiding in the intern() now -- but is a real gain if you
reuse the explicitly interned string in more lookups).

Of course the "cheap pointer equality" doesn't get hit until we're several
layers deep in C function calls, so it's not like this is zippy:  it's just
quicker than you expect at first -- and the time isn't being spent where you
first guess it's being spent.

the-implementation-is-cleverer-than-the-programmer-model-ly
    y'rs  - tim




From vonehr at ira.uka.de  Mon Dec  4 10:30:35 2000
From: vonehr at ira.uka.de (Markus von Ehr)
Date: Mon, 04 Dec 2000 16:30:35 +0100
Subject: mp3 module
Message-ID: <3A2BB89B.D4C30300@ira.uka.de>

Hi,

is there a module to play mp3 files with python?

Thanks for answers,

Markus

Markus von Ehr
vonehr at ira.uka.de


From gregj at pobox.com  Mon Dec 25 04:55:04 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Mon, 25 Dec 2000 09:55:04 GMT
Subject: soundex (revisited)
References: <Var16.263$LU6.109277@typhoon.aracnet.com>
Message-ID: <YPE16.190877$U46.5871952@news1.sttls1.wa.home.com>

"Daniel Klein" <DanielK at aracnet.com> wrote in message
news:Var16.263$LU6.109277 at typhoon.aracnet.com...
> After seeing the post from several days ago on soundex, I gave it whirl to
> see if I could come up with something different (and possibly better),
> following the rules laid down by Knuth:
>
> def get_soundex(name, digits = 3):
>     soundexcodes = "01230120022455012623010202"
>     #               ABCDEFGHIJKLMNOPQRSTUVWXYZ
>     instring = name.upper()
>     soundex = instring[0]
>     last = soundex
>     instring = instring[1:]
>     for char in instring:
>         if 65 <= ord(char) <= 90:
>             sx = soundexcodes[ord(char) - 65]
>             if int(sx) and char != last:
>                 soundex += sx
>                 last = char
>     if len(soundex) < (digits + 1): soundex = (soundex + ("0" * digits))
>     return soundex[:digits + 1]

I see a few problems, mainly in the handling of consecutive consonants. You
are checking for consecutive characters, but the Soundex algorithm specifies
that consecutive character codes be treated as a single code. Both 'mm' and
'mn' are considered consecutive codes because both 'm' and 'n' are coded as
5.

You can (and probably should) use the isalpha() string method to check for
alpha characters, rather than the 'magic numbers' 65 through 90. Likewise
ord(char) - ord('A') is a bit more clear.

Here's a version I wrote. I'm open to any criticisms, suggestions, etc. I
compared my version to the module announced here a while back (I think mine
is a lot more readable; it is certainly shorter). I also compared it to a
Perl version I found and I think my implementation is more robust and
smaller.

def soundex(name, len=4):
    """ soundex module conforming to Knuth's algorithm
        implementation 2000-12-24 by Gregory Jorgensen
        public domain
    """

    # digits holds the soundex values for the alphabet
    digits = '01230120022455012623010202'
    sndx = ''
    fc = ''

    # translate alpha chars in name to soundex digits
    for c in name.upper():
        if c.isalpha():
            if not fc: fc = c   # remember first letter
            d = digits[ord(c)-ord('A')]
            # duplicate consecutive soundex digits are skipped
            if not sndx or (d != sndx[-1]):
                sndx += d

    # replace first digit with first alpha character
    sndx = fc + sndx[1:]

    # remove all 0s from the soundex code
    sndx = sndx.replace('0','')

    # return soundex code padded to len characters
    return (sndx + (len * '0'))[:len]


--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From aleaxit at yahoo.com  Thu Dec 21 09:56:47 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 21 Dec 2000 15:56:47 +0100
Subject: Editing files
References: <Bdd06.5591$K6.674078@juliett.dax.net>
Message-ID: <91t5o102hgl@news2.newsguy.com>

"dv" <lars at stea.nu> wrote in message
news:Bdd06.5591$K6.674078 at juliett.dax.net...
> I'm trying to automate some configuration tasks with python. Is it
possible
> to do a "search and replace" or specify where (which line) in a file
python
> will write?

In modern systems (Unix, Linux, Windows, Mac, ...) files are not
internally 'line oriented', but rather byte-oriented.  So, you
can specify the *byte* offset (with a file object's "seek" method),
but not (easily or directly) "which line".

For files that are not too huge, BY FAR the simplest thing is to
get them all into memory as a list of lines, in one go:

    thelines = thefile.readlines()

now you can modify the lines with great ease, and no problem
at all, and finally write them all out back to the file, after
a thefile.seek(0) to be sure [so you'll write from the start!].
thefile.writelines(thelines) will do it.

Having the file be readable AND rewritable requires opening it
with a "r+" option, by the way, right at the start:
    thefile = open("what.txt","r+")

So, putting it all together, it's simple indeed:

    thefile = open("what.txt","r+")
    thelines = thefile.readlines()
    #
    # here, change thelines as you wish
    #
    thefile.seek(0)
    thefile.writelines(thelines)
    thefile.truncate()
    thefile.close()

The .truncate is only needed if you may have shortened the
total number of bytes in thelines, and it won't be OK on
ALL platforms (but I think Windows and Linux support it).


You may be selective in your _rewriting_, with a little
effort -- overwriting just SOME specific lines -- but that
will work only if your modifications never change the
length of any line, which is quite a rare situation --
probably not worth worrying about.  And you do have to
read the file at the start, to learn the byte offsets
at which the various lines start, so the memory taken/
computer effort doesn't change much, while programming
it IS more complex.  So, I'd skip it, if feasible.


Alex





From rturpin at my-deja.com  Sun Dec 24 15:19:54 2000
From: rturpin at my-deja.com (rturpin at my-deja.com)
Date: Sun, 24 Dec 2000 20:19:54 GMT
Subject: One Python 2.1 idea
References: <m38zp6nk8k.fsf@localhost.localdomain> <mailman.977684587.22689.python-list@python.org>
Message-ID: <925lp7$d2c$1@nnrp1.deja.com>

In article <mailman.977684587.22689.python-list at python.org>,
  "Tim Peters" <tim.one at home.com> wrote:
[Lieven Marchand]
>> I think Common Lisp shows that getting very good
>> speed in Python is quite feasable. ..

In article <mailman.977684587.22689.python-list at python.org>,
  "Tim Peters" <tim.one at home.com> wrote:
> The reasons vary, in large part depending on whether
> or not they've ever worked on a compiler <wink>.  Whatever,
> the fact is that there isn't a Python compiler, nor a
> credible plan for such a beast to come into existence.

Making Python faster doesn't require the creation of an
optimizing compiler, complete with data and control flow
analysis, loop unrolling, expression optimization, code
unfolding, etc.

There likely would be a lot of bang for the buck from
small improvement to the existing infrastructure. Can the
code for list comprehensions be made faster? Does the
existing front-end pre-execute all constant (sub)expressions?
Does it pre-construct fields and methods that are NOT
conditional in class and constructor definitions? Does it
bind constant references to these as tightly as possible?
Are there extensions to the byte code that would increase
efficiency? (And maybe I'm wrong. Maybe most of the low
hanging fruit has been plucked.)

I guess this is my fault, for saying "speed." And yeah, I
would love a full, optimizing, JIT compiler with portable
code generation, so it lives on all platforms. But we all
know that isn't happening any time soon, unless some
company bellies up to the bar with money to make it happen.

I was pushing merely for modest efforts in this direction.
I can only push, since my time is devoted elsewhere.

Russell


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


From root at rainerdeyke.com  Tue Dec 19 13:24:15 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Tue, 19 Dec 2000 18:24:15 GMT
Subject: while true: !!!
References: <B43D149A9AB2D411971300B0D03D7E8B0FBCBB@natasha.auslabs.avaya.com> <v04220801b663769e1800@[192.168.11.33]> <mailman.977147642.18708.python-list@python.org> <91lclt02qr3@news1.newsguy.com> <slrn93st0m.6c1.qrczak@qrnik.zagroda> <91o3l501i59@news2.newsguy.com>
Message-ID: <jJN%5.52075$x6.23325177@news2.rdc2.tx.home.com>

"Alex Martelli" <aleaxit at yahoo.com> wrote in message
news:91o3l501i59 at news2.newsguy.com...
> "Marcin 'Qrczak' Kowalczyk" <qrczak at knm.org.pl> wrote in message
> news:slrn93st0m.6c1.qrczak at qrnik.zagroda...
> > OTOH Python does not equate characters with integers. C, C++ and Java
> > do.
>
> Wrong!  C's types are always slightly fuzzy, but that does
> NOT apply to C++ or Java; e.g., in C++:
>
> void f(int) {
>     cout << 'integer\n';
> }
>
> void f(char) {
>     cout << 'character\n';
> }

'char' is an integer type, same as 'short' and 'long'.  'int' is just
another integer type.  It is possible (and often desirable) to use 'char's
(or 'unsigned char's) to hold small numeric values.  The same does not apply
to single character strings in Python.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From sandj.williams at gte.net  Sat Dec 30 23:39:36 2000
From: sandj.williams at gte.net (Steve Williams)
Date: Sun, 31 Dec 2000 04:39:36 GMT
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <3A4D6A05.EEC97FA6@uswest.net> <slrn94qqvg.mgg.grey@teleute.rpglink.com> <ije36.96813$x6.43604873@news2.rdc2.tx.home.com> <slrn94qt4f.mgg.grey@teleute.rpglink.com> <CMo36.97857$x6.44474719@news2.rdc2.tx.home.com> <slrn94s9qg.134.grey@teleute.rpglink.com> <gdq36.2696$Qb7.302774@newsb.telia.net> <slrn94shbe.150.grey@teleute.rpglink.com> <X8s36.2731$Qb7.305339@newsb.telia.net>
Message-ID: <3A4EB9B5.B4B60F4C@gte.net>

Fredrik Lundh wrote:

> [snitt]

>
> (besides, no matter what you think, "do-while" is spelled "while
> 1-if-break" in Python.  no amount of arguing on comp.lang.python
> will change that)
>
> </F>

Do you have trouble porting your code to other languages?




From fgeiger at datec.at  Thu Dec 28 14:28:11 2000
From: fgeiger at datec.at (Franz GEIGER)
Date: Thu, 28 Dec 2000 20:28:11 +0100
Subject: 8.3 peculiarity if run w/i console
Message-ID: <92g4dn$5ug$1@newsreaderm1.core.theplanet.net>

If I run a script within the Python 2.0 console, the names of the files that
are made by that script are all "abbreviated", e.g.:
ALongFileName.txt becomes ALongF~1.TXT. If I run the script within
PythonWin, the names are o.k.
Does anybody know what's going on here?

FYI: ActiveState Python2.0/202 on WinNT 4.0 Svr.

Thanks in advance and
Best regards
Franz GEIGER





From aahz at panix.com  Mon Dec 11 14:57:42 2000
From: aahz at panix.com (Aahz Maruch)
Date: 11 Dec 2000 11:57:42 -0800
Subject: threads and dictionaries
References: <9139fq$ovj$1@nnrp1.deja.com>
Message-ID: <913bjm$6pn$1@panix3.panix.com>

In article <9139fq$ovj$1 at nnrp1.deja.com>,  <bartig at pinpoint.com> wrote:
>
>I'm writing a program that passes dictionaries into a subroutine that's
>threaded.  Are dictionaries thread safe?  If so how would I go about
>making changes to the dictionaries inside of a thread so that other
>threads can take advantage this?

Dictionaries are as thread-safe as any other Python variable.  That is,
if there's any chance that one thread will try to perform multiple
operations on a dict at the same time that any other thread will try to
perform other operations, you'd better stick in a critical section lock.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

What if there were no rhetorical questions?  --Aahz


From aahz at panix.com  Fri Dec 22 10:16:18 2000
From: aahz at panix.com (Aahz Maruch)
Date: 22 Dec 2000 07:16:18 -0800
Subject: How to make a GUID with plain python
References: <3a413a17.3416402@192.168.0.1> <91voot$cs5$1@violet.singnet.com.sg>
Message-ID: <91vr82$nkp$1@panix6.panix.com>

In article <91voot$cs5$1 at violet.singnet.com.sg>,
Ng Pheng Siong <ngps at post1.com> wrote:
>According to Georg Gogo. BERNHARD <gogo at bluedynamics.com>:
>> I am looking for a way to create a global unique identifier / UUID
>> with python...
>
>Here's an expired I-D describing "the UUID formats and various ways
>of generating them".
>
>    http://www.shore.net/~rsalz/uuid-i-d.txt

That appears to have moved to
http://members.bellatlantic.net/~vze2552x/uuid-i-d.txt

(I was curious to see whether "~rsalz" was Rich Salz.)
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

I don't really mind a person having the last whine, but I do mind
someone else having the last self-righteous whine.  --Aahz


From aleaxit at yahoo.com  Thu Dec 28 17:52:04 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 28 Dec 2000 23:52:04 +0100
Subject: Python vs. PHP (& Java?)
References: <69BC4C6AE83ED311BC9400805FA7B14AF84D2C@il93exp01.css.mot.com> <mailman.978018791.20357.python-list@python.org> <92fv4a$l2l$1@nnrp1.deja.com>
Message-ID: <92gghv02iq1@news2.newsguy.com>

<rturpin at my-deja.com> wrote in message news:92fv4a$l2l$1 at nnrp1.deja.com...
    [snip]
> In contrast, texts in computer languages are more studied
> as artifacts. "How did this work?" "Why did it stop
> working?" "What did the original authors intend?" They,
> the rascals, are no longer around to explain, take credit,
> or defend their honor. This kind of comprehension requires
> a high level of competency in the language.

Are you implying that this _doesn't_ happen a lot for
many written texts in natural languages...?  Literary
analysis is all about such work, after all -- and courts
and lawyers do a lot of that on laws and constitution
articles.

> Knowing one-tenth of Perl is not enough to understand
> most Perl scripts.

Of course not, but, it IS enough to _write_ a lot of somewhat
useful scripts -- I doubt most casual Perl users are above 10%
in their knowledge and understanding of that extremely rich,
complex, and idiosyncratic programming language.  Yet, use
it they do -- and they get some useful work done thereby (as
well, no doubt, as their share of headaches).

The parallel with your anecdote about asking for the toilet's
whereabouts in Japanese is pretty telling, I think:-).

Of course, it's all a matter of degrees -- of shades of grays,
rather than of blacks and whites only!  So, what else is new?

Nothing "is EXACTLY like" anything else (Korzibsky would
insist on such firm avoidance of "the IS of identity", but, as
he doesn't seem to be following this newsgroup regularly,
I'll have to play his role, I guess).  There are always some
distinctions to draw, as well as, often, some similarities,
parallels, or analogies, to notice and meditate upon.

So, when somebody proposes an analogy which most might
consider inappropriate (and parallels between natural and
programming languages are a frequent case), I try to play
the game -- arguing my corner (if I have one -- else, I
just pick one that seems like it might be fun, or of some
challenge, just like one did back in debating-matches:-),
in part, by embroiding on those parallels.  Why is a raven
like a writing desk?  Well, both might be well modeled in
Python, Poe wrote upon both, and, of course, there IS a
'b' in 'both'.


Seriously, for a sec -- I understand, and, of course, agree,
with most of what you're saying.  In actuality, the closest
real and applicable parallel between programming and
natural languages may be based on what they DO have in
common: the human mind.  Learning a new language, in
either sense of the word, can always be a broadening
experience for that peculiar, quirky, interesting organ --
one different from your current knowledge-set even more
so, for the usual 'triangulation best done from distant
basepoints' reasons.  It's a paradigm issue -- I know
(from your excellent diagnostic-manual on quack theories)
that you have a healthy and well-justified dislike for this
word, but here, for once, it IS used appropriately:-).

*Using* a language is subject to different constraint and
uses than *learning* it, of course... and, most often, it's
what one is debating about *programming* languages.


> Computer languages have more precisely defined semantics
> than natural languages. The niggling details are important,
> such as what "[[0]*10]*3" means. Natural languages are
> more fluid, and it is the nuances of the tacit, multiple
> meaning, and almost ambiguous that are important in their
> competence.

In other words, (natural) languages are NOT formulas, because
human _thoughts_ aren't -- they're biological phenomena, not
mathematical abstractions; the key intuition that distinguishes
the mature works of Wittgenstein from his youthful Tractatus.
Yes.

Go further, though: 'mathematical' languages are *bridges*,
between human minds and thoughts on one side, and abstract
symbol-processing engines/environments/settings (real or not
as the case may be) on the other side.  When using them, it's
not *fully* the case any more that "whatever we're talking
about, we're talking about the natural history of human beings",
to restate the key mature-Wittgenstein's insight in his own
words; but, surely, it's still *partly* the case, in as much as
human minds are involved in creating, shaping, using them.

It's from their bridge-nature, I think, that key characteristics
of 'mathematical languages' follow.


> For these reasons, I am a great believer in small computer
> languages, ones that can be explained well in a book no
> thicker than my pinkie. The use of extentions should be
> recognizable, so that when you come across something you
> don't know, you know *what* to look up and where to find
> it. "One way to do everything" IS better than "many ways
> to do anything."

Right -- *because* there are still human beings involved,
though, *as well* as "theoretically less-quirky" symbolic
processing systems on the other side.  Our mind's ability
to grasp and manipulate symbolic abstraction has limits
(ones which an evolutionary psychologist would have good
'just-so stories' ready to explain, but let's not anger S. J.
Gould needlessly by repeating them, shall we).  Simplicity's
power comes from that.

The complexity of natural language springs in good part from
the _adaptive_ characteristics (pace Gould...) of ambiguity in its
use; words serve to hide, as much as they serve to show.

Ambiguity and misunderstanding play no similarly adaptive
roles when we're interacting with an abstract symbol processing
environment (although, even there, human biological traits
are never fully absent -- using a needlessly complicated and
tricky tool may be a way to 'show off', much like a tail so
long, heavy, and showy, that it makes running away from
predators much harder, may be for a male peacock.  "See
what big brains I have, *I* can handle this horridly complex
language even when there's no real need to, *I* am the alpha
male here!", in other words).  "Hiding", as in "information
hiding", is a tool for further _simplification_ here -- one form
of abstracting away what we don't have to keep in mind right
now, so as to minimize the clutter in our thought.


> Of course, being small has its hazards, if core features
> are omitted. That's why I am fond of saying that a language's
> productivity is proportional to its expressive power, and
> inversely proportional to its complexity. (Not that these are
> the only considerations.) That is why I am happy to see
> Guido resist the temptation to continually add new bells and
> whistles.

*Full* agreement here.  "Perfection is reached, not when
there is no longer anything to add, but when there is no
longer anything to take away", in other words (not mine,
of course; but, still, from yet another one of my favourites
among this century's thinkers:-).

Actually, pragmatic issues _augment_ the relative weight
of language simplicity, so I'd revise your formula to
something closer to expressive-power divided by
complexity _squared_.  A simpler language, as well as
helping in terms of not cluttering its user's mind with
more complexity than absolutely necessary, is also liable
to have simpler *implementations* -- and a simpler
language-processor (compiler, interpreter, whatever),
all other things being equal, will be more robust, easier
to port, faster, more adaptable to new environments.  All
big potential pluses in pragmatic terms.

Alternatively, of course, great complexity in implementation
MIGHT be spent towards such targets as optimization, as
you were interestingly arguing Python should be doing now.
But, even if a strongly-optimizing language processor,
being more fragile, exists more precariously, it IS still good
to have a simple, clear, transparent "reference implementation"
to fall back to...


> People tend to focus on the benefit of expressive power,
> while overlooking the problem of complexity. C, Lisp, and
> Python are examples of languages that pack a lot of
> expressive power into simple packages with a small set of
> orthogonal concepts. C++, trying to build OO on top of C
> while retaining maximal compile time efficiency, grew

..._and_ potential for top-notch *run*-time efficiency, too...

> bigger than its expressive power requires. But that set
> of goals is hard to achieve, and to its credit, C++ does
> this.

Again full agreement, and it's hard to find somebody whose
thoughts on these specific issues match mine as well as yours
do (of course, I'd have taken twice as many words to say the
same things less effectively, but, that's just me:-).  People
who appreciate simplicity for all the right reasons often fail
to grasp that C++ is probably pretty close to the best design
*that matches all of its design-goals* -- which, of course, is
not to say that one can't criticize it!, if nothing else because
its complexity proves how overly ambitious all of those goals
were when pursued at once:-).

> Perl is overly complex for what it does. I think
> this was due to its syncretic history, rather than the
> scope of what it achieves.

Not sure about that.  Its design goals DID apparently include
rich, exhuberant, redundant abundance -- it seems to have
matched them closely.  Again, of course, just like for C++,
one can then reflect on the appropriateness of those design
targets, by reasoning backwards from the design that meets
them (again, I guess, meets them pretty well).


> Like Alex, I can't imagine any
> task that, from a pure language viewpoint, is better done
> in Perl than in Python. Practical constraints, such as
> a company's preferred language, or which language someone
> already knows, could determine otherwise.

Right -- I guess I shouldn't have used "somebody with a
loaded gun at my head" as the _typical_ constraint making
Perl usage pragmatically preferable.  Chalk it up to typical
Latin exhuberance... it *is* a constraint that would be pretty
convincing to me.  Faced with other, less-strict constraints
(easy to find free/cheap web-hosts offering Perl for user
written CGI's -- hard for Python -- just for example), I've
chosen to postpone putting up a CGI-enriched personal
website (actually, I don't really have a personal website
at the moment, as magenta took my old one down and I
haven't found another satisfactory host yet); as for the
preferences of companies, with the kind of job market for
programmers that has prevailed over the last 2-3 years,
I'm not sure how much weight I would give to keeping a
job with a company so benighted as to prefer Perl.  Darn,
here I am, exhuberating yet again -- gotta watch it!-)


Alex






From fredrik at effbot.org  Fri Dec 29 15:36:23 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 29 Dec 2000 20:36:23 GMT
Subject: Bug or Feature?
References: <92ij58$74k3p$1@ID-11957.news.dfncis.de>
Message-ID: <bB636.2481$Qb7.271897@newsb.telia.net>

Emile van Sebille wrote:
> This works:
>
> print eval.__doc__

eval's a function.

>>> eval
<built-in function eval>

>
> but this doesn't:
>
> print exec.__doc__

exec's a statement (and a reserved word), just like print,
def, return,etc.  there's simply no "exec" object in there...

also see:
http://python.sourceforge.net/peps/pep-0233.html

</F>






From s2mdalle at titan.vcu.edu  Sat Dec 16 11:28:53 2000
From: s2mdalle at titan.vcu.edu (David Allen)
Date: Sat, 16 Dec 2000 21:28:53 +0500
Subject: Interrupting a thread...
Message-ID: <91hqgk$pu0$1@bob.news.rcn.net>

I'm writing a program which does some networking,
and I have a thread that is devoted to grabbing 
things off of the network so that the program won't
hang while it is reading data.

The problem is, I also occasionally need to interrupt
this data fetching operation, so I need a way to
either raise an exception in the other thread, or
otherwise signal to it to stop what it's doing.
Since the download thread calls other functions
that are not sensitive to the data in the original
object, I can't check to see if some value has 
changed each time I'm reading in a loop.

Ideally, I'd like thread 1 to raise an exception
that would be propagated in thread 2, not thread 1.
Is this possible?  What other techniques might work?

Any help would be appreciated.
-- 
David Allen
http://opop.nols.com/
----------------------------------------
He played the king as if afraid someone else would play the ace. 
- John Mason Brown, drama critic


From sh at ttsoftware.co.uk  Wed Dec  6 12:07:32 2000
From: sh at ttsoftware.co.uk (Steve Horne)
Date: Wed, 06 Dec 2000 17:07:32 +0000
Subject: Sort
References: <NlnX5.1444$yv1.2700363@nnrp5.proxad.net> <oa8s2tsls5foll62r8fpa6nlcpisb6v1ek@4ax.com> <90liij$805$1@nnrp1.deja.com>
Message-ID: <2ass2t8hiltglh6lmip8323ptv7pati4i6@4ax.com>

On Wed, 06 Dec 2000 14:30:42 GMT, Greg Landrum <glandrum at my-deja.com>
wrote:

>In article <oa8s2tsls5foll62r8fpa6nlcpisb6v1ek at 4ax.com>,
>  Steve Horne <sh at ttsoftware.co.uk> wrote:
>>
>> You can specify an alternative comparison function to the sort method.
>> The easiest way to do this is with a lambda...
>>
>> l.sort (lambda a, b : a[2] < b[2])
>Is this something that changed in python 2.0?  The above example
>does not work under 1.5.2 where, according to the docs:

You are right - my bad - I should have used the cmp function, which
gives the correct -1, 0 or 1 result.

l.sort (lambda a, b : cmp (a[2], b[2]))

The < operator only returns 0 or 1, so the list would only be partly
sorted - though sadly enough to fool me on my quick memory-refreshing
test.

>l.sort (lambda a, b : a[2] - b[2])

Surely this doesn't always work either? Even for integers, you can get
values other that -1, 0 and 1. And what about strings, nested lists,
etc?

-- 
Steve Horne
Home : steve at lurking.demon.co.uk
Work : sh at ttsoftware.co.uk


From jepler at inetnebr.com  Mon Dec 11 16:02:02 2000
From: jepler at inetnebr.com (Jeff Epler)
Date: Mon, 11 Dec 2000 21:02:02 GMT
Subject: Python2 distutils question: how to best "autoconfig"...?
References: <9130e60obc@news2.newsguy.com> <82aZ5.3458$Kd1.395204@newsb.telia.net>
Message-ID: <slrn93ag68.8um.jepler@potty.housenet>

On Mon, 11 Dec 2000 19:37:40 GMT, Fredrik Lundh
 <fredrik at effbot.org> wrote:
>>>> print get_base_and_digits()
>(2.0, 53)
>
>(if anyone runs Python on a box that returns anything
>but (2.0, 53), let me know!)

Well, by setting the value of the FPU status word on my x86, I can get
(2.0, 24), but I doubt this is of any real interest.  (The comments
in the fpu_control.h header file suggest that libm will no longer work
properly when single precision is selected, anyhow)

Jeff


From loewis at informatik.hu-berlin.de  Sat Dec  9 05:36:45 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 09 Dec 2000 11:36:45 +0100
Subject: ANN.: Weak Reference Extension Module has been converted to "C"; Beta 2 is out
References: <8usorp$3i3$1@nnrp1.deja.com> <mtaeb1f4q3.fsf@astron.berkeley.edu> <8uvjmj$dt8$1@nnrp1.deja.com> <mtlmuk9hyc.fsf@astron.berkeley.edu> <B63C5A9E.172D7%graham73@telocity.com> <3A176E0F.39BF350B@felinity.com> <KaNR5.5550$QH2.654863@newsb.telia.net> <8v9dl9$si6$1@nnrp1.deja.com> <90i3e3$f8t$1@nnrp1.deja.com> <p6qu28ek7zg.fsf@informatik.hu-berlin.de> <90s2as$ibo$1@nnrp1.deja.com>
Message-ID: <p6q7l59c24i.fsf@informatik.hu-berlin.de>

shindich at my-deja.com writes:

> > That's indeed a big improvement. Congratulations!
> Do I sense sarcasm in your voice? ;)

Not at all. I always thought it would not have the slightest chance of
getting serious attention as long as it was C++.

As the next step, I think packaging through distutils would be good
(and please put a toplevel directory into the archive).

Martin


From rodsenra at correionet.com.br  Thu Dec 14 20:18:31 2000
From: rodsenra at correionet.com.br (Rodrigo Senra)
Date: Thu, 14 Dec 2000 23:18:31 -0200
Subject: Python for REAL Projects ?
References: <3a380d1e.409238@bendigo.news.origin.net.au> <G5KBE1.8F3@world.std.com>
Message-ID: <3A397167.F07D80E4@correionet.com.br>

" dragon at hitech.net.au wrote:
> : Hello Python People !
> : I have been looking at, and trying out Python (under Win NT) and
> : although Python SEEMS interesting can it "cut the mustad" on a real
> : client project ??

Since you have mentioned Delphi in your e-mail,
have you checked out Python for Delphi ? It is very easy to install
and allows to embeed a Python VM inside a Delphi Program among
other interesting things.

I do not have a link at hand, but you can find it in the Vault of
Parnassus or querying www.google.com.
HTH
Rod

-- 
Rodrigo Senra         
Computer Engineer   (GPr Sistemas Ltda)  rodsenra at correionet.com.br 
MSc Student of Reflection (IC- UNICAMP) Rodrigo.Senra at ic.unicamp.br
http://www.ic.unicamp.br/~921234 (see also   http://www.gpr.com.br)




From lists at TheWebsons.com  Sat Dec 23 15:18:40 2000
From: lists at TheWebsons.com (Ben Ocean)
Date: Sat, 23 Dec 2000 12:18:40 -0800
Subject: Why Won't This Work?
In-Reply-To: <20001223184357.G5515@xs4all.nl>
Message-ID: <LOBBJMHOIFOLFIPFANHJIEGACFAA.lists@TheWebsons.com>

Hmmm. *python* at a command prompt, sans *./* is a command. This worked when
I configured 2.0, which I had to rip out because it's not compatible with
Zope. Also, I tried *make test*

[webmaster at thor Zope-2.2.4-src]$ make test
make: *** No rule to make target `test'.  Stop.

Any other ideas as to why I'm not getting my installation of Python 1.5.2 to
respond?
TIA,
BenO

-----Original Message-----
From: Thomas Wouters [mailto:thomas at xs4all.net]
Sent: Saturday, December 23, 2000 9:44 AM
To: Ben Ocean
Cc: python-list at python.org
Subject: Re: Why Won't This Work?


On Sat, Dec 23, 2000 at 06:31:38AM -0800, Ben Ocean wrote:

> Because I'm building Zope, I'm trying to install Python-1.5.2. I've
> successfully run:
> ./configure --with-thread
> make
> and everything appears to be in order. But when I type
> python
> at a command prompt, I get a bash error. What have I not done?

That's because the current directory ('.') isn't in your PATH. Try typing
'./python' instead. Note that the best way to test your python binary is
'make test' -- it'll catch practically everything you can try to catch
yourself, and a lot more to boot.

--
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me
spread!




From grisha at ispol.com  Mon Dec 18 16:51:13 2000
From: grisha at ispol.com (Gregory Trubetskoy)
Date: Mon, 18 Dec 2000 21:51:13 GMT
Subject: ANNOUNCE: mod_python 2.7.1
Message-ID: <Pine.LNX.4.21.0012181646240.20344-100000@saturn.ispol.com>

What it is:

Mod_python is an Apache module that embeds the Python interpreter within
the server. With mod_python you can write web-based applications in Python
that will run many times faster than traditional CGI and will have access
to advanced features such as ability to retain database connections and
other data between hits and access to Apache internals.

http://www.modpython.org/ for info and download.

What's new in 2.7?

   *  Greatly improved documentation. The documents were remade using the
        Python documentation framework. The documentation is now available
        in HTML and PDF, has a much improved consistent look and an index. 
   *  New PythonHandlerModule directive. 
   *  New mod_python.publisher handler. 
   *  New mod_python.util module. 
   *  Several other less significant bug fixes and improvements. 

--
  Gregory (Grisha) Trubetskoy
       grisha at ispol.com



From root at rainerdeyke.com  Sat Dec  9 18:57:58 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Sat, 09 Dec 2000 23:57:58 GMT
Subject: imputils problem
Message-ID: <aGzY5.6237$x6.4746529@news2.rdc2.tx.home.com>

I've been having a problem with imputils: a module imported through a custom
importer seems unable in import .pyd modules.  In particular, if I import
the zipfile module though a custom importer, it is unable to find the zlib
module.  Does anybody here know how to fix this?


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From scarblac at pino.selwerd.nl  Sun Dec 31 07:17:08 2000
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 31 Dec 2000 12:17:08 GMT
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <slrn94shbe.150.grey@teleute.rpglink.com> <X8s36.2731$Qb7.305339@newsb.telia.net> <slrn94smds.150.grey@teleute.rpglink.com> <FOt36.17$vd1.2224@e420r-atl2.usenetserver.com> <slrn94t1vj.150.grey@teleute.rpglink.com> <3A4E95FD.D6C737A7@engcorp.com> <slrn94tqkh.150.grey@teleute.rpglink.com>
Message-ID: <slrn94u8o3.m8h.scarblac@pino.selwerd.nl>

Steve Lamb <grey at despair.rpglink.com> wrote in comp.lang.python:
> On Sat, 30 Dec 2000 21:12:13 -0500, Peter Hansen <peter at engcorp.com> wrote:
> >I understand the point you are making; I'm just pointing out that the
> >opposing views seem well-founded.  Areas where I think you missed the
> >point are in the importance of code maintainability, the increased
> >complexity with your suggested approach, and the feelings of the "entire
> >CS community" on this issue.
> 
>     I seriously doubt that initing a variable before use causes either code
> maintainability problems or increased complexity.

It can hide bugs. If you use an unitialized variable, Python complains
loudly, pointing out some problem with your code, maybe a control flow
bug or a misspelled name or so. Initializing it beforehand makes those
bugs harder to find.

-- 
Remco Gerlich



From gilh at sympatico.ca  Wed Dec 27 09:56:07 2000
From: gilh at sympatico.ca (Gil Hauer)
Date: Wed, 27 Dec 2000 09:56:07 -0500
Subject: interpretter and available memory
Message-ID: <92d01r$1vut$1@news.tht.net>

Hi,

Is there a mechanism for determining how much memory is being consumed by
the interpretter (and how much is left)? I need this functionality in both
1.5.x and 2.0.

    Thanks,
    Gil




From sholden at holdenweb.com  Wed Dec 13 18:55:23 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 13 Dec 2000 18:55:23 -0500
Subject: Use of httplib?
References: <3A37E456.772811C9@nospammy_raqia.com>
Message-ID: <x2UZ5.875$4y4.33024@e420r-atl1.usenetserver.com>

David Lees <DavidL at nospammy_raqia.com> wrote in message
news:3A37E456.772811C9 at nospammy_raqia.com...
> I am running the client code below to try and send and receive text from
> a server.  My problem is that I do not know how to recieve back an
> arbitrary number of lines.  In a test case, when I know the exact number
> of lines as shown below (e.g. 11) I can use readline.  However, in
> general I do not know how many lines there are and the program hangs on
> the connection when  I accidentaly read to the end.  The comment out
> pieces of code also either hang or do not give back all the text.
>
Hmm ... the manual page for httplib reminds us that it is used by urllib,
and although it does implement the HTTP client-side protocol it's actually
much easier to use urllib.  Here's a snippet form a crawler program I wrote
some time ago (warning: not tested under 2.0) which shows that it's very
easy to use urllib to read HTML pages from a server:

    ignored = []
    distance = 1
    while distance < MAXDIST:
    l.append([])
    for URL in l[distance-1]:
        try:
        if URL == '-':
            f = sys.stdin
        else:
            f = urllib.urlopen(URL)
        data = f.read()
        if f is not sys.stdin:
            f.close()

        p = myHTMLParser(fmt, URL)
        p.feed(data)
        links, rubbish = p.close()
        known[URL].setgood()


> I am copying from the Library Reference section 11.4, but it does not
> seem to document the readline, read or readlines methods for http.
>
Bottom line: use urllib, that's what it's for.  Easier, too!  Good luck.

regards
 Steve





From slhath at home.com  Tue Dec 19 19:54:50 2000
From: slhath at home.com (Scott Hathaway)
Date: Wed, 20 Dec 2000 00:54:50 GMT
Subject: xml-rpc module question
References: <dmC%5.75166$yR4.2151642@news1.rdc1.tx.home.com> <j7L%5.879$VF3.1528@news.more.net> <%3T%5.76381$yR4.2212181@news1.rdc1.tx.home.com> <V9T%5.651$Qb7.44185@newsb.telia.net> <KhT%5.652$Qb7.44208@newsb.telia.net>
Message-ID: <urT%5.76389$yR4.2213032@news1.rdc1.tx.home.com>

Fredrik,

I tried your code and I get the following error:  (BTW - I am obviously a
newbie to python :))

C:\Projects\core>python core_client.py
Traceback (most recent call last):
  File "core_client.py", line 5, in ?
    print testing.CubeIt(3)
  File "xmlrpclib.py", line 547, in __call__
    return self.__send(self.__name, args)
  File "xmlrpclib.py", line 627, in __request
    response = self.__transport.request(
AttributeError: 'int' object has no attribute 'request'


Thanks for the help and thanks for writing the xml-rpc library!

Scott




"Fredrik Lundh" <fredrik at effbot.org> wrote in message
news:KhT%5.652$Qb7.44208 at newsb.telia.net...
> I wrote:
> >     def call(self, method, params):
> >         if method == 'CubeIt':
> >             return self.CubeIt(params)
>
> correction: params is a tuple.  tuples cannot be cubed.
>
> better make that:
>
>              return self.CubeIt(params[0])
>
> or, better:
>
>              return apply(self.CubeIt, params)
>
> (anybody know a good book on debugging with print
> statements? ;-)
>
> </F>
>
>




From christian at integration-x.com  Wed Dec  6 09:19:45 2000
From: christian at integration-x.com (Christian Wermuth)
Date: Wed, 06 Dec 2000 15:19:45 +0100
Subject: Accessing "variables" in an extension module
Message-ID: <3A2E4B01.4609EE7E@integration-x.com>

I would like to access "variables" (chars and ints) in a C module
extension for Python from my Python code. Is that possible?
If I wanted to improve the example from "Extending and Embedding the
Python Interpreter" like shown below, what would I need to add to the
spam.c code?

import spam
spam.varstring = "test"
spam.varint = 5
spam.system('dir c:')

Thanks in advance,
Christian Wermuth




From jschmitt at vmlabs.com  Wed Dec 13 15:22:02 2000
From: jschmitt at vmlabs.com (jschmitt at vmlabs.com)
Date: Wed, 13 Dec 2000 20:22:02 GMT
Subject: launching a python program from GNU make (Win2K)
Message-ID: <918lp0$6q4$1@nnrp1.deja.com>

>From my experience, GNU make has the most features that I want to use.
However, I find it so hard to use under Windows.  For example, I wrote
a small python program that I can launch from the command line and
execute without problems.  I can't seem to launch it from GNU make.  I
have the python program's directory in my PATH and also in my
PYTHONPATH and I added .py to my PATHEXT list.

What could the reason be and how could I fix it?  Does it have
something to do with GNU make going out of it's way to launch sh.exe if
it's on your hard drive, rather than launching command.exe to shell out
processes?  If that is the case, how do I get sh.exe to recognize my
python programs?  Is it the '#!python' thing that the Unix people do?

John

PS Choosing a different make is not an option here.


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


From fredrik at effbot.org  Tue Dec 12 12:37:58 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 12 Dec 2000 17:37:58 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu>
Message-ID: <WntZ5.3566$Kd1.414873@newsb.telia.net>

Roy Katz wrote:
> > The real problem is that many Python tutorials fail to explain
> > how things work (or hide it somewhere in the exercises), but
> > that's not really Guido's fault...
>
> Tough for the new Python programmer, then.  It's nt fair to them.

Sorry, but I don't understand that comment.  I've done a lot
of Python training lately, and nobody has had any problems
whatsoever understanding this.

(well, Lisp and Smalltalk programmers seem to have trouble
understanding why anyone would do things in any other way...)

> > (Hint: Everything is a reference to an object.  Variables are
> > named references, not actual objects.  Some objects can be
> > modified in place.  Some cannot be modified.  That's all)
>
> Okay! tell me, which are the objects which cannot be modified in place? I
> assume these are int's, long's, float's, etc.  Strings, lists, objects and
> dictionaries are call-by-reference.  Pardon my choice of diction, but I am
> against inventing new terminiologies (modified-in-place??).

Sigh.  The problem is that you keep using terms that mean
something for you, whether or not they apply to the current
situation.

If you want to learn something new, you have to stop doing
that.

Reset your brain.

Python doesn't use call-by-value, and it doesn't use call-by-
reference either (at least not given your definitions of those
terms).

Realize that there are more than one way to do it.

Here's how it's done in Python:

:::

All Python objects has the following:

    -- some content
    -- an unique identity (a integer, returned by id(x))
    -- a type (returned by type(x))

You cannot change the identity.

You cannot change the type.

Some objects allow you to change the *content* (without
changing the identity or the type, that is).

Some objects don't allow you to change the content
(more below).

The type is represented by a type object, which knows
more about objects of this type (how many bytes of
memory they usually occupy, what methods they have,
etc).

:::

Objects may also have:

    -- zero or more methods (provided by the type object)
    -- zero or more names

Some objects have methods that allow you to change the
contents of the object (modify it in place, that is).

Some objects only have methods that allow you to access
the contents, not change it.

Some objects don't have any methods at all.

Even if they have methods, you can never change the type,
nor the identity.

Things like attribute assignment and item references are just
syntactic sugar (more below).

:::

The names are a bit different -- they're not really properties
of the object, and the object itself doesn't know what it's
called.

An object can have any number of names, or no name at all.

Names live in namespaces (such as a module namespace, an
instance namespace, a function's local namespace).

Namespaces are collections of (name, object reference) pairs.
(Or in other words, they're dictionaries).

When you call a function or a method, it's namespace is
initialized with the arguments you call it with (the names
are taken from the function's argument list, the objects
are those you pass in).

:::

Assignment modify namespaces, not objects.

In other words,

    name = 10

means that you're adding (or replacing) the name "name" in
your local namespace, and making it refer to an integer object
containing the value 10.

    name = 10
    name = 20

means that you're first adding the name "name" to the local
namespace, and making it refer to an integer object containing
the value 10.  You're then replacing the name, making it point
to an integer object containing the value 20.  The original "10"
object isn't affected by this operation, and it doesn't care.

In contrast, if you do:

    name = []
    name.append(1)

you're first adding the name "name" to the local namespace,
making it refer to an empty list object.  You're then calling a
method on that object, telling it to add an integer object to
itself.

Things like "name.attr" and "name[index]" are just syntactic
sugar.  The first corresponds to __setattr__/__getattr__, the
second to __setitem__/__getitem__, depending on which side
of the assignment operator they appear).

That's all.

:::

> I want to know where this is documented.  Specifically, where is the
> dichotomy of ints/longs/floats vs. objects/strings/lists/dictionaries.

There is no dichotomy.  Some objects have methods that let
you modify them, some don't.  Read the language reference.
If you stop thinking in C terms, you may understand what it's
talking about.

> While I'm brainstorming with Python grammar changes, what about a real &
> operator.  No more of this hokey is-it-modifiable-in-place-or-not
> business.  Explicit control is important.

As I just explained, Python doesn't work that way.

</F>






From kampfwurst at kampfesser.net  Wed Dec 20 17:13:08 2000
From: kampfwurst at kampfesser.net (Die Kampfwurst)
Date: Wed, 20 Dec 2000 23:13:08 +0100
Subject: test, please ignore
Message-ID: <3A412EF4.4040903@kampfesser.net>

Just a test. Generated with Mozilla CVS nightly snapshot 17.12.2000.



From removethis_kranki2 at earthlink.net  Fri Dec 22 10:34:11 2000
From: removethis_kranki2 at earthlink.net (Bob W.)
Date: Fri, 22 Dec 2000 15:34:11 GMT
Subject: Newbie Question
References: <05m64tkqev0aavea80ttjse1sj2gd447hg@4ax.com> <91vomv$5i9dg$1@ID-11957.news.dfncis.de>
Message-ID: <4js64tofqr760kgv12ctsjuii3lige6jj4@4ax.com>

Well, I had tried using the "self" reference and that didn't seem to
work either.  "self" would be fine since I had not planned on using
the method outside of the class.  Actually, I was trying to simulate a
private method in C++.  Thanks for your help though.

On Fri, 22 Dec 2000 06:33:36 -0800, "Emile van Sebille"
<emile at fenx.com> wrote:

>Do you intend to be invoking the unbound class method?  Or
>should
>
>os.path.walk( self.szDir, RmwExpireCheck.__CheckFile, self )
>
>be
>
>os.path.walk( self.szDir, self.__CheckFile, self )



From DanielK at aracnet.com  Wed Dec 27 10:34:22 2000
From: DanielK at aracnet.com (Daniel Klein)
Date: Wed, 27 Dec 2000 07:34:22 -0800
Subject: How to question
References: <mailman.977863690.15082.python-list@python.org>
Message-ID: <qUn26.317$LU6.136439@typhoon.aracnet.com>

Don't know about PythonWin but in IDLE you can:

    def cls():
        print "\n" * 50

Daniel Klein

(btw, you should not be sending newsgroup messages in html since there are
many newsreaders that can only 'see' plain text)

"Daniel Kinnaer" <Daniel.Kinnaer at AdValvas.be> wrote in message
news:mailman.977863690.15082.python-list at python.org...
How can one 'clear' or 'cls' the interactive window from PythonWin or form
IDLE? That is : I'd like to clear/cls this window so that all previous
output in the interactive window is erased and only the last 'run' is shown
in the interactive window...  Thanks

best regards,
Daniel




From aleaxit at yahoo.com  Tue Dec 26 04:03:41 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 26 Dec 2000 10:03:41 +0100
Subject: One Python 2.1 idea
References: <mailman.977771949.17911.python-list@python.org> <9292tn$jka$1@nnrp1.deja.com>
Message-ID: <929n9c013m3@news1.newsguy.com>

<rturpin at my-deja.com> wrote in message news:9292tn$jka$1 at nnrp1.deja.com...
    [snip]
> Second, a question. Consider the invocation of a method,
> such as:
>
>     x.foo()
>
> Does this really cause the lookup of the string "foo"
> in x's namespace dictionary? The first time? Every time?

Yes, and yes.  No implicit caching.

> Is this also the case when data members are accessed?

Yes once more.


Alex





From Randy.L.Kemp at motorola.com  Thu Dec 28 09:01:09 2000
From: Randy.L.Kemp at motorola.com (Kemp Randy-W18971)
Date: Thu, 28 Dec 2000 08:01:09 -0600
Subject: Python vs. PHP (& Java?)
Message-ID: <69BC4C6AE83ED311BC9400805FA7B14AF84D2C@il93exp01.css.mot.com>

  Can we nip language comparisons in the bud?  In reality, if someone claims one language is better then another, I would ask to see the studies to back the claim.  In reality, there are really several good languages and each has its own camp.  If a person is really good in a language, they could probably get the language to solve most programming problems.  While I do have my own biases (Java and Python), I recognize and salute the benefits of Perl, PHP, JavaScript, etc.  Suppose we take the arguments for computer languages and extended them to human languages.
1.  Could we argue that English is a better language to use around the world then French or Spanish?
2.  Could we argue that an artificial language, such as Esperanto, is better to use around the world then a natural language, like English?
  Isn't there an axiom in physics that an experimenter influences the outcome of his experiment (physicists, help me out here)?  
    

-----Original Message-----
From: Alex Martelli [mailto:aleaxit at yahoo.com]
Sent: Thursday, December 28, 2000 4:11 AM
To: python-list at python.org
Subject: Re: Python vs PHP (& Java?)


"Tim Hammerquist" <tim at degree.ath.cx> wrote in message
news:slrn94l86m.a3h.tim at degree.ath.cx...
> Ben Ocean <zope at TheWebsons.com> wrote:
> > I'm an intermediate-level programmer that's new to Python (via Zope) and
I
> > like it a lot. However, it appears to me that, while Python is a good
> > replacement for Perl, it cannot replace much of the functionality of
PHP,
> > which can be scripted into HTML (whereas Python, it appears, must be
> > called).
>
> Ok, everyone else made very good points, but this little comment slipped
> out (not that I'm surprised, honestly)  =)   There is no "replacement
> for Perl!"  And Python would do a poor job trying to be Perl (and vice
> versa), nor does anyone here seem to _want_ Python to be/replace Perl.

Not sure what you mean here.  Over my first 2-3 months of Python,
I rewrote _every_ Perl script on my machine in Python, and the gains
in clarity, reliability, and maintainability, were outstanding.

The 'good-applicability niches' of Perl and Python are vastly
overlapped (and not so small as to really deserved being called
'niches', but that's another issue).  I can't think of any program
I would rather write in Perl than in Python, though I guess I
might *have* to, if forced (by somebody pointing a loaded gun at
my head with its safety off, for example).  So, for me, Python HAS
indeed served as "a good replacement for Perl" -- everything I used
to do in Perl, I now do in Python instead; what other definition
of "replacement" are you using instead?


Alex






From nospam at caltech.edu  Fri Dec  8 03:53:21 2000
From: nospam at caltech.edu (Nathaniel Gray)
Date: Fri, 08 Dec 2000 00:53:21 -0800
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu> <976153091.257544@clam-ext>
Message-ID: <3A30A181.4302F102@caltech.edu>

Mark Hadfield wrote:
> 
> "Nathan Gray" <nospam at caltech.edu> wrote in message
> news:3A2EE2AF.AD8A51E2 at caltech.edu...
> > What a rude response to a polite and legitimate question.
> 
> Well I think a more legitimate question would have been:
> 
> "I want to know about Pythons origins. I've looked in the obvious places,
> i.e. www.python.org, and read the stuff there but I want to know more about
> this: ..."

Granted, this would have been a better approach.  I'm not saying it was
the
best question ever, but I have no qualms with calling it legitimate.

It sounded legitimate to me since he was asking about the history of the
language.  This isn't necessarily something a newbie would assume was
on www.python.org. It seems to me that as a community we should be 
encouraging people to read our history, not jumping on them for asking 
about it in the wrong way.

I guess we've all got our own standards for what's a legitimate question
and
what's a waste of our time -- that's fine by me.  I just don't like it
when I
see an obvious newbie flamed for asking a relatively innocuous (if
inane) question.
It's not good for the atmosphere in c.l.p. when somebody who's clearly
never
been here before gets clobbered for something this trivial.

-n8
n8gray <at> caltech <dot> edu
-- 
                          Nathaniel Gray
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
                   California Institute of Technology                    
                     Computation and Neural Systems                      
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._


From dnew at san.rr.com  Mon Dec 18 12:41:40 2000
From: dnew at san.rr.com (Darren New)
Date: Mon, 18 Dec 2000 17:41:40 GMT
Subject: while true: !!!
References: <B43D149A9AB2D411971300B0D03D7E8B0FBCBB@natasha.auslabs.avaya.com> <v04220801b663769e1800@[192.168.11.33]> <mailman.977147642.18708.python-list@python.org> <91lclt02qr3@news1.newsguy.com>
Message-ID: <3A3E4C55.9C785306@san.rr.com>

> > concept that an integer can denote true or false.  How many functions
> > return 0 for success?  (just take a look at the standard C library).

I think that's because there's generally one way to succeed and many ways to
fail.

> Few languages equate 'success' (not raising an exception, for
> languages where 'exception' means something) with 'a boolean
> result' (Icon is rather oriented that way, but that's because
> it _backtracks_ over 'insuccess' -- a very different setup
> than Python's). 

Actually, it's not *quite* like that. "if" in Icon is true for *any* result.
It's only the "no result" that invokes the else part. A failure/exception in
Icon is more like partial functions in math, where it's possible there are
inputs which generate no results at all.  One of the cleaner ways of
handling it I've seen.

-- 
Darren New / Senior MTS & Free Radical / Invisible Worlds Inc.
San Diego, CA, USA (PST).  Cryptokeys on demand.
Personal malapropism generator free with purchase!
Steganography: The manual is hidden in the source code.
Gold mining is only profitable because of the gold found.


From mwh21 at cam.ac.uk  Sun Dec 10 06:33:28 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 10 Dec 2000 11:33:28 +0000
Subject: Bug: Readline getting stuck on Linux and Solaris
References: <lxk8zqwupy2.fsf@ioasun3.epfl.ch> <90m53k$ovh$1@nnrp1.deja.com> <m3u28h48ja.fsf@atrus.jesus.cam.ac.uk> <slrn92u8al.76b.quinn@lira.ugcs.caltech.edu> <m3r93k4kqk.fsf@atrus.jesus.cam.ac.uk> <slrn932eoj.d1j.quinn@ngwee.ugcs.caltech.edu>
Message-ID: <m34s0c4ik7.fsf@atrus.jesus.cam.ac.uk>

quinn at ngwee.ugcs.caltech.edu (Quinn Dunkan) writes:

> On 07 Dec 2000 09:57:23 +0000, Michael Hudson <mwh21 at cam.ac.uk> wrote:
> >quinn at lira.ugcs.caltech.edu (Quinn Dunkan) writes:
> >
> >> Strange.. readline has *never* worked right for me, on IRIX or
> >> Linux, on any version of python (1.4, 1.5.2, 2.0), and any version
> >> of readline (3.0, 4.0, 4.1), if python is compiled with threads.
> >> IRIX (6.4, 6.5, 6.5.2) hangs, Linux (2.0.36, 2.2.12, 2.2.17) exits.
> 
> Well, blow me down and pick me up, I just tried python 2.0 on linux (it was
> 2.0b1 before) and the problem is solved, even with threads.

Ah yes; it was a pretty late fix.

> >Well, I have readline 2.2.1, which may make a difference.  Can you
> >supply a stacktrace?
> >
> >> I just tried your 'while 1: pass' thing, and the behaviour is the same there
> >> too.
> >
> >That really is odd, because readline doesn't have control then and
> >should be out of the picture.  Does control-C work as expected here if
> >readline isn't enabled, or if you hit it before the first prompt
> >appears (this may be tricky).
> 
> It's not tricky on an Indy :) In any case, on the irix machine, I've
> been using 2.0-for-real and it still locks.  It apparently works ok
> if I hit control C quickly: I get 'import site' failed, presumably
> because it died of a KeyboardInterrupt.  However, even a 'while 1:
> pass' ^C locks.  I got a backtrace by suspending the locked python,
> and killng it with a SIGILL to get a core.

Well, this sounds like a readline bug to me (or maybe an IRIX oddity
that readline fails to work it's way around).

Do other proggies that use readline work properly?

> (gdb) bt
> #0  0xfa3ab88 in __ioctl () at lio.c:201
> #1  0xfa434f0 in ioctl () at nftw.c:184
> #2  0x5ffd2a74 in rl_restart_output ()
> #3  0x5ffdd0f4 in _rl_clean_up_for_exit ()
> #4  0x5ffdd860 in rl_cleanup_after_signal ()
> #5  0x5ffdd304 in _rl_erase_entire_line ()

Was this backtrace from the "while 1: pass" hang?  If it was, it's
mightily strange that we seem to be in readline code.

> Once again, Irix has no problems if I compile --without-threads
> (which "fixed" the linux problem too, so I assumed they were the
> same bug).  This is Irix 6.5, with gcc 2.8.1 and realine 4.  On
> Monday I'll try with readline 2 and see if it's still there.

I suspect this will make no difference - but it's probably worth
checking.

> ^C works properly in all instances if readline is disabled (via 'mv
> readline.so spam', if there's a more graceful way I don't know it).

No, that's probably the best way.

All I can think of is that there's some fuck-up in restoring signals
(maybe a sysV vs. BSD misunderstanding?).  On catching a signal,
readline does some clean-up, restores the application's original
signal handler and re-sends the signal.  So if for some reason
readline is unable to restore Python's handler, one could see how it
might just spin.  But this is very rapidly getting beyond my
experience - I don't think I've ever even seen an IRIX box...

Ooh, I wonder if readline uses signal(2) rather than sigaction to
reset the the signal - that would probably go wrong.  If that is the
problem, I can't really see a way around it, other than to hack
readline into using sigaction...  do the IRIX manpages say anything
scary about signal/sigaction interaction?

disconnected-ly y'rs
M.

-- 
48. The best book on programming for the layman is "Alice in
    Wonderland";  but  that's because it's  the best book on 
    anything for the layman.
  -- Alan Perlis, http://www.cs.yale.edu/homes/perlis-alan/quotes.html


From hoopyfrood at antionline.org  Thu Dec 28 14:09:46 2000
From: hoopyfrood at antionline.org (R. A.)
Date: Thu, 28 Dec 2000 19:09:46 GMT
Subject: Announce: Mississippi Python Interest Group
Message-ID: <3A4B9156.458DA8B9@antionline.org>

MS PIGgie (MS Python Interest Group) is a new group for people
interested in Python.  The new website and email list are intended to
develop into useful resources, and membership is loosely defined.

Mississippi Needs Elegant Code
Rob Andrews
-- 
Mississippi Python Interest Group
http://www.lowerstandard.com/python/mspiggie


From see_signature at 127.0.0.1  Thu Dec 14 17:22:44 2000
From: see_signature at 127.0.0.1 (Leo Razoumov)
Date: 14 Dec 2000 14:22:44 -0800
Subject: Bug: Readline getting stuck on Linux and Solaris
References: <lxk8zqwupy2.fsf@ioasun3.epfl.ch> <90m53k$ovh$1@nnrp1.deja.com> <m3u28h48ja.fsf@atrus.jesus.cam.ac.uk> <slrn92u8al.76b.quinn@lira.ugcs.caltech.edu> <m3r93k4kqk.fsf@atrus.jesus.cam.ac.uk> <slrn932eoj.d1j.quinn@ngwee.ugcs.caltech.edu> <m3aea12o2z.fsf@atrus.jesus.cam.ac.uk> <slrn93fsoc.5pr.quinn@dinar.ugcs.caltech.edu>
Message-ID: <ja6snnqk5hn.fsf@balu.qualcomm.com>

I recently compiled Python 2.0 for SUN Solaris 2.6 (SunOS 5.6)

Indeed, I had bad luck with readline-4.1 (it linked but some Python tests dumped
core). 

However it all went fine when I linked python against readline-4.0 (as found in
gdb-5.0 distro) and replaced -ltermcap with -ltermlib.

Everything works just fine now!

Just my two cents,
--Leo [slonika at yahoo.com]



>>>>> "QD" == Quinn Dunkan <quinn at dinar.ugcs.caltech.edu> writes:

    QD> On 12 Dec 2000 23:53:56 +0000, Michael Hudson <mwh21 at cam.ac.uk> wrote:
    >> quinn at ngwee.ugcs.caltech.edu (Quinn Dunkan) writes:
    >>> (gdb) bt
    >>> #0  0xfa3ab88 in __ioctl () at lio.c:201
    >>> #1  0xfa434f0 in ioctl () at nftw.c:184
    >>> #2  0x5ffd2a74 in rl_restart_output ()
    >>> #3  0x5ffdd0f4 in _rl_clean_up_for_exit ()
    >>> #4  0x5ffdd860 in rl_cleanup_after_signal ()
    >>> #5  0x5ffdd304 in _rl_erase_entire_line ()
    >> 
    >> Can I see some more of this, please?

    QD> Here are some more details tracebacks (|fmt'ed to shut up slrn):

    QD> (dbx) where
    >> 0 __ioctl(0x0, 0x54d7, 0x5fff9d68, 0x0, 0xc, 0xc, 0x7fe35d90, 0x0)
    QD> ["/xlv51/6.5.5f/work/irix/lib/libc/libc_n32_M3/sys/ioctl.s":20, 0xfaf7dbc] 1
    QD> _ioctl(0x0, 0x54d7, 0x5fff9d68, 0x0, 0xc, 0xc, 0x7fe35d90, 0x0)
    QD> ["/xlv51/6.5.5f/work/irix/lib/libc/libc_n32_M3/sys/ioctlSCI.c":28, 0xfb00720]
    QD>    2 ___new_tcsetattr(0x0, 0x54d7, 0x5fff9d68, 0x0, 0xc3ac600, 0xc,
    QD> 0x7fe35d90, 0x0)
    QD> ["/xlv51/6.5.5f/work/irix/lib/libc/libc_n32_M3/term/new_tcsetattr.c":61,
    QD> 0xfb03878]
    QD>    3 tcsetattr(__fd = 0, __act = 21519, __t = 0x5fff9d68)
    QD> ["/usr/include/sys/termios.h":136, 0x5ffc62f4]
    QD>    4 set_tty_settings(tty = 0, tiop = 0x5fff9d68)
    QD> ["/usr/local/src/readline-4.0/rltty.c":428, 0x5ffc6544]
    QD>    5 rl_deprep_terminal() ["/usr/local/src/readline-4.0/rltty.c":581,
    QD> 0x5ffc6954]
    QD>    6 rl_cleanup_after_signal() ["/usr/local/src/readline-4.0/signals.c":361,
    QD> 0x5ffd655c]
    QD>    7 rl_signal_handler(sig = 2) ["/usr/local/src/readline-4.0/signals.c":157,
    QD> 0x5ffd5f84]
    QD>    8 sig_fixup_mask(0x0, 0x54d7, 0x5fff9d68, 0x0, 0xc, 0xc, 0x7fe35d90, 0x0)
    QD> ["/xlv51/6.5.5f/work/eoe/lib/libpthread/libpthread_n32_M3/sig.c":446,
    QD> 0xc38ebb0]
    QD>    9 _sigtramp(0x0, 0x54d7, 0x0, 0x0, 0xc, 0xc, 0x7fe35d90, 0x0)
    QD> ["/xlv51/6.5.5f/work/irix/lib/libc/libc_n32_M3/signal/sigtramp.s":71,
    QD> 0xfac93c0]
    QD>    10 _ksigaction(0x2, 0x7fe362f8, 0x5fff9ea0, 0x0, 0x11e1000, 0x7fe362f8,
    QD> 0x7fe363c0, 0xc3a4cb8)
    QD> ["/xlv51/6.5.5f/work/irix/lib/libc/libc_n32_M3/signal/ksigaction.s":23,
    QD> 0xfac92ec]
    QD>    11 _SGIPT_libc_sigaction(0x2, 0x7fe362f8, 0x5fff9ea0, 0x5ffd5f20,
    QD> 0x11e1000, 0x7fe362f8, 0x7fe363c0, 0xc3a4cb8)
    QD> ["/xlv51/6.5.5f/work/eoe/lib/libpthread/libpthread_n32_M3/sig.c":419,
    QD> 0xc38eab8]
    QD>    12 ptctl(0x2, 0x7fe362f8, 0x5fff9ea0, 0x0, 0x11e1000, 0x7fe362f8,
    QD> 0x7fe363c0, 0xc3a4cb8)
    QD> ["/xlv51/6.5.5f/work/eoe/lib/libpthread/libpthread_n32_M3/libcthread.c":149,
    QD> 0xc387cd4]
    QD>    13 _sigaction(0x2, 0x7fe362f8, 0x5fff9ea0, 0x0, 0x11e1000, 0x7fe362f8,
    QD> 0x7fe363c0, 0xc3a4cb8)
    QD> ["/xlv51/6.5.5f/work/irix/lib/libc/libc_n32_M3/signal/sigaction.c":26,
    QD> 0xfac98b4]
    QD>    14 rl_set_sighandler(sig = 2, handler = 0x5ffd5f20, ohandler = 0x5fff9ea0)
    QD> ["/usr/local/src/readline-4.0/signals.c":242, 0x5ffd60f4]
    QD>    15 rl_maybe_set_sighandler(sig = 2, handler = 0x5ffd5f20, ohandler =
    QD> 0x5fff9ea0) ["/usr/local/src/readline-4.0/signals.c":259, 0x5ffd6158]
    QD>    16 rl_set_signals() ["/usr/local/src/readline-4.0/signals.c":272,
    QD> 0x5ffd61f0]
    QD>    17 rl_reset_after_signal() ["/usr/local/src/readline-4.0/signals.c":371,
    QD> 0x5ffd65c4]
    QD>    18 rl_signal_handler(sig = 2) ["/usr/local/src/readline-4.0/signals.c":179,
    QD> 0x5ffd5ff8]
    QD>    19 sig_fixup_mask(0x2, 0x7fe362f8, 0x5fff9ea0, 0x0, 0x11e1000, 0x7fe362f8,
    QD> 0x7fe363c0, 0xc3a4cb8)
    QD> ["/xlv51/6.5.5f/work/eoe/lib/libpthread/libpthread_n32_M3/sig.c":446,
    QD> 0xc38ebb0]
    QD>    20 _sigtramp(0x0, 0x7fe362f8, 0x0, 0x0, 0x11e1000, 0x7fe362f8, 0x7fe363c0,
    QD> 0xc3a4cb8)
    QD> ["/xlv51/6.5.5f/work/irix/lib/libc/libc_n32_M3/signal/sigtramp.s":71,
    QD> 0xfac93c0]
    QD>    21 _ksigaction(0x2, 0x7fe36918, 0x5fff9ea0, 0x0, 0x11e1000, 0x7fe36918,
    QD> 0x7fe369e0, 0xc3a4cb8)
    QD> ["/xlv51/6.5.5f/work/irix/lib/libc/libc_n32_M3/signal/ksigaction.s":23,
    QD> 0xfac92ec]
    QD>    22 _SGIPT_libc_sigaction(0x2, 0x7fe36918, 0x5fff9ea0, 0x5ffd5f20,
    QD> 0x11e1000, 0x7fe36918, 0x7fe369e0, 0xc3a4cb8)
    QD> ["/xlv51/6.5.5f/work/eoe/lib/libpthread/libpthread_n32_M3/sig.c":419,
    QD> 0xc38eab8]

    QD> [ ... 11--21 loop forever ... ]

    QD> I get this trace when I do a ^C on an Indigo2 (IRIX64 6.5), even if I do
    QD> the ^C after a 'while 1: pass'.  

    QD> Another Indigo2 with the same stats gives:
    QD> (dbx) where
    >> 0 _getpid(0x1, 0x7fcd30b0, 0x0, 0x0, 0x0, 0x1, 0x0, 0xc374f0c)
    QD> ["/xlv47/6.5.8m/work/irix/lib/libc/libc_n32_M4/proc/getpid.s":15, 0xfabb9f8]
    QD>    1 <Unknown>() [< unknown >, 0x5ffdd33c]

    QD> when it locks after a while 1 ^C, and

    >> 0 __ioctl(0x0, 0x54d5, 0x7fabcb40, 0x0, 0x0, 0x0, 0x0, 0x0)
    QD> ["/xlv47/6.5.8m/work/irix/lib/libc/libc_n32_M4/sys/ioctl.s":20, 0xfafebf8]
    QD>    1 _ioctl(0x0, 0x54d5, 0x7fabcb40, 0x0, 0x0, 0x0, 0x0, 0x0)
    QD> ["/xlv47/6.5.8m/work/irix/lib/libc/libc_n32_M4/sys/ioctlSCI.c":28, 0xfb0789c]
    QD>    2 ___new_tcgetattr(0x0, 0x54d5, 0x7fabcb40, 0x0, 0x0, 0x0, 0x0, 0x0)
    QD> ["/xlv47/6.5.8m/work/irix/lib/libc/libc_n32_M4/term/new_tcgetattr.c":26,
    QD> 0xfb0aa60]
    QD>    3 <Unknown>() [< unknown >, 0x5ffd22c8]

    QD> (varies, but always begins with the ioctls) after a ^C after the prompt first
    QD> appears.

    QD> Ok, on an Indy1 (plain IRIX 6.5), I get the same sort of thing.  Which is
    QD> weird, because I coulda sworn when I tried it before, 'while 1: pass' ^C
    QD> worked right on the indy but not on the indigo.

    QD> Oh, and I used dbx because gdb isn't on all the machines, and it seems to give
    QD> more info anyway (maybe gdb doesn't agree with sgi cc compiled libs).  The
    QD> second indigo2 and the indy compiled python and readline with gcc, while the
    QD> first indigo2 uses sgi cc, so perhaps that accounts for the more complete,
    QD> consistent, and less confusing info from the first one. All of them lock up on
    QD> ^C, even under 'while 1: pass'.  They all work correctly if readline is not
    QD> imported, or if I manage to hit ^C before readline is imported.  Two of them
    QD> work fine when threads aren't compiled in, and I assume the third one would
    QD> too, but I haven't tested it.

    QD> I also grepped through the preprocessed source of readline, and found
    QD> sigactions() but no signals() (well, except in the included protos).  I also
    QD> tried readline-2.2.1; no luck.


    QD> But that first traceback is nice, because it manages to implicate signals,
    QD> readline, and libpthread all at once :)

    QD> ... just what we need!  Two compilers, two ABIs, 32 bit and 64 bit arches,
    QD> multiple library versions, and inconsistent errors!  Irix is such fun.

    >> -- 
    >> In general, I'd recommend injecting LSD directly into your temples,
    >> Syd-Barret-style, before mucking with Motif's resource framework.
    >> The former has far lower odds of leading directly to terminal
    >> insanity.                                            -- Dan Martinez

    QD> Y'know, those random quote sigs can be eerily apropos.  "terminal" insanity
    QD> indeed.


From donod at home.com  Wed Dec  6 22:33:58 2000
From: donod at home.com (Donald O'Donnell)
Date: Thu, 07 Dec 2000 03:33:58 GMT
Subject: Finding values on sequences and .index()
References: <90lvnm$k2a$1@nnrp1.deja.com>
Message-ID: <3A2F0531.EBFD0C72@home.com>

rhamorim at my-deja.com wrote:
> 
> Hi.
> 
> I have a list that may have many ocurrences of the same data. How do I
> find the indexes for all ocurrences of it? list.index(value) only works
> for the first ocurrence.
...

How about using a list comprehension (ver 2.0) like this:

>>> a = [1, 2, 3, 4, 1, 5, 6, 7, 1, 8, 9, 1]
>>> n = 1
>>> x = [i for i in range(len(a)) if a[i]==n]
>>> x
[0, 4, 8, 11]

Or doing it the old fashioned way (ver 1.5):

>>> x = []
>>> for i in range(len(a)):
... 	if a[i] == n: x.append(i)
...
>>> x
[0, 4, 8, 11]

Hope this helps.

Don


From mikael at isy.liu.se  Mon Dec 11 07:06:07 2000
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Mon, 11 Dec 2000 13:06:07 +0100 (MET)
Subject: variable variables
In-Reply-To: <Pine.LNX.4.21.0012111127460.5044-100000@relax.dk>
Message-ID: <XFMail.001211130607.mikael@isy.liu.se>

On 11-Dec-00 pyth at relax.dk wrote:
 >  Sorry Mikael, for using your time... I should have been more specific...
 >  What i was thinking of, was something like this:

Well... At least I made you specify your question.

 >  Thanx for tryin' Mikael :o)

No problem.

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    11-Dec-00
Time:    13:04:03

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

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



From gregj at pobox.com  Sat Dec 16 02:28:09 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Sat, 16 Dec 2000 07:28:09 GMT
Subject: smtp and Groupwise
References: <91dtck$fgo$1@nnrp1.deja.com>
Message-ID: <dQE_5.153245$U46.4996917@news1.sttls1.wa.home.com>

"Brad" <blosh at tsbrass.com> wrote in message
news:91dtck$fgo$1 at nnrp1.deja.com...
> I am having a problem sending an email using smtplib.  we have an HPUX
10.20
> box and a novell server with groupwise for our pobox.  from unix, i am
trying
> to send an email using by smtp server on the novell server which works if
it
> is in house, not if it out house.  what am i overlooking?  tried the same
> thing from windows, same result....

One possibility: SMTP servers are often configured to reject connections
from networks or hosts they don't know. They also usually reject mail that
is not from a trusted host or network, or not addressed to a host or domain
the mail server accepts mail for. Mail servers that accept mail from
anywhere and send mail to anywhere are called open relays, and they are
considered a bad thing because spammers use them.

>From your outside machine try telnetting to the mail server on port 25 (SMTP
port) to see if it accepts your connection (stuff you type follows > in
example below):

> telnet my.mail.server 25
220 *** server name, etc.
> HELO your.workstation.name
250 ...

If you get this far the SMTP server is accepting your connection. If you
don't get this far you'll have to talk to your mail server administrator to
see if they can open the mail server up for you to use from outside.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From hwanjoyu at uiuc.edu  Wed Dec 20 12:56:28 2000
From: hwanjoyu at uiuc.edu (Hwanjo Yu)
Date: Wed, 20 Dec 2000 11:56:28 -0600
Subject: Q: dynamic function call
Message-ID: <Ko606.141$G66.2664@vixen.cso.uiuc.edu>

Hi,

Is there any way to bind a function name at runtime ?
For instance, we don't know the function name to call, but we know the
function name is saved in a variable, "fname".
How to call it in this case ?
Thanks in advance.





From joconnor at mail.cybermesa.com  Wed Dec  6 13:27:25 2000
From: joconnor at mail.cybermesa.com (Jay O'Connor)
Date: Wed, 06 Dec 2000 18:27:25 GMT
Subject: Schwartzian Transform (was Re: Sort)
Message-ID: <200012061827.LAA03198@mail.cybermesa.com>

An embedded and charset-unspecified text was scrubbed...
Name: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20001206/f6ddfd54/attachment.ksh>

From fredws_snb at ionet.net  Wed Dec  6 15:57:37 2000
From: fredws_snb at ionet.net (Fred Schroeder)
Date: Wed, 6 Dec 2000 14:57:37 -0600
Subject: Splash screen with Tkinter
Message-ID: <90m8oi$c3e1@news.cis.okstate.edu>

Could somebody tell me how to put up a splash screen with Tkinter? or if it
is very involved tell me where to look, I have tried several searches, but
have had no luck.
TIA
Fred




From joconnor at mail.cybermesa.com  Thu Dec 14 17:40:21 2000
From: joconnor at mail.cybermesa.com (Jay O'Connor)
Date: Thu, 14 Dec 2000 22:40:21 GMT
Subject: Getting a list of member functions
Message-ID: <200012142240.PAA14787@mail.cybermesa.com>

An embedded and charset-unspecified text was scrubbed...
Name: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20001214/bc0d5a55/attachment.ksh>

From SBrunning at trisystems.co.uk  Mon Dec  4 12:44:30 2000
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Mon, 4 Dec 2000 17:44:30 -0000 
Subject: Mailing lists are moving
Message-ID: <31575A892FF6D1118F5800600846864D5B153D@intrepid>

Barry,
Now, that is *excellent* news! Thanks.

Is the mailing list going to get all the usenet postings that it has missed
over the past couple of weeks? If so, it's going to make a *right* mess of
my mail-server - "Oh sir... it's only *wafer* thin". Still, I'll have
something to keep me occupied for a while!

I'll-have-to-get-the-hang-of-this-'ly y'rs'-stuff-again-ly y'rs,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From mikael at isy.liu.se  Fri Dec 22 02:10:31 2000
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Fri, 22 Dec 2000 08:10:31 +0100 (MET)
Subject: teekay
In-Reply-To: <F5N%5.108769$IP1.3655247@news1.giganews.com>
Message-ID: <XFMail.001222081031.mikael@isy.liu.se>

On 19-Dec-00 Nick Bensema wrote:
 >  In article <x%D%5.106612$IP1.3577747 at news1.giganews.com>,
 >  Nick Bensema <nickb at fnord.io.com> wrote:
 > >Tk would make a great baby name.
 >  
 >  It just occurred to me that if your baby was named Tk and you wanted
 >  to make him/her giggle, you could Tcl/Tk.

His/her humorous friends would probably call him/her decay...

Merry Xmas

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    22-Dec-00
Time:    08:07:19

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

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


From jvanar1 at my-deja.com  Fri Dec 29 12:51:03 2000
From: jvanar1 at my-deja.com (jvanar1 at my-deja.com)
Date: Fri, 29 Dec 2000 17:51:03 GMT
Subject: Need help with string method
References: <92if8v$iu1$1@nnrp1.deja.com> <slrn94pgo7.jbc.scarblac@pino.selwerd.nl>
Message-ID: <92iiu4$lvp$1@nnrp1.deja.com>

Right on the money.
#!/usr/local/bin/python2.0 cleans me right up , thanks a million...
-Jim


In article <slrn94pgo7.jbc.scarblac at pino.selwerd.nl>,
  scarblac at pino.selwerd.nl wrote:
> jvanar1 at my-deja.com <jvanar1 at my-deja.com> wrote in comp.lang.python:
> > I have a pretty dumb question here.  I can't seem to get any string
> > methods working from *within* a script.  here is a super-simple
example
> > that yields an error
> >
> > >>import sys
> > >>import string
> > >>theString = 'This is a string'
> > >>print theString.find('string')
> >
> > now this works fine from the python command line but gives an error
> > when run as a script. Here is the error message:
> >
> > Traceback (innermost last):
> >   File "./clean", line 18, in ?
> >     print theString.find('string')
> > AttributeError: 'string' object has no attribute 'find'
>
> Version mismatch. I think you tried it with a 2.0 command line but the
> script invokes the 1.5.2 interpreter (or similar). Try printing
sys.version
> from the script, and check the #! line (if it's on unix).
>
> --
> Remco Gerlich
>


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


From arildh at pasta.cs.uit.no  Fri Dec  8 03:27:18 2000
From: arildh at pasta.cs.uit.no (Arild Hansen)
Date: Fri, 8 Dec 2000 09:27:18 +0100
Subject: Tkinter frame resizeing
In-Reply-To: <t30nhn9kq4a5a3@corp.supernews.com>
References: <t30nhn9kq4a5a3@corp.supernews.com>
Message-ID: <Pine.LNX.4.21.0012080925520.931-100000@arildh.pasta.cs.uit.no>


To enable / disable the user to change the window size use:

self.master.resizable(1, 1)

Question 2 I can't help you with.

A. Hansen




On Fri, 8 Dec 2000, Tony wrote:

> 1.  How do you turn on the resize feature off for frames?
> 
> 2.  In Java you can modal windows?  How is this achieved with python using
> Tkinter?
> 
> 
> 



From andreas at andreas-jung.com  Tue Dec 12 07:26:04 2000
From: andreas at andreas-jung.com (Andreas Jung)
Date: Tue, 12 Dec 2000 13:26:04 +0100
Subject: Q: Print whithout \n
In-Reply-To: <aboZ5.3538$Kd1.410322@newsb.telia.net>; from henrik.martensson@frontconsulting.se on Tue, Dec 12, 2000 at 11:43:02AM +0000
References: <aboZ5.3538$Kd1.410322@newsb.telia.net>
Message-ID: <20001212132604.A27873@yetix.sz-sb.de>

On Tue, Dec 12, 2000 at 11:43:02AM +0000, Henrik M?rtensson wrote:
> Hi
> 
> How do you print a string whithout a linefeed in the end.

Use sys.stdout.write()

Andreas 



From gregj at pobox.com  Tue Dec 12 03:15:58 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Tue, 12 Dec 2000 08:15:58 GMT
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com>
Message-ID: <29lZ5.138780$U46.4638547@news1.sttls1.wa.home.com>

"Jaroslav Gergic" <j_gergic at yahoo.com> wrote:

> one thing I dislike in many Python tutorials is
> the style of infinite 'while' loop with 'break' statements:

Learning the syntax of a language will satisfy the compiler, but learning
the idioms will make your code accessible to other programmers, and more
important it will make other programmer's code useful to you. All computer
languages are idiomatic to some degree. It doesn't matter if the idioms are
the "right" way to do things or if they even make sense at first sight; what
is important is that the idioms are commonly used and understood. Reading
code written by someone who doesn't know the idioms is like talking to
someone who knows your language but none of the figures of speech or common
usages (rather like listening to Al Gore).

FORTRAN programs generally use I, J, and K as loop indices because variables
starting with I, J, K are implicitly integers (type declarations are
optional in FORTRAN). This idiom was perpetuated by programmers even when
they moved out of FORTRAN into languages with explicit typing (C, C++,
Pascal) and languages with no typing (Visual Basic, Python).

The C/C++ loop idioms are:

  for (; ;) { ... }

and less common:

  while (1) { ... }

There are any number of ways to write these loops, but the important point
is that the idiom is understood by C/C++ programmers. In "The C Programming
Language" the authors evolve the implementation of the strcpy() function
(copy a string) from the explicit but clunky loop a Basic or FORTRAN
programmer might write to the terse but elegant C idiom that confounds so
many beginniners: "while (*t++ = *s++) ;"

Aside from the loop construction, the variable names s and t are C idioms
introduced by Kernighan & Ritchie. Condemnations of the C idioms--especially
those involving pointers--crop up in the newgroups regularly. A flame war
broke out a few years ago in comp.lang.c++ over which was better: for ( i=0;
i < N; i++ ) or for (i=0; i < N; ++i ). The result is exactly the same, and
K&R use the first (postincrement) idiom, but it turns out that ++i is
usually a tiny bit more efficient than i++. I don't know what the resolution
was, but I notice the preincrement usage is much more common in C++ code
than it is in C.

My advice: learn the idioms of the languages you use.

"...I'm afraid when I'm in this idiom, I sometimes get a bit, uh, sort of
carried away."
    -- Sir Launcelot in "Monty Python and the Holy Grail"

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From gregj at pobox.com  Wed Dec 20 01:08:20 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Wed, 20 Dec 2000 06:08:20 GMT
Subject: while true: !!!
References: <B43D149A9AB2D411971300B0D03D7E8B0FBCBB@natasha.auslabs.avaya.com> <v04220801b663769e1800@[192.168.11.33]> <mailman.977147642.18708.python-list@python.org> <91lclt02qr3@news1.newsguy.com> <slrn93st0m.6c1.qrczak@qrnik.zagroda> <91o3l501i59@news2.newsguy.com> <jJN%5.52075$x6.23325177@news2.rdc2.tx.home.com> <91of2t01um1@news2.newsguy.com>
Message-ID: <o1Y%5.172255$U46.5441899@news1.sttls1.wa.home.com>

"Alex Martelli" <aleaxit at yahoo.com> wrote in message
news:91of2t01um1 at news2.newsguy.com...
> If you mean 'char is a numeric type' (in C but not Python),
> ...
> I don't know of any distinction drawn _in C_ between
> 'integer' and 'non-integer' numeric types, that would make
> 'integer' somehow a significant C term -- am I just having
> a brainstorm...?).

You're having a brainstorm. There is definitely a distinction between
integer and non-integer types in C. And char is most definitely an integer
type.

C has four integer types that may be signed or unsigned: char, short, int,
long. In K&R C They do not have to be different sizes; the definition only
requires that sizeof(char) <= sizeof(short) <= sizeof(int) <= sizeof(long).
Typical x86 compilers have 8-bit chars, 16-bit shorts, 16- or 32-bit ints,
and 32-bit longs. K&R says that the integer types should map to the
"natural" sizes for the host hardware. ANSI C specifies minimum ranges for
the integer types.

K&R C has two floating-point types: float and double. ANSI C added long
double. Again K&R only requires that the precision of float <= precision of
double, and ANSI C defines minimum precisions.

Just some of the differences between integer and floating-point types: The
results of bitwise operators are only defined for int types. Bit fields must
be one of the int types. ANSI C only allows integer types as the switch(
... ) condition. And so on.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From erno-news at erno.iki.fi  Wed Dec 27 18:29:05 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 28 Dec 2000 01:29:05 +0200
Subject: Python and Napster
References: <000c01c06fe3$efc29520$102b2bc1@cadlab.it> <92ap0g12b6r@news1.newsguy.com> <sx226.18970$RC1.511898@newsread1.prod.itd.earthlink.net> <20001227164900.DF7E6A84F@darjeeling.zadka.site.co.il> <mailman.977910196.13387.python-list@python.org> <92dptl11vqa@news2.newsguy.com>
Message-ID: <kuzohhqw8u.fsf@lasipalatsi.fi>

Alex Martelli <aleaxit at yahoo.com> writes:

| What's the transfer-resumption transaction, then?

<URL:http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html#sec14.35>

  -- erno


From ssthapa at harper.uchicago.edu  Thu Dec  7 03:26:19 2000
From: ssthapa at harper.uchicago.edu (Suchandra Thapa)
Date: Thu, 07 Dec 2000 08:26:19 GMT
Subject: parallel port access?
References: <3A268378.4DA1B831@sympatico.ca> <3A272881.6F444750@engcorp.com> <3A2D6761.814ABE8E@bioeng.ucsd.edu>
Message-ID: <slrn92su1q.12g.ssthapa@localhost.localdomain>

Curtis Jensen <cjensen at bioeng.ucsd.edu> wrote:
>There has been so many questiona about sieral and parallel port access
>throuhg Python.  How come it hasn't been put into the distrabution?  I
>realize that it is platform specific, but so are other things that are
>in Python.  Or even put it in as a platform specific module like the gl
>module.

    The parallel port and serial port controls are really platform specific.
Windows has a standard serial interface but for the parallel interface,
windows 98 allows direct access but nt requires a kernel driver.  Unix systems
probably have there own oddities ranging from non-pc machines which may not 
have a parallel or serial port to whatever access mechanisms linux/*BSD/HURD
have.  The task would probably come fairly close to writing a new driver 
for each platform.


-- 
------------------------------------------------------------------
			    |
Suchandra Thapa             | "There are only two kinds of math books. 
s-thapaNO at SPAMuchicago.edu  | Those you cannot read beyond the first 
			    | sentence, and those you cannot read 
			    | beyond the first page."
			    |                       -C.N. Yang
------------------------------------------------------------------


From tim.one at home.com  Fri Dec 22 15:21:19 2000
From: tim.one at home.com (Tim Peters)
Date: Fri, 22 Dec 2000 15:21:19 -0500
Subject: ANNOUNCE: soundex.py
In-Reply-To: <slrn94787k.1q7.amk@207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCECJIFAA.tim.one@home.com>

[A.M. Kuchling]
> Does this mean that Modules/soundex.c can be dropped from the Python
> CVS tree?

Alas, I don't think so, unless we get Special Dispensation from Guido.
Until PEP 5 is accepted, we don't have a well-defined way to get rid of a
module.  The distributed 2.0 docs call the soundex module "obsolete", but
don't say anything about it ever going away.  While the details of PEP 5
aren't all agreed upon yet, I believe most people agree with the basic
notion that modules can't go away without an explicit warning about that
first.




From tcmay at got.net  Fri Dec 15 18:19:16 2000
From: tcmay at got.net (Tim May)
Date: Fri, 15 Dec 2000 15:19:16 -0800
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <3a36c85e.4409136@News.CIS.DFN.DE> <owen_nospam-1312001131590001@cs204-90.fsmodem.washington.edu> <3A37DD31.C9A5F18C@pacbell.net> <91bhlh$c64u$1@nntp6.u.washington.edu> <3A395197.CB2974D5@home.com> <L_r_5.7453$n51.136546@e420r-atl1.usenetserver.com> <steve976915950@serf.org>
Message-ID: <tcmay-E53C56.15191615122000@news.got.net>

In article <steve976915950 at serf.org>, steve at serf.org (A SERFer) wrote:


> Indeed we are!  We keep hearing a circular argument:
> 1. Cincom are committed to Smalltalk in the long term.
> 2. We must charge these prices if Smalltalk is to survive within Cincom.
> 
> Its like a sort of quantum mechanical uncertainty thing... each statement
> on its own makes sense, but put together, they are meaningless.

Really? Translate it to this:

1. Intel is committed to the Pentium in the long term.
2. We must charge these prices if the Pentium is to survive within Intel.

Or translate it to any number of other softare tools, such as chip 
layout software, PCB software, design rule checking tools. Or factory 
management tools. Or data bases.

1. Oracle is committed to its database products in the long term.
2. We much charge these prices if these products are to survive withing 
Oracle.

Fact is, some software products are quite expensive, some are moderately 
expensive, some are inexpensive, and some are even free.

My group at Intel, many years ago, paid about $25K per seat for a 
product from Intellicorp (no relation to us).

I won't pretend to know why some products do well at certain prices, or 
whether Cincom is charging "too much."


> Having done some research into Java's history, I would guess that Sun are
> in fact making money from Java:  one of the motivations was to provide
> a common platform and a common set of APIs for internal Sun use as well
> as a product.  Even if Java had only succeeded as an internal tool, it
> would have cut costs, and so made money for Sun.

I assure you, Sun is _not_ making money from Java. At least not 
directly. Maybe through some tangential benefits to their real products, 
but I doubt even that. Sun machines would be selling even if Java had 
never appeared. 


--Tim May

-- 
Timothy C. May         tcmay at got.net        Corralitos, California
Political: Co-founder Cypherpunks/crypto anarchy/Cyphernomicon
Technical: physics/soft errors/Smalltalk/Squeak/agents/games/Go
Personal: 1951/UCSB/Intel '74-'86/retired/investor/motorcycles/guns




From pearu at cens.ioc.ee  Thu Dec 28 05:07:03 2000
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Thu, 28 Dec 2000 12:07:03 +0200
Subject: Python 2.0 parser module bug(?) and a Q
In-Reply-To: <p6qofxxe43j.fsf@informatik.hu-berlin.de>
References: <Pine.LNX.4.10.10012272204060.330-100000@kev.ioc.ee> <p6qofxxe43j.fsf@informatik.hu-berlin.de>
Message-ID: <Pine.LNX.4.10.10012281200430.23773-100000@kev.ioc.ee>


On 28 Dec 2000, Martin von Loewis wrote:

> Pearu Peterson <pearu at ioc.ee> writes:
> 
> > I have found that parser module may have a bug in its functions tuple2ast
> > or suite.
> 
> Please report bugs on sourceforge.net/projects/python; on this list,
> they might easily be overlooked.

OK, will do.

> > And my question is (not really related to above):
> > 
> >   Is there inverse function for parser.suite? 
> 
> A colleague of mine has an "unparser" module for Python AST; that is
> not (yet) released to the public, and works not (yet) for 2.0. Please
> contact me if you need more details.

Thanks, but I already wrote a simple inverse parser.suite my self.
See the function ast2string below.

Any constructive critisism is welcome!

Thanks,
	Pearu

from types import *
import string
import parser,token

def _asttuple2str(t,indent=[0]):
    if not isinstance(t,TupleType):
        if isinstance(t,StringType):
            return t+' '
        return '' # if line_info=1 in ast2tuple or ast.totuple
    if not isinstance(t[0],TupleType):
        if token.tok_name.has_key(t[0]):
            if token.tok_name[t[0]]=='INDENT':
                indent[0]+=1
                return '\t'
            if token.tok_name[t[0]]=='DEDENT':
                indent[0]-=1
                return '\r'
            if token.tok_name[t[0]]=='NEWLINE':
                return '\n'+indent[0]*'\t'
            if token.tok_name[t[0]]=='ENDMARKER':
                return t[1]
    return string.join(map(_asttuple2str,t[1:]),'')
def asttuple2str(t,tab=4*' '):
    return _asttuple2str(t).replace('\t\r','').replace('\t',tab)
def ast2string(ast):
    return asttuple2str(ast.totuple())




From ten.llun at ttam  Sun Dec 17 18:51:45 2000
From: ten.llun at ttam (Matthew Schinckel)
Date: Mon, 18 Dec 2000 10:21:45 +1030
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <915jcf0c6o@news2.newsguy.com> <9162p3$7t2$1@animus.fel.iae.nl> <MBWZ5.334$h67.37748@newsread1.prod.itd.earthlink.net> <3A397C74.6C8906E1@cosc.canterbury.ac.nz> <91d5520m6j@news1.newsguy.com> <wvw_5.10652$h67.670217@newsread1.prod.itd.earthlink.net> <slrn93mh41.1nk.scarblac@pino.selwerd.nl>
Message-ID: <3a401c31_5@news.chariot.net.au>

In message <slrn93mh41.1nk.scarblac at pino.selwerd.nl>, Remco Gerlich wrote:
> Oldayz <sill at localhost.kitenet.net> wrote in comp.lang.python:
> > Hey, how about if you think while 1: looks ugly
> > you do true =3N 1 somewhere on top and use
> > while true: ? I think it's much more obvious
> > and clean and you don't need to change syntax.
> 
> Yes! And you don't need the ugly "break" statement, you can simply do
> "true =3N 0" and the loop will end! (Remember to set it back to 1 later)
> 
:-)

(I'm the guy who forgets to set it back :-)

In my freshman Pascal course, all those years ago (although, until last year
they still taught the same course), I was really pleased that I came up with
(may be bad syntax, haven't used Pascal since 93 :)

HELL_FREEZES_OVER =3N 0

repeat:
 # do stuff
until HELL_FREEZES_OVER


-- 
Matthew Schinckel
matt at null dot net



From emile at fenx.com  Wed Dec  6 08:23:45 2000
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 06 Dec 2000 05:23:45 -0800
Subject: [OT] XML introduction
References: <d8qX5.227$rZ6.6363@m2newsread.uni2.es>
Message-ID: <3A2E3DE1.8060809@fenx.com>

http://www.oasis-open.org/cover/xmlIntro.html

Emile van Sebille
emile at fenx.com

Arturo P?rez Mulas wrote:

> Hello,
> 
>  
> 
> this is kind of off-topic, but not quite. I cannot find any 
> comprehensive introduction
> 
> to XML that gives real information on what is it used for. As far as I 
> could grasp,
> 
> it allows to define structures to store data cretaing your html-like 
> tags, and rules
> 
> to format them on a xml-compliant browser... but I keep reading things 
> about
> 
> "developing web sites or applications with xml" and then I get lost.
> 
>  
> 
> Anyone any pointer or hint, please?
> 
>  
> 
> Thanks
> 
>  
> 
> Arturo P?rez



From claird at starbase.neosoft.com  Sun Dec 31 12:48:18 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 31 Dec 2000 11:48:18 -0600
Subject: Pythonic concurrency (was: Erlang (was Re: Good introduction to functional programming with Python?))
References: <mailman.977984412.30211.python-list@python.org> <92f0il01a5u@news1.newsguy.com> <86wvckqw4h.fsf@localhost.my.domain> <G6AyoC.CK1@world.std.com>
Message-ID: <92A081D417204A91.AF29627FD1BF4835.0C336CAB4734825B@lp.airnews.net>

In article <G6AyoC.CK1 at world.std.com>, Will Ware <wware at world.std.com> wrote:
>Brad Knotwell (knotwell at ix.netcom.com) wrote:
>> I'm now wondering if Stackless Python's uthreads are cooperatively or
>> pre-emptively multi-tasked.
>
>Unless you're digging into their inner workings, microthreads will appear
>to you as preemptively multitasking.
			.
			.
			.
Is this true:  microthreads block on system calls (I/O
operations, most notoriously) and on invocations of
externally-coded extensions?  
-- 

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


From grante at visi.com  Mon Dec 18 12:45:55 2000
From: grante at visi.com (Grant Edwards)
Date: 18 Dec 2000 11:45:55 -0600
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Dec 18)
Message-ID: <FC292AEAD3D45E69.7A8DC95A0D485F4F.DB35E5138BC5B200@lp.airnews.net>

    Kirby Urner discusses using Python for crypto (and math in
    general).
        http://deja.com/=dnc/getdoc.xp?AN=703617917

    Fredrick Lundh provides an excellent description of Python
    variable/parameter/assignment semantics.
        http://deja.com/=dnc/getdoc.xp?AN=704134659

    Jon Nicoll is working on a module to provide direct access to
    parallel port lines under Win95.
        http://deja.com/=dnc/getdoc.xp?AN=704416116

    Call for Papers: LCTES 2001 (ACM SIGPLAN Workshop)
        http://deja.com/=dnc/getdoc.xp?AN=704719354


========================================================================

Everything you want is probably one or two clicks away in these pages:

      Python.org's Python Language Website is the traditional
      center of Pythonia
          http://www.python.org

      PythonWare complements the digest you're reading with the
      daily python url
           http://www.pythonware.com/daily  

      comp.lang.python.announce announces new Python software.  Be
      sure to scan this newly-revitalized newsgroup at least weekly.
          http://deja.com/group/comp.lang.python.announce

      Andrew Kuchling writes marvelous summaries twice a month of
      the action on the python-dev mailing list, where the future
      of Python is truly determined
	  http://www.amk.ca/python/dev

      The Vaults of Parnassus ambitiously collects Python resources
          http://www.vex.net/~x/parnassus/   

      Much of Python's real work takes place on Special-Interest Group
      mailing lists
          http://www.python.org/sigs/

      Python Consortium emerges as an independent nexus of activity
          http://www.python.org/consortium

      Cetus does much of the same
          http://www.cetus-links.de/oo_python.html

      Python FAQTS
          http://python.faqts.com/

      Python To-Do List anticipates some of Python's future direction
          http://www.python.org/cgi-bin/todo.py

      Python Journal is at work on its second issue
          http://www.pythonjournal.com

      Links2Go is a new semi-automated link collection; it's impressive
      what AI can generate
          http://www.links2go.com/search?search=python

      Archive probing trick of the trade:
          http://www.dejanews.com/dnquery.xp?QRY=&DBS=2&ST=PS&defaultOp=AND&LNG=ALL&format=threaded&showsort=date&maxhits=100&groups=comp.lang.python


Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://purl.org/thecliff/python/url.html
or
  http://www.dejanews.com/dnquery.xp?QRY=~g%20comp.lang.python%20Python-URL%21


Suggestions/corrections for next week's posting are always welcome.
[http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
e-mail us ideas directly.]

To receive a new issue of this posting in e-mail each Monday morning,
ask <claird at neosoft.com> to subscribe.  Mention "Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.
-- 

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


From aahz at panix.com  Tue Dec 12 17:48:17 2000
From: aahz at panix.com (Aahz Maruch)
Date: 12 Dec 2000 14:48:17 -0800
Subject: quoting styles (was Re: Label-Value)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <kupuix71b6.fsf@lasipalatsi.fi> <9167qo$g4s$1@panix2.panix.com> <mtxZ5.3611$Kd1.423896@newsb.telia.net>
Message-ID: <9169vh$k9i$1@panix6.panix.com>

In article <mtxZ5.3611$Kd1.423896 at newsb.telia.net>,
Fredrik Lundh <fredrik at effbot.org> wrote:
>Aahz wrote:
>>
>> What's "jeopardy style" quoting?
>
>When you see the answer before the question.

Oh, *duh*.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

What if there were no rhetorical questions?  --Aahz


From paul.moore at uk.origin-it.com  Tue Dec 19 08:06:17 2000
From: paul.moore at uk.origin-it.com (Paul Moore)
Date: Tue, 19 Dec 2000 14:06:17 +0100
Subject: How to rewrite this as list comprehension?
References: <91nj2q$s3p$1@buty.wanadoo.nl>
Message-ID: <OVs=OikIak7qdTmmq0E5rkRKrGiz@4ax.com>

On Tue, 19 Dec 2000 13:07:49 +0100, "Serge Beaumont"
<serge at sergebeaumont.com> wrote:

>Hello,
>
>I'm trying to get my head around comprehensions. How would I rewrite this
>method?
>
>  def deadThings(self, type = None):
>    deadThings = self.destroyed[:]
>    for thing in self.things:
>      if thing.destroyed:
>        if type:
>          if isinstance(thing, type):
>            deadThings.append(thing)
>        else:
>          deadThings.append(thing)
>    return deadThings

I think it goes:

deadThings = self.destroyed[:]
deadThings += [x for x in self.things
                 if (x.destroyed and
                    (not type or isinstance(x, type)))]

Start from the inside and work out. The nasty (not type or ...) bit is
a translation of the "if type: if isinstance(): else:" bit into an
expression as opposed to a statement. In C I might have written
    (type ? isinstance() : true)
but I'm not sure it's much clearer.

Whether you find the list comprehension clearer is a matter of taste.
Just because it exists, doesn't mean it is the right answer in all
cases...

Paul



From bowman at montana.com  Mon Dec 25 09:25:32 2000
From: bowman at montana.com (bowman)
Date: Mon, 25 Dec 2000 07:25:32 -0700
Subject: ready for C or C++?
References: <20001225031852.04296.00003766@ng-fp1.aol.com>
Message-ID: <mSI16.9502$wE3.10504@newsfeed.slurp.net>

Xmnemonic <xmnemonic at aol.com> wrote in message
> If I've learned Python and become proficient with it, am I ready to learn
C or
> C++?

yeah, but why would you want to? seriously, it shouldn't be a problem. C++
can be complex, but can have a higher level of abstraction, more in keeping
with Python. C is simpler but this implies if you want a dictionary, for
instance,
you will have to learn a lot more about the underlying data structure, since
you'll be building your own linked list, hash table or however you choose to
implement it. otoh, if you learn the nuts and bolts using C, you'll be able
to
see the concepts of more abstract languages a bit easier.

if you've spent some happy hours trying to build a usable hash, you'll
appreciate some of the arcane points, like why Java's first cut at a hash
degenerated into a list when fed a bunch of 'http://.....' keys...






From g_will at cyberus.ca  Thu Dec 21 23:33:51 2000
From: g_will at cyberus.ca (Gordon Williams)
Date: Thu, 21 Dec 2000 23:33:51 -0500
Subject: Line Benchmarks - Tkinter vs. wxPython
Message-ID: <91uljk$12as$1@news2.ottawa.cyberus.ca>

Hello All,

In November of 1999 I did some benchmark tests to gauge the performance
difference between Tkinter and wxPython when large numbers of lines are
plotted.  At that time wxPython outperformed Tkinter by a large margin as
can be seen in the results that are repeated below.  As a full year has
passed and there are new versions of Python, Tkinter and wxPython, I felt
that it was time to rerun the tests to see if there were any significant
changes for better or worst.
While Tkinter has improved significantly, it is still far behind wxPython in
this aspect by a margin of 20 to 30 times.

The same computer was used to generate the new set of results, so the
results are directly comparable.  As described below, a repeating sine wave
was created and put into a horizontal scrollable window.  These are the
results
for the line plotting time only.  Time in Seconds. (W95, Python 2.0,
wxPython 2.2.2)

Points      Tkinter      Tkinter   Tkinter    wxPython     wxPython
                (standard) (mod 1)  (mod 2)  (direct draw) (bitmap)
1000        0.163        0.162      0.157         0.0078            0.028
2000        0.324          --            --             0.0130
0.041
5000        0.817          --            --             0.0390
0.092
10000      1.62          1.63        1.64           0.0590            0.157

See the notes below for more information on the configuration for each run.

Conclusions:
1) The biggest change is in the performance of Tkinter (standard).
Previously, the time to plot lines using the standard Tkinter was quadratic
with the
number of points (ugh!). Now it is linear and the results are very similar
to both of the modified Tkinter runs.  An improvement was made to speed up
the_flatten method where most of the time was spent previously.

2) Tkinter mod 1 and 2 give almost the same results.  There is a slight
improvement over the last years results by 15 to 20%.  Some other changes
must have been made to the code to get these efficiencies.

3) wxPython results have changed very little.  Maybe they have gotten 5 to
10% slower, but it is difficult to say without doing a more though study.
Direct drawing onto to screen is still roughly 3 times faster than using a
bitmap.

4) wxPython still beats Tkinter by a wide margin.  Although the number are
not as huge as they were a year ago, wx still goes 20 - 30 times faster.

If you need raw speed to get large numbers of lines out to the screen,
wxPython is the way to go.

It would be interesting to have this repeated on a Unix box to get
comparable results.

Regards,

Gordon Williams
Dec. 2000



########################################################
November 1999 (Windows 95, Python 1.5.2, wxPython ?)

I have done some benchmark testing to see the difference in speed between
plotting multi point lines using Tkinter and wxPython.  I have looked at 3
different methods for doing it in Tk and 2 different methods in wx.  The
differences are astounding for the 1000 to 10000 point lines that I
examined.  Generally, wx was found to be in the order of 200 to 1800 times
faster than the standard Tkinter.  Tk can be sped up by ten to 50 times by
changing Tkinter so it bypasses the _flatten method or by making the Tk
calls directly.

A repeating sine wave was created and put into a horizontal scrollable
window of about 500 points wide and 300 points high.  These are the results
for the line plotting time only.Time in Seconds on a P166 with 96MB RAM.

Points      Tkinter      Tkinter   Tkinter    wxPython     wxPython
                (standard) (mod 1)  (mod 2)  (direct draw) (bitmap)
1000        1.190        0.192      .202         0.007            0.026
2000        4.406        0.385        --           0.014            0.042
5000      25.170        0.967        --           0.027            0.080
10000  100.260        1.941     2.064        0.054            0.148

Notes:
1) Tk standard is the standard usage with no modifications.
2) Tk mod 1 has Tkinter modified so a flat tuple of points is passed to
create_line and _flatten is by-passed.
3) Tk mod 2 used the apply method  e.g.
apply(self.can.tk.call,(self.can._w,'create','line') + points +
('-width',1,'-fill',colour,'-tags',"tag"))
4) wx direct draw draws to the window directly.
5) wx bitmap draws to a bitmap and then blit

Conclusions:
1) The time to plot lines using the standard Tkinter is quadratic with the
number of points (ugh!).  This is due to the time spent in _flatten.
_flatten is very versatile but agonizingly slow for large numbers of lines.
2) Tkinter mod 1 and 2 give the same results.  Times are linear with the
number of points plotted as both cases dont use the _flatten method.  This
means that a flat tuple of points has to be used (a small constraint).  The
speed increase in doing this is from 10 to 50 times over the standard.
3) wxPython beats Tk by a very! large margin even when Tk is called
directly.  Direct drawing onto to screen is roughly 3 times faster than
using a bitmap.

Thanks to:
Greg Landrum for providing some information on direct calls to tk.
Ionel Simionescu for providing the wxPython code that I was able to modify.

############################################################
I know I am going to get questions on where is the code, so here it is.  I
have not supplied my version of Tkinter (Tkinter_gw.py) because it is large.

############################################################
#Ionel2_test_spd.py (for draw to bitmap)
__author__ = "Ionel Simionescu"  # messed up by Gordon Williams
#---------------------------------------------------------------------------
from wxPython.wx import *
import time
from math import sin

#---------------------------------------------------------------------------
t0 = time.clock()

N = 1000                # the number of line points


points = []         # here we store the points
for k in range(N):
    points.append( (k, 50+ 50*sin(k/10.)) )
#print points
#print points

t1 = time.clock()
print 'Data is ready! Total data preparation time:  %f ' %  (t1 - t0)

#---------------------------------------------------------------------------

bmpW = N #changed
bmpH = 300

#---------------------------------------------------------------------------
class myCanvas(wxScrolledWindow):
    def __init__(self, parent, ID, position, size):
        wxScrolledWindow.__init__(self, parent, ID, position, size,
wxSUNKEN_BORDER)
        self.SetBackgroundColour(wxNamedColor("WHITE"))

        self.maxWidth  = 500
        self.maxHeight = 300
        self.SetScrollbars(20, 0, N/20, 0) #self.maxWidth/20,
self.maxHeight/20)

        self.bmp = wxEmptyBitmap(bmpW, bmpH, -1)
        #self.bmp.Create(bmpW, bmpH, -1)
        #
        self.bmp_w = self.bmp.GetWidth()
        self.bmp_h = self.bmp.GetHeight()

        # flag used to redraw only when necessary
        self.new_data = 1
        EVT_PAINT(self, self.OnPaint)

    def OnPaint(self, event):
        dc = wxPaintDC(self)
        self.PrepareDC(dc)
        self.DoDrawing(dc)

    def DoDrawing(self, dc):
        t0 = time.clock()
        pen1 = wxPen(wxNamedColour('RED'))

        tmp_dc = wxMemoryDC()
        tmp_dc.SelectObject(self.bmp)

        # if new data, update the bitmap drawing
        if self.new_data:
            tmp_dc.BeginDrawing()
            tmp_dc.Clear()
            tmp_dc.SetPen(pen1)
            # draw the lines
            tmp_dc.DrawLines( points ) #changed
            tmp_dc.EndDrawing()
            # reset the new_data flag
            self.new_data = 0

        # copy the bitmap to the canvas
        dc.Blit( 0, 0, self.bmp_w, self.bmp_h, tmp_dc, 0, 0 )

        t1 = time.clock()
        print 'Plotting is ready! Total drawing time:  %f ' %  (t1 - t0)



#---------------------------------------------------------------------------
class myFrame(wxFrame):
    def __init__(self):
        wxFrame.__init__(
            self, NULL, -1, "wx_plot_lines",
            wxDefaultPosition, size= (500,300)#wxSize(500, 300)
        )
        myCanvas(self, -1, wxPoint(0,0), self.GetClientSize())

#---------------------------------------------------------------------------
class myApp(wxApp):
    def OnInit(self):
        myFrame().Show(TRUE)
        return TRUE

#---------------------------------------------------------------------------
if __name__ == '__main__':
    app = myApp(0)
    app.MainLoop()

#######################################################
#Ionel1_test_spd.py (for direct draw method)
__author__ = "Ionel Simionescu" # messed up by Gordon Williams
#---------------------------------------------------------------------------
from wxPython.wx import *
from math import sin
import time

#---------------------------------------------------------------------------
t0 = time.clock()

N = 10000                # the number of line points


points = []         # here we store the points
for k in range(N):
    points.append( (k, 50+ 50*sin(k/10.)) )
#print points

t1 = time.clock()
print 'Data is ready! Total data preparation time:  %f ' %  (t1 - t0)


#---------------------------------------------------------------------------
class myCanvas(wxScrolledWindow):
    def __init__(self, parent, ID, position, size):
        wxScrolledWindow.__init__(self, parent, ID, position, size,
wxSUNKEN_BORDER)
        self.SetBackgroundColour(wxNamedColor("WHITE"))

        self.maxWidth  = 500
        self.maxHeight = 300
        self.SetScrollbars(20, 0, N/20, 0) #self.maxWidth/20,
self.maxHeight/20)
        EVT_PAINT(self, self.OnPaint)

    def OnPaint(self, event):
      dc = wxPaintDC(self)
      self.PrepareDC(dc)
      self.DoDrawing1(dc)

    def DoDrawing1(self, dc):
        t0 = time.clock()

        dc.BeginDrawing()
        pen1 = wxPen(wxNamedColour('RED'))
        dc.SetPen(pen1)
        #
        # draw line_points
        dc.DrawLines(points, 0)
        #
        dc.EndDrawing()

        t1 = time.clock()
        print 'Plotting is ready! Total drawing time:  %f ' %  (t1 - t0)



class myFrame(wxFrame):
    def __init__(self):
        wxFrame.__init__(
            self, NULL, -1, "wx_plot_lines",
            wxDefaultPosition,wxSize(500,300)
        )

        myCanvas(self, -1, wxPoint(0,0), self.GetClientSize())

#---------------------------------------------------------------------------
class myApp(wxApp):
    def OnInit(self):
        myFrame().Show(TRUE)
        return TRUE

#---------------------------------------------------------------------------
def test():
    app = myApp(0)
    app.MainLoop()

if __name__ == '__main__':
    test()

##########################################################
# File for benchmarking multi-point lines
# in wxPython and Tkinter.
#
# This file is for Tkinter
# Change the commented sections as required.
#
# Author: Gordon Williams 99/11/15
#



import time
from math import sin
#from Tkinter_gw import *
from Tkinter import *

class LayoutFrames:
    def __init__(self,master):
        # ********** Frame Layouts ***************
        # Main Frame Layout

        self.data= Frame(master)
        self.data.pack(side=TOP,fill=X)

        self.scroll= Frame(self.data)
        self.scroll.pack(side=BOTTOM,fill=X)


class Plot:

    def __init__(self, master, master_scroll=None):

# dataframe with canvas and scollbar
        self.scrollbar = Scrollbar(master_scroll, orient=HORIZONTAL)
        self.can = Canvas(master,cursor="crosshair", width=500,
height=300,relief=RIDGE, bd=2,bg="lightyellow",
xscrollcommand=self.scrollbar.set)
        self.scrollbar.config(command=self.can.xview)
        self.scrollbar.pack(side=BOTTOM, fill=X)
        self.can.pack(side=LEFT)


    def makelines(self,points, tag="misc",colour="black"):
        t0 = time.clock()
#>>>> comment out as required (use for standard and mod1 runs)
        #self.can.create_line(points,width= 1, tags=tag, fill=colour)

#>>>> comment out as required (use for mod2 runs)
        apply(self.can.tk.call,(self.can._w,'create','line') + points +
            ('-width',1,'-fill',colour,'-tags',"tag"));


        self.can.config(scrollregion=self.can.bbox(ALL))

        t1 = time.clock()
        print 'Plotting is ready! Total drawing time:  %f ' %  (t1 - t0)



class PlotResults:
    def __init__(self, dataFrame, scrollFrame):

    # ********** PLOTTING DATA ***************
        self.analog = Plot(dataFrame,scrollFrame)


    def analogPlot(self):
        self.analog.makelines(linePoints, "dataA","black" )


#************ start of main **************************
def main():

    root= Tk()
    frame=LayoutFrames(root)

    analog=PlotResults( frame.data, frame.scroll)
    analog.analogPlot()

    root.mainloop()
    try:
        root.destroy() # removes root
    except:
        pass


# make a bit of data
N= 1000   # number of points
points = []         # here we store the points
for k in range(N):
    points.extend( [k, 50+ 50*sin(k/10.)] )
#print points


linePoints= tuple(points)  # must be in tupple form for Tkinter_gw
main()
#######################################################
Enjoy!!





From moshez at zadka.site.co.il  Thu Dec 28 11:22:03 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: 28 Dec 2000 16:22:03 -0000
Subject: Python vs PHP (& Java?)
In-Reply-To: <92f3kp01cqe@news1.newsguy.com>
References: <92f3kp01cqe@news1.newsguy.com>, <mailman.977897649.4877.python-list@python.org> <slrn94l86m.a3h.tim@degree.ath.cx>
Message-ID: <20001228162203.7163.qmail@stimpy.scso.com>

On Thu, 28 Dec 2000 11:10:31 +0100, "Alex Martelli" <aleaxit at yahoo.com> wrote:

> The 'good-applicability niches' of Perl and Python are vastly
> overlapped (and not so small as to really deserved being called
> 'niches', but that's another issue).  I can't think of any program
> I would rather write in Perl than in Python

Here's an example, coming from one of the major anti-Perl zealots
around here: Apache modules. Yes, now that we have mod_snake, you
can do these things in Python, in theory: but mod_snake is still
young, while mod_perl is quite mature.

Let me assure you that I miss Python's elegance and clarity sorely,
so I hope mod_snake will mature quickly <wink>

(BTW: Are there plans for quixote/mod_snake integration? That'd
be ultra-cool: one of the benefits of HTML::Mason is that it
integrates really well with mod_perl)



From fredp at mygale.org.nospam  Wed Dec  6 03:24:09 2000
From: fredp at mygale.org.nospam (Fred Pacquier)
Date: 6 Dec 2000 08:24:09 GMT
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <3A2DD2CA.14A5D7E2@engcorp.com>
Message-ID: <90025162APaCmAnRDLM@161.48.76.20>

peter at engcorp.com (Peter Hansen) said :

>Nigel wrote:
>> 
>>  I am doing some research and I would be grateful if anybody could give
>> me some history on the origins of Python.
>> 
>> --
>> Worker
>> Disk Drive Computer Books
>> http://disk.vstorecomputers.com
>
>Sounds like an interesting and useful endeavor.  This might give you a
>good start:

ROTFL - great stuff ! :-)))

-- 
YAFAP : http://www.multimania.com/fredp/


From loewis at informatik.hu-berlin.de  Wed Dec 27 19:50:48 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 28 Dec 2000 01:50:48 +0100
Subject: Python Database Interfaces - Any Standard?
References: <RUo16.204045$RP.37382354@news1.rdc1.md.home.com> <kuk88pu0cd.fsf@lasipalatsi.fi> <Gwu16.205577$RP.37521696@news1.rdc1.md.home.com> <3A483B53.ACB0C04D@python.net>
Message-ID: <p6qu27pe5cn.fsf@informatik.hu-berlin.de>

"Thomas A. Bryan" <tbryan at python.net> writes:

> I haven't done much with Python and databases (yet), but I think that the 
> idea is to be able to write 
> 
> import DCOracle
> dbh = DCOracle.Connect( connect_string )
> query = dbh.cursor()
> query.execute( query_string )
> 
> 
> To change to PostgreSQL, you'd have to change the import and the connection.

If you anticipate change to another database, you write

import DCOracle as DB
dbh = DB.Connect( connect_string )
query = dbh.cursor()
query.execute( query_string )

For PostgreSQL, you only need to write

import pg as DB

Now, it is common that Python provides a wrapper module, like the os
module or the anydbm module. However, there was no demand for such a
wrapper in Python, yet.

Regards,
Martin


From gustav at morpheus.demon.co.uk  Sat Dec 16 17:38:51 2000
From: gustav at morpheus.demon.co.uk (Paul Moore)
Date: Sat, 16 Dec 2000 22:38:51 +0000
Subject: Loading startup file question
References: <Pine.LNX.4.10.10012161137360.655-100000@kev.ioc.ee>
Message-ID: <nkrn3tkcja891l8scji8vjcpenlt08jig0@4ax.com>

On Sat, 16 Dec 2000 11:39:27 +0200, Pearu Peterson <pearu at cens.ioc.ee>
wrote:
>sh> python -i -c 'from Numeric import *'
>
>But then python does not load $PYTHONSTARTUP.
>
>So, my question is: Is there a simple way to force python to load startup
>file when both -i and -c option are used?

It's a bit ugly, but there is always (on one line)

python -i -c "import os; execfile(os.environ['PYTHONSTARTUP'])"
          -c 'from Numeric import *'

Paul.


From ccrain at austin.rr.com  Mon Dec 18 10:03:56 2000
From: ccrain at austin.rr.com (Charles Crain)
Date: Mon, 18 Dec 2000 09:03:56 -0600
Subject: another callback question
Message-ID: <t3s9sn46q9c196@corp.supernews.com>

Thanks to everyone who replied to my question about the currentThread()
warning.  I have another, more involved question this time.

Again, I am attempting to write a .pyd wrapper for a C library that uses
callbacks.  I want to propagate the callbacks into Python.  I have gotten as
far as creating a new PyThreadState for the callback thread, then using
PyEval_AcquireThread()/PyEval_ReleaseThread() within the callbacks.  My
problem, however, is this: the callbacks are not always invoked from this
"callback thread."  Sometimes, the callbacks are invoked SYNCHRONOUSLY from
within calls intp this C library.

So for instance, I may have a python-exported function call a function in
this C API, which, from within the function, invokes one of the callbacks.
The problem here is that when I get to the callback itself, since it was
called from within the interpreter, I already have a PyThreadState attached
to the thread, and PyEval_AcquireThread() blows up with a fatal error.  And
I can't use PyThreadState_Get() as a check, because this call will die with
a fatal error if the current thread state is NULL.  So far, I have only come
up with two possible solutions to this, neither of them very elegant.

1.  Given that I know all the C API functions that might invoke callbacks
synchonously (which I don't, but I can find out by experimentation), I can
do a Py_BEGIN_ALLOW_THREADS before the call and a Py_END_ALLOW_THREADS
after.  Since Py_BEGIN_ALLOW_THREADS, among other things, sets the current
PyThreadState to NULL (right?), the subsequent call to
PyEval_AcquireThreads() should work.  I have had limited success with this
technique however, since I have yet to determine all the places that I need
to do this.

2.  Instead of saying PyEval_AcquireThread()/PyEval_ReleaseThread(), I
instead do the following:

my_callback()
{
PyEval_AcquireLock();
pTempThread = PyThreadState_Swap(pCallbackThread);

// Do stuff, including the call to PyEval_CallObject()

pCallbackThread = PyThreadState_Swap(pTempThread);
PyEval_ReleaseLock();
}

That seems to work all the time (on Windows at least).  My questions are:

1.  Is there a better way to solve this problem and/or is one of my current
solutions better than the other?
2.  It occurs to me that, if this callback is called synchonously, the
global interpreter lock might already be held upon entry into the callback.
The dox say that calling PyEval_AcquireLock() from a thread that already
owns the lock will deadlock.  Clearly this doesn't happen on Windows, as my
stuff works, but I am afraid that this fact arises from the fact that
Windows critical sections are re-entrant, whereas similar lock objects on
other OS's may not be.  Does anyone forsee this being a problem were I to
try to port this module to Linux for instance?






From kirschh at lionbioscience.com  Wed Dec 13 02:21:39 2000
From: kirschh at lionbioscience.com (Harald Kirsch)
Date: 13 Dec 2000 08:21:39 +0100
Subject: running a pipe of commands from python
Message-ID: <yv24s08pz0c.fsf@lionsp093.lion-ag.de>

In Tcl I just write

   exec doThis | doThat |toMore | finishUp >output <input 2>errors

In the Python reference manual I find os.exec(), os.fork(), popen2()
etc, but none of these is as easy to use as Tcl's exec. What is
Python's equivalent (and I am sure there must be somehting)?

  Harald Kirsch

-- 
----------------+------------------------------------------------------
Harald Kirsch   | kirschh at lionbioscience.com | "How old is the epsilon?"
LION Bioscience | +49 6221 4038 172          |        -- Paul Erd?s
       *** Please do not send me copies of your posts. ***


From aleaxit at yahoo.com  Fri Dec 15 09:38:29 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 15 Dec 2000 15:38:29 +0100
Subject: xreadlines (was Re: while true: !!!)
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <slrn93g37o.9tj.jepler@potty.housenet> <f3%Z5.146409$U46.4768601@news1.sttls1.wa.home.com> <91a6it02flt@news2.newsguy.com> <slrn93k8sn.lb.jepler@potty.housenet>
Message-ID: <91dae70t7i@news1.newsguy.com>

"Jeff Epler" <jepler at inetnebr.com> wrote in message
news:slrn93k8sn.lb.jepler at potty.housenet...
> On Thu, 14 Dec 2000 11:14:38 +0100, Alex Martelli
>  <aleaxit at yahoo.com> wrote:
> >...if it was able to iterate over already-opened file-like objects,
rather
> >than having to open actual files itself, it WOULD conceptually solve it
> >all.  I wonder how hard it would be to generalize fileinput in this
way --
> >probably not very (the in-place-rewriting option would be incompatible
> >with having already-opened file-like objects in the list, of course).
>
> xreadlines takes a file object, not a filename, as its argument.

Good for it (what IS it?-), but I was talking about the fileinput
module; I thought it would be nice for the user to generalize the
functionality (and enhance the speed) of fileinput, rather than
introducing a completely new and separate module.


> >If [using readline] was a performance problem, it could of course
> >also be fixed in a future fileinput version without changing code that
> >uses it (again, in-place-rewriting would probably have to inhibit the
> >optimization, although that isn't entirely clear).
>
> xreadlines internally uses fp.readlines(sizehint), so
> for line in xreadlines.xreadlines(fp): pass
> is just about as fast as
> for line in fp.readlines(): pass
> except that it will not require storage for the whole file.

Now *THAT* is interesting indeed!  The closest I got was about
a 2:1 ratio, in the LinesOf class which I posted today -- using
the same technique.  I wonder what optimizations I missed that
would allow twice-as-good performance as that!  Do tell, it's
always nice to learn little (or big) tricks like that...!


Alex





From lg at rgz.ru  Mon Dec 25 06:13:26 2000
From: lg at rgz.ru (lg at rgz.ru)
Date: 25 Dec 2000 14:03:26 +0250
Subject: is there any troubles with select.select in python
Message-ID: <3a47297e@news.edunet.ru>

subj?
i've got python script with long time select.select waiting with no timeout
and in half a hour or may be hour .. python exited on signal 11.
gdb shows the top of stack - 0x80b1649 in select_select (self=0x0, args=0x829c4cc) at ./selectmodule.c:177

-- 
zev


From pearu at cens.ioc.ee  Sun Dec 31 02:26:40 2000
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Sun, 31 Dec 2000 09:26:40 +0200
Subject: Solution: Re: Lazy Python usage (hack request) (fwd)
Message-ID: <Pine.LNX.4.10.10012310924140.14042-100000@kev.ioc.ee>

For some reasons this message did not show up in c.l.p.
But here it is.

---------- Forwarded message ----------
Date: Sun, 31 Dec 2000 02:56:00 +0200
From: Pearu Peterson <pearu at cens.ioc.ee>
Newsgroups: comp.lang.python
Subject: Solution: Re: Lazy Python usage (hack request)


Hi again,

I have implemented a Python C/API module `lazy' that you can use to put a
Python dictionary to a "lazy" mode.
You can find the lazy module in

http://cens.ioc.ee/~pearu/misc/lazy-0.1.tar.gz

Here is a Python session showing lazy in action:

  >>> import lazy
  >>> d = {'a':4}
  >>> lazy.set(d,lambda k:3)              # set d to lazy mode
  >>> print d
  {'_lazy_cb': <function <lambda> at 0x8112e24>, 'a': 4}
  >>> d['b']
  3
  >>> print d
  {'_lazy_cb': <function <lambda> at 0x8112e24>, 'b': 3, 'a': 4}
  >>> del d['_lazy_cb']                   # disable lazy mode
  >>> d['c']
  Traceback (most recent call last):
    File "<stdin>", line 1, in ?
    KeyError: c

And another example:

>>> import lazy
>>> def fun(k):
...     print 'Creating variable %s'%(`k`)
...     return k
... 
>>> lazy.set(locals(),fun)
>>> print a+b+c
Creating variable 'a'
Creating variable 'b'
Creating variable 'c'
abc
>>> print lazy.set.__doc__
lazy.set(dict,func) - set dictionary dict to a "lazy" mode with
a callback function func.

"lazy" mode means that the following statement
    dict[<name>] = func(<name>)
is executed whenever dict has no key <name>.

WARNING: if dict is set to "lazy" mode, dict.has_key(<name>)
returns always 1 and dict[<name>] is set to func(<name>).

To disable the "lazy" mode for dict, do
    del dict['_lazy_cb']


Be lazy and have fun!-)

Pearu




From aleaxit at yahoo.com  Fri Dec 22 03:06:46 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 22 Dec 2000 09:06:46 +0100
Subject: saving dictionaries
References: <3A4300C7.FA4192AF@wielun.dhs.org>
Message-ID: <91v23201ph8@news2.newsguy.com>

"QdlatY" <qdlaty at wielun.dhs.org> wrote in message
news:3A4300C7.FA4192AF at wielun.dhs.org...
> Hello!
>
> How to easily save and load contents of dictionaries (2 elements + key)
> to file?

The standard module pickle (and its speedier version cPickle) is a
good solution.  You may also want to look into shelve and anydbm,
which do far more than just 'load and save' but are still connected
to solving persistency issues.

Here's a simple script showing off pickle use: if you run it with
arguments, as in:

    python dipik.py uno one due two tre three

it will build (and save to file 'dipik.pickled') a dictionary
{'uno': 'one', 'due': 'two', 'tre': 'three'}.

If you then call it without arguments, it will load the dict
from that same file and display it.  It's intended to be a
very simple and purely didactic script, of course -- nothing
complicated or 'advanced' in it!

-- cut dipik.py
import sys, pickle

filename = 'dipik.pickled'

numargs = len(sys.argv)

if numargs==1:
    # no args: load and display the dictionary
    #     that was last saved to dipik.pickled
    dict = pickle.load(open(filename,"rb"))
    keys = dict.keys()
    keys.sort()
    for key in keys:
        print "%20s %20s" % (key, dict[key])
else:
    # args: build a dictionary using alternate
    #     args as keys and values, then save it
    dict = {}
    if numargs%2 == 0:
        # odd number of args, ignore last one
        numargs -= 1
    for i in range(1,numargs,2):
        key, value = sys.argv[i:i+2]
        dict[key] = value
    pickle.dump(dict, open(filename,"wb"))
-- end dipik.py


Alex





From Blair.Lowe at compeng.net  Fri Dec  8 16:40:56 2000
From: Blair.Lowe at compeng.net (Blair Lowe)
Date: Fri, 8 Dec 2000 14:40:56 -0700
Subject: SSL how to?
In-Reply-To: <yZ3Y5.3016$9B.89007@news1.oke.nextra.no>
References: <yZ3Y5.3016$9B.89007@news1.oke.nextra.no>
Message-ID: <v04220800b65703c65420@[192.168.11.33]>

Python 2.0 is supposed to support it.

I have version 1.5 working with the M2Crypto library. Ken Kinder 
pointed me to http://www.vex.net/parnassus/apyllo.py?find=ssl to get 
it.

For M2Crypto, you need some other stuff too (openssl, swig, and a 
couple of other things). I got the other bits from rpm's on the net 
(only if you have a system that can use ropm's like LINUX), otherwise 
it is difficult to set up since you may need to do the old tar -xvf 
<tarball>, ./configure, make, make install dance to get the 
supporting packages.

In python 1.5, after installing M2Crypto, everything is quite simple.

I assume that python 2.0 is easier, but I have not got there yet.

At 12:58 +0100 2000/12/08, Syver Enstad wrote:
>Anyone know if SSL is painless to use under python. Does httplib support it?

Hope this helps,
Blair.
Computer Engineering Inc. http://www.compeng.net
Phone: 780 499 5687 (9 - 5 MST) Fax:   780 435 0693 (24 Hours)



From jbauer at rubic.com  Mon Dec 11 10:28:44 2000
From: jbauer at rubic.com (Jeff Bauer)
Date: Mon, 11 Dec 2000 09:28:44 -0600
Subject: MySQL + Python + Win32
Message-ID: <3A34F2AC.E069C506@rubic.com>

Recently a friend with no Python nor MySQL experience
wanted to try using both on a Win32 platform.  After
telling him it should work fine I decided to check it
out -- since almost all my experience with MySQL has 
been on Linux/Solaris platforms.  The following is a
brief portion of the reply I sent him.

-------------------------------------------------------

This looks pretty straightforward.  I installed Python2.0, 
MySQL 3.23.28-gamma, and the MySQL-python-0.3.0-win32-2.zip
files onto an NT4.0 system.

I used Gerhard's win32 MySQLdb Python interface, compiled
from Andy Dustman's sources:

  http://home.t-online.de/home/err666/

Don't forget to run "python setup.py install" from the
MySQL-python-0.3.0 directory.

First, I test it against a known Unix database server.

D:\jbauer\tmp>python
Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
>>> import MySQLdb
>>> connection = MySQLdb.connect(host='domitian', db='fhc', user='jbauer', passwd='XXX')
>>> c = connection.cursor()
>>> c.execute("select ID, LNAME from CLIENT")
450L
>>> for r in c.fetchall():
...     print r[0], r[1]
...
024123 Adelson
019868 Akers
026959 Akers
011502 Ambrose
032352 Anderson
  .
  .
  .

Next, I try connecting to the local NT machine:

>>> ntconn = MySQLdb.connect(host='albatross', db='mysql', user='root')
>>> c = ntconn.cursor()
>>> c.execute("select * from User")
4L
>>> for row in c.fetchall():
...     print row
...
('localhost', 'root', '', 'Y', 'Y', 'Y', 'Y', 'Y', 'Y', 'Y', 'Y', 'Y',
'Y', 'Y', 'Y', 'Y', 'Y')
('%', '', '', 'N', 'N', 'N', 'N', 'N', 'N', 'N', 'N', 'N', 'N', 'N',
'N', 'N', 'N')
('localhost', '', '', 'Y', 'Y', 'Y', 'Y', 'Y', 'Y', 'Y', 'Y', 'Y', 'Y',
'Y', 'Y', 'Y', 'Y')
('%', 'root', '', 'Y', 'Y', 'Y', 'Y', 'Y', 'Y', 'Y', 'Y', 'Y', 'Y', 'Y',
'Y', 'Y', 'Y')
>>>

Everything looks peachy. Good Luck!

-Jeff



From fredrik at effbot.org  Fri Dec  8 10:31:56 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 08 Dec 2000 15:31:56 GMT
Subject: Q for regexp experts
References: <yv2aea77xw2.fsf@lionsp093.lion-ag.de>
Message-ID: <M97Y5.2589$O%.399681@newsc.telia.net>

Harald Kirsch wrote:
> Even with perl-compatible regular expressions it seems to be
> impossible to specify a pattern which matches
>
>    `the longest substring not containing separator' SEP
>
> Put another way, from a given starting point, the pattern shall
> match everything until just before the next occurence of SEP.

why not just use non-greedy repeat and a group?

    m = re.match("(.*?)SEP", text)
    print m.group(1)

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From pbjorn at uswest.net  Thu Dec 28 00:57:30 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Wed, 27 Dec 2000 22:57:30 -0700
Subject: Some Python 2.1 ideas
References: <mailman.977590932.19335.python-list@python.org>
Message-ID: <3A4AD64A.48761A0F@uswest.net>

Bob Alexander wrote:

> Here are some thoughts on possible enhancements to Python. If discussion
> shows any positive enthusiasm maybe they could turn into PEPs.
> [snip]

> ----
> String method to remove trailing newline

Patch number 103029 on sourceforge.

-- bjorn




From Edgar at EdgarSchwarz.de  Sat Dec 23 04:05:04 2000
From: Edgar at EdgarSchwarz.de (Edgar at EdgarSchwarz.de)
Date: 23 Dec 2000 09:05:04 GMT
Subject: mod_python.dll needed
Message-ID: <921ps0$l2e$1@riker.addcom.de>

Hi,
some days ago I installed Apache 1.3.14 on my W98. Python 1.5.2 was already there.
Then I got mod_python-2.7.1 which tells me I still need mod_python.dll. But I cant't
find it on the webpage. And I can't compile it because my Windows is for the children
to play and doesn't have a C compiler.
So, where can I find mod_python.dll ?

Merry Christmas, Edgar

P.S. I only learned about Python two months ago and loved it at once. It reminds me
a little bit of my other favorite environment.
<OT> Native Oberon System3 http://www.oberon.ethz.ch/native/ </OT>

-- 
edgar at edgarschwarz.de                    http://www.edgarschwarz.de
*          DOSenfreie Zone.        Running Native Oberon.         *
Make it as simple as possible, but not simpler.     Albert Einstein


From aleaxit at yahoo.com  Fri Dec  8 04:09:14 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 8 Dec 2000 10:09:14 +0100
Subject: UnboundLocalError: Local variable 'Tee' referenced before assignment
References: <90mru0$cv4$1@nnrp1.deja.com> <%FFX5.2390$O%.371404@newsc.telia.net> <90osgm$vs$1@nnrp1.deja.com> <3A30285A.46813B28@cosc.canterbury.ac.nz>
Message-ID: <90q8h80rko@news2.newsguy.com>

"Greg Ewing" <greg at cosc.canterbury.ac.nz> wrote in message
news:3A30285A.46813B28 at cosc.canterbury.ac.nz...
    [snip]
> As a final refinement, you may want to enclose the
> operation in a try...finally, to make sure that the
> old value of stdout doesn't get lost if there is
> an exception raised while doing your stuff:
>
>   old_stdout = sys.stdout
>   sys.stdout = Tee( reportfile, sys.stdout )
>   try:
>     # do some stuff
>   finally:
>     sys.stdout = old_stdout

Very good, but, not _quite_ final... while in some versions
of Python object's finalizers are called just as soon as an
object's reference count goes to zero (and such versions
include the current 'CPython'), this is not a guaranteed
characteristic of the language -- in other versions (such as
Jython and Python .NET), finalization might be delayed or
even non-guaranteed.

If one can, therefore, it may be considered preferable to
finalize one's objects explicitly if that finalization matters
for other purposes than just making available the memory
the objects use -- closing files and tearing down database
connections being typical examples.

It may therefore be even better to change this finally
clause into a more elaborate one, for compatibility with
all Python dialects...:

    finally:
        try: sys.stdout.close()
        finally: sys.stdout = old_stdout

or even more, if one wants to hide exceptions that would
come if the Tee class doesn't define a close method, e.g:
    finally:
        try:
            closer = getattr(sys.stdout,'close',None)
            if not closer is None: closer()
        finally: sys.stdout = old_stdout
or
    finally:
        try:
            try: sys.stdout.close()
            except AttributeError: pass
        finally: sys.stdout = old_stdout


On the other hand, on the principle "do the simplest
thing you can possibly get away with", the much simpler
and cleaner form you use could be kept today, maybe
with just a comment about explicit finalization needing
to be added if this is ever ported to some Python
version without guarantee of immediate finalization
on reference count going to zero -- just to alert any
possibly-unaware-of-this-issue future maintainer!


Alex





From rob at ZOOstation.cc  Wed Dec  6 01:32:17 2000
From: rob at ZOOstation.cc (Rob Brown-Bayliss)
Date: 05 Dec 2000 18:32:17 -1200
Subject: [pygtk] Python and Orbit Examples
In-Reply-To: <Pine.LNX.4.21.0012050828550.24722-100000@tupelo.daa.com.au>
Message-ID: <200012060632.eB66WIK27614@ZOOstation.cc>

>or getting it to work with gnome-python, that is a bit more difficult.

Sounds a little ominous...

What I am doing (or planning at this stage) is to write a POS and Stock 
management system for a small retail chain.

Would I be better of giving Pyhton a miss, or ditching the Corba idea? 

I have very little knowledge of linux programming, though I do have 
years of basic programming on the Amiga (like 18 I think if include vic 20 and c64)
and a few months VC++ on windows.  I have used Linux for 18 months now, and played
a bit with C and a very very little with Python. 

I am confident I can tackle the task in either language, but I know python will be
quicker to develop with.

I had looked to corba as a means of communicating between the stores, I could always
use some form of file transefer, but it's not as elegant or nearly as real time.




-- 

  Rob Brown-Bayliss
 ---======o======---

  www.ZOOstation.cc



From jeremy.lamplough at motorola.com  Tue Dec  5 16:44:33 2000
From: jeremy.lamplough at motorola.com (Jeremy Lamplough)
Date: Wed, 06 Dec 2000 08:44:33 +1100
Subject: urllib2, proxy, basic authentication
References: <90hk8k$fqh$1@bob.news.rcn.net>
Message-ID: <3A2D61C1.9182431D@motorola.com>

try setting the proxy host prior to importing the other modules.

import os, sys
os.environ['http_proxy'] = 'http://[PROXY_HOST]:5865'
import urllib2
from urllib2 import urlopen

The way i'm doing basic stuff, with NO authentication is (to download a
stock quote)

import urllib
import sys
import string
 
http_proxy = 'http://proxy.mycom...'
opener = urllib.FancyURLopener({'http' : http_proxy})
stocks = sys.argv[1:]
 
for stock in stocks:
   file = opener.open("http://finance.yahoo.com/d/quotes.csv?s=" + stock
+ "&f=sl1d1t1c1ohgv&e=.csv")
   print string.strip(file.read())


hope this helps,

jeremy


Albert Chin-A-Young wrote:
> 
> I'm trying to use urlopen in urllib2.py with an HTTP proxy to connect
> to a site requiring BASIC authentication. My code follows:
> 
>   import os, sys, urllib2
>   from urllib2 import urlopen
> 
>   os.environ['http_proxy'] = 'http://[PROXY_HOST]:5865'
> 
>   authinfo = urllib2.HTTPBasicAuthHandler ()
>   authinfo.add_password ('[REALM]', 'http://[URL]/', '[LOGIN]',
>   '[PASSWORD]')
> 
>   opener = urllib2.build_opener (authinfo)
>   urllib2.install_opener (opener)
> 
>   url = urlopen ('http://[URL]/')
>   print url.info ()
>   url.close ()
> 
> This is not working. Does urllib2 support this configuration?
> 
>     File "/tmp/a.py", line 15, in ?
>       url = urlopen ('http://updates.thewrittenword.com/')
>     File "/opt/TWWfsw/pkgutils12/lib/python20/lib/python2.0/urllib2.py",
>   line 137, in urlopen
>       return _opener.open(url, data)
>     File "/opt/TWWfsw/pkgutils12/lib/python20/lib/python2.0/urllib2.py",
>   line 325, in open
>       '_open', req)
>     File "/opt/TWWfsw/pkgutils12/lib/python20/lib/python2.0/urllib2.py",
>   line 304, in _call_chain
>       result = func(*args)
>     File "/opt/TWWfsw/pkgutils12/lib/python20/lib/python2.0/urllib2.py",
>   line 764, in http_open
>       return self.parent.error('http', req, fp, code, msg, hdrs)
>     File "/opt/TWWfsw/pkgutils12/lib/python20/lib/python2.0/urllib2.py",
>   line 351, in error
>       return self._call_chain(*args)
>     File "/opt/TWWfsw/pkgutils12/lib/python20/lib/python2.0/urllib2.py",
>   line 304, in _call_chain
>       result = func(*args)
>     File "/opt/TWWfsw/pkgutils12/lib/python20/lib/python2.0/urllib2.py",
>   line 430, in http_error_default
>       raise HTTPError(req.get_full_url(), code, msg, hdrs, fp)
>   urllib2.HTTPError: HTTP Error 401: Authorization Required
> 
> --
> albert chin (china @at@ thewrittenword .dot. com)


From mikael at isy.liu.se  Mon Dec 11 05:22:02 2000
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Mon, 11 Dec 2000 11:22:02 +0100 (MET)
Subject: variable variables
In-Reply-To: <Pine.LNX.4.21.0012111052110.5044-100000@relax.dk>
Message-ID: <XFMail.001211112202.mikael@isy.liu.se>

On 11-Dec-00 pyth at relax.dk wrote:
 >  Is it possible to make variable variables? :o)

I guess you want to be able to have multiple references to the same
variable, and that you want to be able to change the value of that
variable in such a way that all references refer to the same new value.

Others may correct me, but you need references to a reference. I would 
use an instace of a class that encapsule the data. Then you can have 
multiple references to an instance, and change the data inside the 
instance.

HTH 

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    11-Dec-00
Time:    11:13:57

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

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



From support at internetdiscovery.com  Thu Dec 21 14:17:27 2000
From: support at internetdiscovery.com (Mike Clarkson)
Date: Thu, 21 Dec 2000 19:17:27 GMT
Subject: WWW_URLEcho with dde
References: <3a32b7e6.8685100@24.0.228.33> <9100sp$a0b$1@nnrp1.deja.com>
Message-ID: <3a41090f.5947518@nntp.ix.netcom.com>

On Sun, 10 Dec 2000 13:36:24 GMT, Roger Abbott <rogerha at aol.com>
wrote:

>In article <3a32b7e6.8685100 at 24.0.228.33>,
>  support @ internetdiscovery.com (Mike Clarkson) wrote:
>> Has anyone used the dde module in Python to register a URLEcho from
>> netscape or Internet Explorer? Looking at the source code to dde I see
>> comments about Netscape's use of poke, which indicates the author
>> may have been playing with URLEcho.
>>
>> I'd like to start a dde server and register it with netscape to send
>> me an echo when it loads a web page. Has anyone done this?
>
>Yes, in VB. The web site below has a free WebSpy program which does it.
>It comes with source code, I am sure you can figure out how it works.

Thanks, but I was looking for a solution in Python with dde.py.

Here is what I tried: I can see the messages coming back from
Netscape, so the Echo has been registered, but Python takes no action.

Mike.

import win32ui
from pywin.mfc import object
import dde

class MySystemTopic(object.Object):
	def __init__(self):
		object.Object.__init__(self,
dde.CreateServerSystemTopic())
	def Exec(self, cmd):
		print "System Topic asked to exec", cmd

class MyPokeTopic(object.Object):
	def __init__(self, topicName):
		object.Object.__init__(self,
dde.CreateTopic(topicName))
	def Poke(self, cmd):
		print "Other Topic asked to poke ", cmd

class MyRequestTopic(object.Object):
	def __init__(self, topicName):
		object.Object.__init__(self,
dde.CreateTopic(topicName))
	def Request(self, cmd):
		print "Other Topic asked to poke ", cmd

server = dde.CreateServer()
server.AddTopic(MySystemTopic())
server.AddTopic(MyPokeTopic("WWW_URLEcho"))
server.AddTopic(MyRequestTopic("WWW_URLEcho"))
server.Create('MyApp')
print 'created myapp'

conversation = dde.CreateConversation(server)

conversation.ConnectTo("Netscape", "WWW_RegisterURLEcho")

s = '"MyApp"'
sl = conversation.Poke(s)
print 'Retval is "%s"'% s

del conversation

while 1:
	win32ui.PumpWaitingMessages(0, -1)



From brian_zhouNOSPAM at agilent.com  Sat Dec  9 12:35:33 2000
From: brian_zhouNOSPAM at agilent.com (Brian Zhou)
Date: Sat, 9 Dec 2000 09:35:33 -0800
Subject: Unit Testing tool
References: <90s69s$f85$1@troll.powertech.no>
Message-ID: <976383323.699817@emperor.labs.agilent.com>

ftp://ftp.ddj.com/1999/1999_04/jpython.zip
It's quite simple and easily understood, see Dr. Dobb's Journal issue for
detail. Although it was written for jpython, I think you can use it in
CPython as well.

-Brian

"Syver Enstad" <syver at NOSPAMcyberwatcher.com> wrote in message
news:90s69s$f85$1 at troll.powertech.no...
> What is the Python unit testing tool preferred by readers/writers from
> comp.lang.python?
>
>
>




From pblanchette at pixelsystems.com  Tue Dec  5 16:57:03 2000
From: pblanchette at pixelsystems.com (Patrick Blanchette)
Date: Tue, 05 Dec 2000 21:57:03 GMT
Subject: win32security.AdjustTokenPrivileges
References: <3A2D378E.7050504@pixelsystems.com> <dqbX5.38997$hk4.1468335@news.infostrada.it>
Message-ID: <3A2D649E.1090609@pixelsystems.com>

In fact, the declaration of OpenProcessToken in the Python extensions 
for Windows is :
int = OpenProcessToken( processHandle, desiredAccess ).
I suppose the integer returns is the token handle.

Regards

Carlo Bertuccini wrote:

> Hi Patrick,
> 
>> hProcess = GetCurrentProcess()
>> processToken = OpenProcessToken(hProcess, TOKEN_ADJUST_PRIVILEGES |
>> TOKEN_QUERY)
> 
> 
> There is something strange in your code ... this is the declaration of
> OpenProcessToken:
> 
> BOOL OpenProcessToken(
>     HANDLE ProcessHandle, // handle to process
>     DWORD DesiredAccess, // desired access to process
>     PHANDLE TokenHandle  // pointer to handle of open access token
>    );
> 
> The "ProcessToken" is not the Token Handle, but a boolean value: the result
> of the function.
> The AdjustTokenPrivileges needs the real Token Handle in order to set the
> new privileges.
> 
> Regards,
> --
> - Carlo -
> c.bertu at libero.it
> 
> 
> 
> 
> 
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20001205/b674a241/attachment.html>

From slhath at home.com  Wed Dec 27 19:21:47 2000
From: slhath at home.com (Scott Hathaway)
Date: Thu, 28 Dec 2000 00:21:47 GMT
Subject: Python vs PHP (& Java?)
References: <mailman.977897649.4877.python-list@python.org> <92cdn00hck@news2.newsguy.com> <87k88mxcf4.fsf@eugene.prosa.it> <D722E432FF377E01.657BAF2ECAA117A3.909A66D671189EE3@lp.airnews.net>
Message-ID: <vIv26.94619$yR4.2590443@news1.rdc1.tx.home.com>

It should be noted that PHP already support a command line interface.  I use
it heavily in this manner for doing scripting on Windows NT.

Also, I have fallen in love with python, but I would not use it for
server-side web scripting.  PHP is very easy and extremely powerful for this
task.  Why not use both?

Scott


"Cameron Laird" <claird at starbase.neosoft.com> wrote in message
news:D722E432FF377E01.657BAF2ECAA117A3.909A66D671189EE3 at lp.airnews.net...
> In article <87k88mxcf4.fsf at eugene.prosa.it>,
> >"Alex Martelli" <aleaxit at yahoo.com> writes:
> David N. Welton <davidw at prosa.it> wrote:
> >
> >> So, _what_ "much functionality" of PHP couldn't be "replaced" by
> >> Python solutions?
> >
> >Personally, I use Tcl for this sort of thing, but I think that Python
> >is pretty good too, for a very important reason:
> >
> >PHP is used on the web and only on the web.  Python, Tcl, Java, and
> >company are languages that you can take and reuse elsewhere.
> .
> .
> .
> Incidentally, the PHPers certainly are aware of
> this situation, and it's likely that PHP will
> become available through command-line and/or
> MySQL stored-procedure applications, sometime
> in 2001.  I expect to write more on this, pro-
> bably in January.
> --
>
> Cameron Laird <claird at NeoSoft.com>
> Business:  http://www.Phaseit.net
> Personal:  http://starbase.neosoft.com/~claird/home.html




From cnelson9 at gte.net  Tue Dec 19 15:03:46 2000
From: cnelson9 at gte.net (Clifford J. Nelson)
Date: Tue, 19 Dec 2000 20:03:46 GMT
Subject: Python in "math through programming" curriculum
References: <2g2t3t06r2agrrjmfnr41eu4lc2u39a8qi@4ax.com> <3A3EFF47.67DE69CF@gte.net> <513v3toslkjkggh9oumlf20qc9bi1gdtif@4ax.com>
Message-ID: <3A3FC677.FDB19937@gte.net>


Kirby Urner wrote:
> 
> "Clifford J. Nelson" <cnelson9 at gte.net> wrote:
> 
> >I don't think Python should be a standard at all, because of
> >my experience with it on the Mac. Keep the options open for
> >computer programming languages.
> >

> I'm not sure I need to attach much weight to any of your
> opinions.  ...
> 
> Kirby

I'm sure I don't attach ANY weight to any of your opinions.
I've seen your web sites about Synergetics and programming!

  Cliff Nelson


From aleaxit at yahoo.com  Fri Dec 15 08:02:48 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 15 Dec 2000 14:02:48 +0100
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <tcmay-FCA22F.14073713122000@news.got.net> <918v7a02scn@news1.newsguy.com> <tcmay-04E80C.16181313122000@news.got.net> <91a1ed02a4j@news2.newsguy.com> <3A389B1A.4F33FFD4@cfmu.eurocontrol.be> <91ae8s02llf@news2.newsguy.com> <3A397778.2E6D84D3@cosc.canterbury.ac.nz> <M0f_5.27169$x6.14244092@news2.rdc2.tx.home.com>
Message-ID: <91d4qs0lu7@news1.newsguy.com>

"Rainer Deyke" <root at rainerdeyke.com> wrote in message
news:M0f_5.27169$x6.14244092 at news2.rdc2.tx.home.com...
> "Greg Ewing" <greg at cosc.canterbury.ac.nz> wrote in message
> news:3A397778.2E6D84D3 at cosc.canterbury.ac.nz...
> > To mention something vaguely Python-relevant:
> > if Python had a convenient syntax for declaring
> > singletons, it could be a pretty good language for
> > Interactive Fiction...
>
> Meta-classes can turn a class statement into a singleton:
>
> class A(Singleton):
>   ...
>
> (where Singleton is an instance of the (meta-)class that implements
> singletons, not listed here because I haven't written or found it yet)
>
> It would be nice if the Singleton above was in module __builtin__ or at
> least somewhere in the standard library.

I beg to differ -- it's *great* that no such metaclass is anywhere
handy: the last thing we should do is encourage wider use of a
flawed design pattern.  It's bad enough that Singleton has such
a catchy name... it already makes it too-widely used!-)  (Please
see the 'direct printer access' thread for a post of mine, from a
few minutes ago, addressing this issue more directly).


Alex





From mbel44 at dial.pipex.net  Fri Dec 15 05:44:57 2000
From: mbel44 at dial.pipex.net (Toby Dickenson)
Date: Fri, 15 Dec 2000 10:44:57 +0000
Subject: Smalltalk and Python
References: <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <91a4ht02e59@news2.newsguy.com> <0j8_5.25310$x6.13684532@news2.rdc2.tx.home.com> <91bb6402lvp@news1.newsguy.com>
Message-ID: <1etj3tom2b3k07g2htmo9o315r5bn6vh8q@4ax.com>

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

>    def __init__(self):
>        self.a = A()
>        self.a.g = self.g


Note quite - self.a.g is now a circular reference.

Toby Dickenson
tdickenson at geminidataloggers.com


From sdm7g at virginia.edu  Sat Dec 23 00:52:06 2000
From: sdm7g at virginia.edu (Steven D. Majewski)
Date: Sat, 23 Dec 2000 00:52:06 -0500 (EST)
Subject: Python interpreter in python?
In-Reply-To: <91vuan$l97@news.Informatik.Uni-Oldenburg.DE>
References: <mailman.977335880.30954.python-list@python.org> <91vuan$l97@news.Informatik.Uni-Oldenburg.DE>
Message-ID: <Pine.NXT.4.21.0012230026080.233-100000@localhost.virginia.edu>


On 22 Dec 2000 jnix at medi.physik.uni-oldenburg.de wrote:

> This opens a huge number of choices -  [ ... ]
> 
> But I would certainly like to see a Python VM written in Modula-3..

While on the topic of "blue sky" projects:

 Now that I've been playing with the python objective-c bridge module
on MacOSX, I can see some real advantages to reimplementing part of
python in objective-c. The objc-runtime has some of the same sort of
plumbing that python's runtime in c does. Objective C code doesn't
have to be "wrapped"  to be accessable to python.  Well -- it's
actually dynamically wrapped by the runtime, but you don't have to
do any special codeing to make objc libraries python callable.  
I think you have the same sort of thing with COM classes on PythonWin,
except objc is even easier. ( Step 1 of several C++ GUI projects always
seems to be extending C++ with more dynamic, introspective features
to make it more like objc-c and thus usable for dynamically adding
features and widgets to a GUI. )  
 However, to get all of the advantages, you would have to incompatibly
shake up the python object hierarchy. ( i.e. one of the advantages would
be to solve the extension class problem by making most python objects
classes -- if not python classes, at least obj-c classes. 

 -- Steve Majewski <sdm7g at Virginia.EDU> 




From simon108 at my-deja.com  Tue Dec 12 17:40:11 2000
From: simon108 at my-deja.com (Simon)
Date: Tue, 12 Dec 2000 22:40:11 GMT
Subject: Script vs Program
References: <y39V5.524$OC.159868@typhoon.aracnet.com> <bHnV5.96503$U46.3094705@news1.sttls1.wa.home.com> <9069v6$ifi$1@nnrp1.deja.com> <krFV5.102015$U46.3205226@news1.sttls1.wa.home.com>
Message-ID: <9169g9$87o$1@nnrp1.deja.com>

In article <krFV5.102015$U46.3205226 at news1.sttls1.wa.home.com>,
  "Greg Jorgensen" <gregj at pobox.com> wrote:
> <simon108 at my-deja.com> wrote in message news:9069v6
$ifi$1 at nnrp1.deja.com...
>
> > Java is not called scripting language because you have a compiler
> > that "compiles" your script, therefore entering in the category of
> > compile languages instead of interpreted as Python.
> > In any case you can call your Python code scripts programs if they
> > actually perform any type of i/o .
>
> I'm not sure what your point is. Java source code is compiled into
byte
> codes that are executed by a JVM. Python works the same way, but the
compile
> step is generally invisible.
>
> I don't know what I/O has to do with anything. Are my C programs
scripts if
> they don't perform I/O?
>
> --
> Greg Jorgensen
> Deschooling Society
> Portland, Oregon, USA
> gregj at pobox.com
>
>

--
 "In Python, like in C, procedures are just functions that don't return
a value. In fact, technically speaking, procedures do return a value,
albeit a rather boring one. This value is called None (it's a built-in
name). Writing the value None is normally suppressed by the interpreter
if it would be the only value written."
By this quote taken from the Python tutorial I will try to state why I
consider I/O a part in writing a program in Python. If you write a set
of functions and variable declarations in Python , you are not really
writing a program but defining resources for your program(s) to use. In
fact there is no need for a program if you directly use the command
interpreter and obtain the results by manipulating your pre-defined
functions and variables (although it may be argued that the fact that
the input of a set of instructions to obtain the results in the
interpreter might be called a program).
Now I would say that to write a function by itself is not a program, to
use the function through the Python interpreter is not a program, A
program would be a set of instructions to get the values, manipulate
those functions and return a result and I will add to this, that this
set of instructions should have the ability to be recalled and executed
automatically(by means of the interpreter or by executing its object
code). In that sense I am saying that if you replicate in a file those
instructions you performed in the interpreter manually and are able to
execute them automatically from that file (being the input the
definitions of the variables and functions previously defined, or the
users direct input, and the output the results of such manipulation)
then you have a strict definition of a program.
Now I am not going to argue whether the functions or procedures
returning a "none" value should be consider programs but I would say
that you can call the file with the instructions to manipulate the
functions and variables and that return a result or output a script but
the file with the variables and function definition by itself(if they
do not produce any results or outputs different than none) that you can
not call that a program.

You can gather all the ingredients to make bread, but without a recipe
you do not have bread, you have ingredients.


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


From aleaxit at yahoo.com  Wed Dec 27 11:33:05 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 27 Dec 2000 17:33:05 +0100
Subject: A Suggestion for Python Dictionary/Class
References: <920fe9$c20$1@hecate.umd.edu> <920m9h011uh@news2.newsguy.com> <92cvte$b0i$1@nnrp1.deja.com>
Message-ID: <92d5k5016bt@news2.newsguy.com>

<billtj at my-deja.com> wrote in message news:92cvte$b0i$1 at nnrp1.deja.com...
> Thanks for all the suggestions in making the Python class more like
> C/C++ struct or class.  However, it looks that the current solution is
> based on a "convention" to derive Python classes from a certain class.

"Convention"???


> I am not too familiar with how the Python interpreter works.  But based
> on my current understanding, the current Python class is implemented
> using dictionary, unlike in C/C++.  Therefore Python's
>
>     var_1.member_1.member_2.member_3
>
> will invoke three dictionary lookups, while C/C++'s

Four, actually: var_1 also has to be looked up (if it's a _local_
variable, then THAT 'look-up' is generally optimized away).

>     var_1.member_1.member_2.member_3
>
> is a (short) constant time operation, after compilation, of course.

Correct.  The compiler knows at compile-time the static types
of all objects involved -- enough of those types to know about
their *data* members, at least, since data can't be virtual in
C++; there may be virtual *bases* involved, which require levels
of indirections, but, if there are, the compiler will know -- it
will take longer, but the appropriate machinecode will have been
placed there by the compiler to minimize the inevitable overhead;
and -- a key idea of C++ -- the whole language is organized so
that you don't have to pay in terms of runtime for features you
are not actually using, at least in theory.


> My question is whether by defining a special class (call is struct?) in
> Python will result in better Python run-time performance.  For example,

Not by itself, no, because Python's typing is "latent", aka dynamic --
which is where so much of the language's power and flexibility come
from.  The compiler cannot determine statically, at compile-time,
what the runtime of, e.g., var_1 will be -- not generally, and, at
this time, it doesn't even *try*.

So, once it has looked-up (or solved more rapidly, if it's a local
variable) the object corresponding to var_1, it will call its
"get-attribute" C function passing it the string "member_1"; this
call, in itself, is an overhead quite comparable to a virtual-function
call in C++ (call it a few machine instructions -- ones which C++
itself is not paying at all when no virtual behavior is involved,
but you can see Python [as, say, Java] as being 'always virtual').

It is then up to the 'type-object' to define its own appropriate
'get-attribute' function that turns the string 'member_1' into a
Python object pointer.  If the object's type for var_1 is 'class
instance', its C-coded 'get-attribute' starts with a dictionary
lookup in the instance's dictionary, then, if that fails, with
one in the class-object's dictionary, then up the 'bases' DAG,
etc, etc.  See Python-2.0\Objects\classobject.c -- there are lots
of steps involved (e.g., checking for a __getattr__ method), but
none of them is complicated.

It's not hard to define a type with fewer abilities, such as
the 'struct' you propose (having it have a special syntax as
you wish would be harder, as you'd have to mess with Python
internals a little bit; but just adding the semantics, a snap).

Actually, since 'struct' is already the name of an important
builtin module (which lets your Python code handle actual
binary-layout structs, just like C or C++ code would), we had
better name it differently -- say, 'bill'.

>     struct MyStruct:
>         def __init__ (self):
>             self.member_1
>             self.member_2
>             self.member_3

That might become something like:

import bill

def MyStruct(v1=None, v2=None, v3=None):
    return bill.bill(member_1=v1, member_2=v2, member_3=v3)

Or, to save some memory in each bill-object, you might want
'bill.struct' to be the factory function for factory objects
that encode all knowledge about how a given object kind can
possibly be; each 'bill instance' object would then have the
overhead of a single pointer to its 'bill-kind' -- it's not
THAT much of an extra saving, and complexity zooms up fast.


> Because it is a "struct" instead of a "class", we cannot add any new
> member.  Can then the Python interpreter be optimized to handle struct's
> so that the run-time performance is better than a class?  Coming from

Nah -- though the bill object itself _might_ save you a few machine
cycles per access over using a class-instance object, deciding to
accord some special object-type a superior status, to the point of
changing the interpreter itself to honor that status, is something
you'll never, but NEVER, get past Guido.  That is how comes Python is
as powerful as it is while remaining simple -- NO ten zillion special
cases for 'convenience', '10% speed boosts', and other such ad-hockeries.


> C/C++ background and never saw the Python C code, I imagine that the

Since you're blessed with a C/C++ background -- why not download
and study the C sources?  They're blessedly CLEAN and SIMPLE -- you'll
wish all C code you ever had to maintain was written to such standards!-)

> interpreter allocates memory space for three labels (which are
> pointers/handles to Python objects?) and during
> interpretation/compilation, Python replaces the references for member_1,
> member_2, and member_3 simply to the pointers to the corresponding
> labels.  (I may not be precise in describing this, but hopefully you get
> the idea).  So no dictionary lookup is involved.

You might choose to do it differently than with Python dictionary
lookups (which are VERY FAST INDEED, as it happens), but you'd
still have (in your bill object) to translate string->offset at
runtime ('interned' strings, probably -- but that's designed to
make them even faster *for dictionary lookup purposes*...!).

If you really want to wring some real speed out of this scheme,
you have to get some measure of type-inference or static typing
into the Python compiler/interpreted; make the compiler able to
do what it already does for local variables, also for attributes
of some specially singled-out kind of object... but first it
has to 'pin down' a certain variable as SURE to refer at all
times to THAT kind of object and no other.  Two pretty hard
steps to get past Guido!-)


Alex





From grey at despair.rpglink.com  Mon Dec 18 14:05:18 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Mon, 18 Dec 2000 19:05:18 -0000
Subject: while true: !!!
References: <B43D149A9AB2D411971300B0D03D7E8B0FBCBB@natasha.auslabs.avaya.com> <v04220801b663769e1800@[192.168.11.33]> <mailman.977147642.18708.python-list@python.org> <91lclt02qr3@news1.newsguy.com>
Message-ID: <slrn93snve.tvg.grey@teleute.rpglink.com>

On Mon, 18 Dec 2000 17:05:51 +0100, Alex Martelli <aleaxit at yahoo.com> wrote:
>I think it would be inconsistent, un-Pythonic, to add a fundamental type for
>'conceptual' reasons, without any practical, measurable benefit.  For
>example, Python does away with the distinction between 'a character' and 'a
>string of length 1' -- a distinction that is religiously maintained in C,
>Pascal, Java, etc, and does have a basis in both 'conceptual' and
>machine-level considerations -- yet the strong conceptual *simplification*
>arising from just doing away with the distinction appears to be quite
>positive.

    I can attest to that.  I have tough times in C, Pascal and other such
languages because of one consideration they make that Perl and Python do not.
Data is data.  Forgiving me my discussion earlier where I said Python doesn't
do that (when compared to Perl) it still does it in all the ways that count.
I don't have to worry about how large or small of a block of space to allocate
to my strings or arrays.  It's a string, not a string[80] that I might one day
want to stuff 81 characters into.  :/


-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From loewis at informatik.hu-berlin.de  Wed Dec 27 20:22:08 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 28 Dec 2000 02:22:08 +0100
Subject: removing directories
References: <Pine.A41.3.96.1001227175036.37924A-100000@aix2.uottawa.ca>
Message-ID: <p6qito5e3wf.fsf@informatik.hu-berlin.de>

<abouf066 at aix2.uottawa.ca> writes:

> 	Is there a way to force removal of directories even if they
> contain files and/or directories. 

It depends on the operating system, but on most systems, you can't
remove a directory if that still contains files. I recommend to remove
the files first, then remove the the directory.

Regards,
Martin



From jasonic at nomadicsltd.com  Sat Dec 30 13:35:34 2000
From: jasonic at nomadicsltd.com (Jason Cunliffe)
Date: Sat, 30 Dec 2000 13:35:34 -0500
Subject: PythonWin prompts for win32api methods
Message-ID: <t4saigog72et1c@corp.supernews.com>

I have been exploring win32api methods using PythonWin.

One of the features I like most about PythonWin is the popup prompts for
objects and methods complete with doc strings where present.
When I import win32api it does do this. It is easy enough to flip back and
forth to the win32 helpfiles, but5 so much nicere to do it inline.

How can I enable this feature please?


Thanks

- Jason
___________________________________________________________
Jason CUNLIFFE = NOMADICS['Interactive Art and Technology']





From Blair.Lowe at compeng.net  Mon Dec  4 18:14:57 2000
From: Blair.Lowe at compeng.net (Blair Lowe)
Date: Mon, 4 Dec 2000 16:14:57 -0700
Subject: SSL with python
In-Reply-To: <v04220807b651d381086e@[192.168.11.33]>
References: <9097pc+pspc@eGroups.com>
 <v04220804b64f0e47b847@[192.168.11.33]> <3A29DBF6.6A074D47@tridog.com>
 <v04220807b651d381086e@[192.168.11.33]>
Message-ID: <v04220809b651d51b68b2@[192.168.11.33]>

More on this:

My email to Val bounced, but I found another item from 
comp.lang.python where someone suggested using M2Crypto Toolkit (also 
at the vaults) instead of ssl_wrapper 1.1.

I shall check this out.

Blair.

At 16:06 -0700 2000/12/04, Blair Lowe wrote:
>Hi again,
>
>I emailed the author of the ssl_wrapper python lib: Val Shkolnikov 
><nvsoft1 at worldnet.att.net>.
>
>Hopefully we can get it from him.
>
>Maybe he reads this list too.
>
>Blair.
>
>At 22:36 -0700 2000/12/02, Ken Kinder wrote:
>>  >From a simple search of http://www.vex.net/parnassus/ -- the 
>>premier python module
>>directory, I found a few. I suggest just going through the results yourself.
>>http://www.vex.net/parnassus/apyllo.py?find=ssl
>>
>>ssl_wrapper 1.1 looks very promising but it's a broken link. Does anyone know
>>where a mirror of that one is?
>>
>>Blair Lowe wrote:
>>
>>>  -----BEGIN PGP SIGNED MESSAGE-----
>>>
>>>  Hi,
>>>
>>>  I am new with python, and I was wondering if anyone has got it to
>>>  work with SSL connections.
>>>
>>>  We are trying to do some socket level manipulation using urllib, and
>>>  httplib, and we are getting some messages saying that we are using
>>>  straight http over an SSL connection.
>>>
>>>  Does anyone know of any information about SSL and python (web or 
>>>printed work)?
>>>
>>>  Thanks,
>>>  Blair.
>>>
[snip]
Computer Engineering Inc. http://www.compeng.net
Phone: 780 499 5687 (9 - 5 MST) Fax:   780 435 0693 (24 Hours)



From dsh8290 at rit.edu  Tue Dec 26 18:07:57 2000
From: dsh8290 at rit.edu (D-Man)
Date: Tue, 26 Dec 2000 18:07:57 -0500
Subject: Why Won't This Work?
In-Reply-To: <LOBBJMHOIFOLFIPFANHJKEFGCFAA.lists@TheWebsons.com>; from lists@thewebsons.com on Sat, Dec 23, 2000 at 06:31:38AM -0800
References: <LOBBJMHOIFOLFIPFANHJKEFGCFAA.lists@TheWebsons.com>
Message-ID: <20001226180757.C8309@westsidecnc.rh.rit.edu>

You mentioned in a different thread that you have RH6.2.  I would
suggest installing from an rpm instead of building it using make.
This will help ease installation/upgrade issues (and makes it simpler
to do).  If you want to build an rpm from the source use the command

rpm -tb the_source.tar.gz

HTH,
-D

On Sat, Dec 23, 2000 at 06:31:38AM -0800, Ben Ocean wrote:
> Hi;
> Because I'm building Zope, I'm trying to install Python-1.5.2. I've
> successfully run:
> ./configure --with-thread
> make
> and everything appears to be in order. But when I type
> python
> at a command prompt, I get a bash error. What have I not done?
> TIA,
> BenO
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list



From jh at web.de  Sat Dec  9 07:06:57 2000
From: jh at web.de (Jürgen Hermann)
Date: Sat, 9 Dec 2000 13:06:57 +0100
Subject: Dijkstra's Shortest Path algorithm
References: <Pine.GSO.4.21.0012090042370.15110-100000@y.glue.umd.edu>
Message-ID: <90t78p$5r1$03$1@news.t-online.com>

"Roy Katz" <katz at Glue.umd.edu> schrieb im Newsbeitrag
news:Pine.GSO.4.21.0012090042370.15110-100000 at y.glue.umd.edu...
> as edges).  It took a good two or three days for me in C++, and
> I was wondering if Python (or Numerical Python) has anything like this
> (a Graph module??).  If there isn't, perhaps I should post my class at
> SourceForge and try to make an extension out of it :)

Maybe you should try to implement it in Python and NOT take 2-3 days. ;))

Tongue-in-cheek-ingly yours, J?rgen




From qrczak at knm.org.pl  Mon Dec 18 12:54:10 2000
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 18 Dec 2000 17:54:10 GMT
Subject: what is a closure?
References: <3a39033b@news.xtml.co.uk> <S18_5.3929$Kd1.455131@newsb.telia.net> <8nck3tocsg8kp1u2g1du6nd3lt9gfeckr2@4ax.com> <91dkvu$7mk$1@nnrp1.deja.com> <ulsr3tctvq7m03hhsofogrl3r6cn3e9f8q@4ax.com>
Message-ID: <slrn93sjq2.en9.qrczak@qrnik.zagroda>

Mon, 18 Dec 2000 11:30:03 +0000, Steve Horne <sh at ttsoftware.co.uk> pisze:

> In short, for the standard behaviour of a closure, would the value
> of a be bound at the time the lambda is defined, or is some kind
> of reference created.

IMHO reference.

Many languages where function closures are important are functional
languages, which avoid mutable objects at all, so that question does
not exist.

The standard way of passing things in Python (and in many other
languages) is by reference. Otherwise it would not be clear how much
to duplicate, because there is no formal distinction beteween an object
"owning" another object and "referring" to it.

If someone wants to pass by value, perhaps he should redesign the
problem to avoid mutability. Then passing by reference is always safe.

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


From pbjorn at uswest.net  Sun Dec 24 14:24:13 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Sun, 24 Dec 2000 12:24:13 -0700
Subject: Some Python 2.1 ideas
References: <mailman.977590932.19335.python-list@python.org> <Cw616.23054$qN6.3308205@typhoon.we.rr.com> <3A452881.7686E717@uswest.net> <%vb16.187233$U46.5801214@news1.sttls1.wa.home.com>
Message-ID: <3A464D5D.B541C04E@uswest.net>

Greg Jorgensen wrote:

> "Bjorn Pettersen" <pbjorn at uswest.net> wrote in message
> [snip]
> How should these be handled?
>
>   "foo\n\n\n"
>   "foo\n\r\n\r"

"foo\n\n\n" -> "foo\n\n"
"foo\n\r\n\r" -> "foo\n\r"

> In other words, does the chomp() method always remove 0-2 characters, or all
> trailing /n/r characters? Is it a good idea to add a string method that

string.rstrip() is probably sufficient if you want to strip all whitespace from
the end...

> makes assumptions about what the file.readline and file.readlines methods
> return? Wouldn't it be cleaner to enhance the file object with additional
> methods:
>
>   f. readline_chomp()
>   f.readlines_chomp()

This would be a good idea, except it wouldn't fit with the standard Python idiom
for reading files:

   while 1:
      line = f.readline_chomp()
      if not line: break # would always break

> Or perhaps add an attribute to the file object to indicate if newlines
> should be stripped for text files?
>
>   f.chomp = 1  # 1=strip trailing /n and /r characters, 0=leave them
> (default)

I'd be hesitant about adding complexity to the file object interface since any
changes might have to be cascaded into StringIO etc.



From b.wolff at gedia.com  Wed Dec 27 03:27:19 2000
From: b.wolff at gedia.com (Wolff, Bernhard)
Date: Wed, 27 Dec 2000 09:27:19 +0100
Subject: AW: How to Grab Keyboards keys?
Message-ID: <0A7AA89ABE04D311801600105A5F76ACB2AE54@GD-DE-AT-01-230>

Calvin Fong wrote:

> Hi, do you know the function to grab keyboards event in python
> Thank a lot.
> 
> 
> 
	I don't have UNIX/Python at hand.
	Therefore I send you some lines for Windows.
	I hope it's helpful.
	Regards Bernhard Wolff

	import msvcrt

	while 1:
	  if msvcrt.kbhit():
	    c = msvcrt.getch()
	    if c != 'q':
	      msvcrt.putch(c)
	    else:
	      break  





From obou at my-deja.com  Tue Dec  5 02:49:57 2000
From: obou at my-deja.com (obou at my-deja.com)
Date: Tue, 05 Dec 2000 07:49:57 GMT
Subject: SAP R/3 automation with win32com
References: <90flpu$gpu$1@nnrp1.deja.com> <kl5n2tsfk3hud1c37gjo56290sueccuegk@4ax.com>
Message-ID: <90i6n6$hg0$1@nnrp1.deja.com>

In article <kl5n2tsfk3hud1c37gjo56290sueccuegk at 4ax.com>,
  dale at out-think.NOSPAM.co.uk wrote:
> obou at my-deja.com wrote:
>
> >Hello!
> >
> >I'm trying to do SAP R/3 automation with win32com using the SAP DCOM
> >Connector (a tool that presents SAP Business objects as COM
Components).
> >The business objects (BAPI) are kind of objects with parameters and
> >methods.
> >Everything works fine, except the function call (the most important
> >part of course), and it could be because of the way I have to give
the
> >different parameters to the BAPI method. I have a VB example and
don't
> >know if the way I convert the VB function call to python function
call
> >is correct.
> >Maybe someone who knows both VB and Python could help me.
> >
> >Here's the Python code:
> >
> >from win32com.client import Dispatch
> >
> ># Create a connection object
> >sapConn = Dispatch('SAP.CCSession.1')
> >
> ># Insert connection informations
> >syst = raw_input('System :')
> >uname = raw_input('Username: ')
> >pwd = raw_input('Password: ')
> >client = raw_input('Client: ')
> >sapConn.PutSessionInfo(syst, uname, pwd, 'EN', client)
> >
> ># Log in
> >sapConn.Logon()
> >
> ># Create a sales order business object
> >salesOrd = sapConn.CreateInstance('SAP.SalesOrder.1')
> >
> ># Create the parameters for calling the method CreateFromData
> >header = salesOrd.DimAs('BapiCreateFromData', 'OrderHeaderIn')
> >itemsIn = salesOrd.DimAs('BapiCreateFromData', 'OrderItemsIn')
> >partners = salesOrd.DimAs('BapiCreateFromData', 'OrderPartners')
> >ret = salesOrd.DimAs('BapiCreateFromData', 'Return')
> >
> ># Initialize structure with data.
> >partners.AddNew()
> >partners.Fields("Partn_Role").Value = 'AG'
> >partners.Fields("Partn_Numb").Value = '0000001234'
> >partners.Update()
> >
> ># Method call. Everything works fine up to this call!!!
> >salesOrd.BapiCreateFromData(OrderHeaderIn = header,
> >                            OrderItemsIn = itemsIn,
> >                            OrderPartners = partners,
> >                            Return = ret)
> >
> >In VB this function call looks like this:
> >salesOrd.BapiCreateFromData OrderHeaderIn:=header,_
> >                          OrderItemsIn:=itemsIn,_
> >                          OrderPartners:=partners,_
> >                          Return:=ret
> >
> >I tried it with or without the parameter names, with or without the
> >return value in the parameters list, I always get this exception:
> >
> >Traceback (most recent call last):
> >  File "d:\python20\pythonwin\pywin\framework\scriptutils.py", line
> >301, in RunScript
> >    exec codeObject in __main__.__dict__
> >  File "d:\sapconn.py", line 35, in ?
> >    Return = ret)
> >  File "<COMObject CreateInstance>", line 2, in BapiCreateFromData
> >  File "d:\python20\win32com\client\dynamic.py", line 234, in
> >_ApplyTypes_
> >    result = apply(self._oleobj_.InvokeTypes, (dispid, LCID, wFlags,
> >retType, argTypes) + args)
> >com_error: (-2147352567, 'Exception occurred.', (0, None, "Table
> >doesn't contain 1 record\015\012(COM_PARAMETER_NO_SINGLE_RECORD):
error
> >code 16388", None, 0, -2147467260), None)
> >
> >Thank you for your assistance,
> >
> >Olivier Boudry
> >
> >
> >
> >Sent via Deja.com http://www.deja.com/
> >Before you buy.
>
> I can't see anything obviously wrong so there's a chance that the
error might be in the preceding
> code. Have you tried running through this interactively and checking
that the objects you are
> creating all look OK?
>
> Also, do you know to what table the error refers?
> --
> Dale Strickland-Clark
> Out-Think Ltd
> Business Technology Consultants
>
>

Thanks for your answer!

I've run the whole thing in an interactive window and my objects are
ok. I'm "translating" a Visual Basic example that works on my system.
So I thought I made something wrong in the last function call.
I don't know VB at all and don't really know how to translate VB
functions calls to Python function calls. Especially when using COM.
There are some strange differences. In my example, there are in VB
function calls that takes 3 parameters where in Python I have to put
only 2 parameters in the brackets and get the third as a return value.
The salesOrd.DimAs function calls are such cases. In VB the return
value is written as a third parameter and the function has no return
value.
Is there a way to identify which parameter have to stay in the brackets
and which ones will be sent as return value?

Olivier Boudry



Sent via Deja.com http://www.deja.com/
Before you buy.


From urner at alumni.princeton.edu  Tue Dec 19 20:13:32 2000
From: urner at alumni.princeton.edu (Kirby Urner)
Date: Tue, 19 Dec 2000 17:13:32 -0800
Subject: Python in "math through programming" curriculum
References: <2g2t3t06r2agrrjmfnr41eu4lc2u39a8qi@4ax.com> <3A3EFF47.67DE69CF@gte.net> <91n9i30pr7@news2.newsguy.com> <slrn9402bt.fa6.neelk@alum.mit.edu>
Message-ID: <ci104tcj1jgqdqfqecsaaao7hgpbu46195@4ax.com>

neelk at alum.mit.edu (Neelakantan Krishnaswami) wrote:

>I disagree with you here about Python's appropriateness. Conceptually,
>the strongly typed FPLs are the best languages there are for learning
>about the mathematics of computer programming.

It could be there's a difference in emphasis here.  My 
curriculum writing features Python as a useful tool 
for exploring math concepts generally, not the mathematics
of computer programming specifically.  We're more 
interested in regular calculus than lambda calculus.

In a lot of ways, this is simply about rescuing kids
from the clutches of the calculator freaks and giving
them a real computer language to play with.  You have
all these schools with fancy computers, but in math
class it's all graphing calculators, with computers
regarded as alien to the subject.

I've dinked around with Haskell some, but haven't 
seen anything specifically geared for 8th graders 
(say).  I think Python's chief competitor for right
now might be DrScheme, as the Rice University folks
have put enormous effort into coming up with 
supporting curriculum materials.  On the other hand,
the emphasis there also seems to be on computer 
math in particular, not so much garden variety 
math math (not that there's always a sharp 
distinction of course).

Kirby



From arturo_perez at wanadoo.es  Tue Dec  5 10:57:48 2000
From: arturo_perez at wanadoo.es (=?iso-8859-1?Q?Arturo_P=E9rez_Mulas?=)
Date: Tue, 05 Dec 2000 15:57:48 GMT
Subject: [DAO Database in win32] What is more efficient
Message-ID: <0g8X5.152$7U6.3922@m2newsread.uni2.es>

Hello,

I am accessing an Access database under Windows NT from Python. I am
using Mark's python windows extensions to access through the DAO
objects, and I have a question and a performance comment:

a) When I open a connection and create a recordset executing an arbitrary
SQL instruction, as follows:

db = dbengine.OpenDatabase("filename")
SQL_query = 'SELECT Table1.Column1, Table2.Colum2  FROM Table1 INNER JOIN Table2 ON (Table1.Num = Table2.Num)  WHERE (((Table1.ID)="name"));'
myrecordset = self.db.OpenRecordset(SQL_BCSet_query)


Is there any way to check if the resulting recordset is empty?



b) I need to add a register to a database but being sure that one particular field is not repeated; I have no access to the predefined structure of the
database, so I need to check manually if such a register exists before adding it; which of the following is more efficient?:

        - make an SQL query restricting the table records with "WHERE field = desired_value" and then check if the resulting recordset is empty

        - retrieve the whole table to a recordset, and check the records one by one:

                myrecordset.MoveFirst()
                while not myrecordset.EOF:
                    if myrecordset.Fields(n) == desired_value:
                        print "found in db"
                        break
                    myrecordset.MoveNext()

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20001205/ea4818fe/attachment.html>

From erno-news at erno.iki.fi  Sun Dec 24 23:59:43 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 25 Dec 2000 06:59:43 +0200
Subject: html filtering
References: <3A46900D.BD118A0A@wielun.dhs.org>
Message-ID: <ku7l4pt7sw.fsf@lasipalatsi.fi>

>>>>> "QdlatY" == QdlatY  <qdlaty at wielun.dhs.org> writes:

| Hello!  Big problem in writeing message boards is filtering html
| tags. I use very simple method: Each occurence of "<" sign or ">"
| sign I replace with &lt; and &gt;. Is this safe method?

no. there have been bugs in some browsers making them interpret
'<' and '>' with the 8th bit set as those characters also.
there may be others. this was a popular topic on bugtraq at one
point, i suggest you check the archives.

| Maybe You've got better ideas?

no, but good luck :)

  -- erno



From pearu at cens.ioc.ee  Sat Dec 30 19:56:00 2000
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Sun, 31 Dec 2000 02:56:00 +0200
Subject: Solution: Re: Lazy Python usage (hack request)
In-Reply-To: <Pine.LNX.4.10.10012301443330.13756-100000@kev.ioc.ee>
References: <Pine.LNX.4.10.10012301443330.13756-100000@kev.ioc.ee>
Message-ID: <Pine.LNX.4.10.10012310238260.2303-100000@kev.ioc.ee>

Hi again,

I have implemented a Python C/API module `lazy' that you can use to put a
Python dictionary to a "lazy" mode.
You can find the lazy module in

http://cens.ioc.ee/~pearu/misc/lazy-0.1.tar.gz

Here is a Python session showing lazy in action:

  >>> import lazy
  >>> d = {'a':4}
  >>> lazy.set(d,lambda k:3)              # set d to lazy mode
  >>> print d
  {'_lazy_cb': <function <lambda> at 0x8112e24>, 'a': 4}
  >>> d['b']
  3
  >>> print d
  {'_lazy_cb': <function <lambda> at 0x8112e24>, 'b': 3, 'a': 4}
  >>> del d['_lazy_cb']                   # disable lazy mode
  >>> d['c']
  Traceback (most recent call last):
    File "<stdin>", line 1, in ?
    KeyError: c

And another example:

>>> import lazy
>>> def fun(k):
...     print 'Creating variable %s'%(`k`)
...     return k
... 
>>> lazy.set(locals(),fun)
>>> print a+b+c
Creating variable 'a'
Creating variable 'b'
Creating variable 'c'
abc
>>> print lazy.set.__doc__
lazy.set(dict,func) - set dictionary dict to a "lazy" mode with
a callback function func.

"lazy" mode means that the following statement
    dict[<name>] = func(<name>)
is executed whenever dict has no key <name>.

WARNING: if dict is set to "lazy" mode, dict.has_key(<name>)
returns always 1 and dict[<name>] is set to func(<name>).

To disable the "lazy" mode for dict, do
    del dict['_lazy_cb']


Be lazy and have fun!-)

Pearu



From syver at NOSPAMcyberwatcher.com  Fri Dec  8 08:58:33 2000
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Fri, 8 Dec 2000 14:58:33 +0100
Subject: My first attempt at python
References: <n2j13tcc0il8k1i3dtjkdckglo871ln744@4ax.com>
Message-ID: <KJ5Y5.3134$9B.90669@news1.oke.nextra.no>

"Jason Stewart" <kahuna01 at hotmail.com> wrote in message
news:n2j13tcc0il8k1i3dtjkdckglo871ln744 at 4ax.com...

Hi Jason!
Couldn't you just do this:
'String that has spaces in it'.split(' ') # split on a space




From darrell at dorb.com  Tue Dec 26 15:56:44 2000
From: darrell at dorb.com (Darrell Gallion)
Date: Tue, 26 Dec 2000 15:56:44 -0500
Subject: One Python 2.1 idea
References: <LNBBLJKPBEHFEDALKOLCAEHNIFAA.tim.one@home.com>
Message-ID: <034e01c06f7e$5b4561d0$6401a8c0@home>

From: "Tim Peters"
> >
> > Am I wildly off-base...?  If not, it seems to me that procuring
> > an acceptable diff-program is going to be the largest part of
> > making a patch to submit...:-).
>
> Easy one first:  the Cygwin diff works fine on Windows; or if you grab a
> copy of the Python source off of SourceForge, cvs diff works OK too.
There
> are detailed instructions for setting up cvs and ssh for use with
> SourceForge on Windows at:
>
>     http://python.sourceforge.net/winssh.txt
>
>
Another source of Unix stuff

http://www.weihenstephan.de/~syring/win32/UnxUtils.html

--Darrell




From dalke at acm.org  Sun Dec 31 10:23:55 2000
From: dalke at acm.org (Andrew Dalke)
Date: Sun, 31 Dec 2000 08:23:55 -0700
Subject: Problems with the re module on IRIX
References: <3A4F432C.386EFFE3@ikb.mavt.ethz.ch>
Message-ID: <92nj0m$oml$1@slb1.atl.mindspring.net>

Yogi Parish wrote:
>I cannot get the re module to run on an SGI. (1.5.2) Has anybody
>had similar experiences with such a problem (make test fails
>in test_re)....

Searching the back newsgroup messages would bring up a few
matches.  There's a problem with SGI's optimizer.  Recompile
the re file without optimizations or use gcc.  (But older
gccs also have problems and set max int to -1.)

                    Andrew
                    dalke at acm.org





From mwh21 at cam.ac.uk  Wed Dec  6 15:08:41 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 06 Dec 2000 20:08:41 +0000
Subject: Bug: Readline getting stuck on Linux and Solaris
References: <lxk8zqwupy2.fsf@ioasun3.epfl.ch> <90m53k$ovh$1@nnrp1.deja.com>
Message-ID: <m3u28h48ja.fsf@atrus.jesus.cam.ac.uk>

sragsdale at my-deja.com writes:

> In article <lxk8zqwupy2.fsf at ioasun3.epfl.ch>,
>   Jan Kybic <Jan.Kybic at epfl.ch> wrote:
> > Hi,
> > 	I am using Python 2.0, I downloaded it in the RPM form and I
> > also compiled it from source on both Linux RedHat 6.2 and Solaris. In
> > both cases, when I have the readline module enabled, the keyboard
> > input gets stuck.
> 
> I'm having similar problems with Python 2.0 hanging in Irix.  What's
> most annoying is that this happens on the Python command line when you
> hit control-C.  In Python 1.5.2 this is a KeyboardInterrupt event.  In
> Python 2.0 this hangs the interpreter which starts using 100% of the
> processor.  I can't kill it with more control-C's.  If I disable the
> readline (4.1) module this problem goes away.  This problem does not
> happen in Linux 2.2.15.  I can't find this bug listed on the sourceforge
> bug list.

Hmm.  This *used* to happen on Linux (with slightly different
symptoms; control-C would kill the interpreter, not hang it), but the
bug got fixed.  A couple of things to try: does control-C do what it
should when you're not sitting at the ">>> " prompt?  E.g. if you type

>>> while 1:
...    pass
...

and whack control-C, what happens?

The other thing is to try building without threads.

The problem with Linux was a sigaction/signal mismatch - but I don't
know a lot about signal handling and nothing at all about IRIX, so I
can't say if I'd expect trouble here....

Cheers,
M.

-- 
  This makes it possible to pass complex object hierarchies to 
  a C coder who thinks computer science has made no worthwhile
  advancements since the invention of the pointer.
                                       -- Gordon McMillan, 30 Jul 1998


From glenfant at equod.com.nospam  Tue Dec 19 18:08:18 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Wed, 20 Dec 2000 00:08:18 +0100
Subject: Wanted: Python White Papers
References: <91oj8v$i0r$1@nnrp1.deja.com>
Message-ID: <91opg9$l6f$1@reader1.imaginet.fr>

Our previous application mailer (builds text messages and sends them through
an SMTP server) was built using WebObjects (java) application server.
It samples periodially the database and builds/sends plain text messages
built from new database records.

- It's only capable of text/plain messages.
- It "eats" about 34 Mb of RAM, and ALWAYS (works as a daemon).
- No alarm possible when the SMTP server is down and marks messages as sent.
- Sometimes... Dr. Watson is coming :((

I made a Python workaround within 4 day (I was a Python newbie)
- Makes plain text and html pretty multipart mail (very easy).
- Uses 2 Mb and only when working.
- About 30% less source code.
- Can detect when SMTP is down and tries to send the mails later.
- Never seen Dr. Watson.
- But about 40% more process time for the same thing (but it's 12 seconds
every hour).

I'm definitely convicted to use of Java or C only when nothing else is
possible.

This is my Whitepaper :))

Go to www.zope.org . There are links to big companies which Web sites are
powered by zope/python.

I know that in France, lots of IT project managers don't like to use open
source products because no open source project sales people will invite them
to the restaurant ;-)

<noahspurrier at my-deja.com> a ?crit dans le message news:
91oj8v$i0r$1 at nnrp1.deja.com...
> Is anyone aware of any white papers to explain the use of Python
> at any big, high-profile companies?
> I'm trying to use a "look-how-these-cool-companies-use-python"
> argument to defend the use of Python for a project. That won't be my
> only argument, of course :-) But currently Python is not winning many
> management converts here based solely on its technical merits.
> Many still say, "What's Python? Never heard of it...". It helps
> a lot when I say "Google uses it!", but it would help more if I
> knew HOW it is used and WHY.
>
> Actually, the only high-profile company that uses Python that I
> can think of is Google. Is anyone familiar with their architecture?
> How does Python fit into their design?
>
> Is there anyone out there from Google that can represent Google's use
> of Python? What are Google's plans for the future of Python in their
> architecture? I heard that they might be switching to Java.
>
> Yours,
> Noah Spurrier
>
>
> Sent via Deja.com
> http://www.deja.com/



From andrer at stud.cs.uit.no  Tue Dec 12 04:47:24 2000
From: andrer at stud.cs.uit.no (=?iso-8859-1?Q?Andr=E9?= Larsen Risnes)
Date: Tue, 12 Dec 2000 10:47:24 +0100
Subject: compiled or interpreted?
References: <w8jZ5.74702$IP1.2487256@news1.giganews.com>
Message-ID: <3A35F42C.A506F753@stud.cs.uit.no>

Matt Frey wrote:
> 
> Is Python a compiled language that can be run on a platform like c++, or is
> it interpreted like java, or is it in a class of its own?

Its interpreted, like perl. See http://www.python.org/doc/Summary.html

alr


From alawhead at vcn.bc.ca  Mon Dec 11 17:18:11 2000
From: alawhead at vcn.bc.ca (Alexander Lawhead)
Date: 11 Dec 2000 22:18:11 GMT
Subject: Python-like compiled language
Message-ID: <913jr3$dmv$1@sylvester.vcn.bc.ca>

This may seem like an odd question, but what is the most Python-like
compiled language? I'm a new programmer and have been learning on Python.
I'm trying to build some command-line driven programs to do some image
processing and am trying to find a Python-like compiled language that
would be faster than Python+NumPy. Ideally I guess I should learn C and
combine Python + C extensions. For the moment, I'd like to start on a new
compiled language, preferably one that's oriented towards math/science
(hopefully with built-in or standard matrix libraries). I've fallen in
love with Python+NumPy's matrix capabilities and am far more interested in
a language that comes close in this regard. Object oriented capability is
not necessary.

So far, I've been pointed to F, a subset of Fortran 90. Is there anything
else out there? Wide usage is not necessary, but free availability under
Linux or WinXX is!

Thanks,

Alexander



From aleaxit at yahoo.com  Wed Dec 27 11:57:48 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 27 Dec 2000 17:57:48 +0100
Subject: A Suggestion for Python Colon Syntax
References: <mailman.977546107.21808.python-list@python.org> <92cu71$9m4$1@nnrp1.deja.com>
Message-ID: <92d72h017po@news2.newsguy.com>

<billtj at my-deja.com> wrote in message news:92cu71$9m4$1 at nnrp1.deja.com...
    [snip]
> > So does Guido, but it's rare, and in this case you're ignoring that he
> > *already decided*.
>
> Tim, people opinion can change over time, and hopefully Python does not
have
> the principle that once a feature is already decided, it can never be
> changed.  I think to survive, a language has to be flexible to follow
> (rapidly) changing technology and therefore modify some of its own
> characteristics.

Funny enough, Python survived and thrived for 10 years based on a LOT
of respect for the installed code-base -- and therefore UTTER prudence
in "adding features", because you just don't "retract" a feature later
(breaking people's good code).


> I think there is some inconsitency here.  If Python is really not about
> minimizing keystrokes but more about maximining readability, then I have
> at least one counter-point:  Why don't add a pair of parentheses after the
> clauses such as:
>
>     if (xxxx xxxx xxxx):

Yecch.  Redundant punctuation -- the WORST kind, too... ubiquitous
parentheses!


> Just like in Tcl, why not make everything consistent?  Like

Because "A foolish consistency is the hobgoblin of little minds"
(Ralph Waldo Emerson)?


> It will then be easier to memorize the basic syntax of Python.  Because
every
> time I have to switch between C/C++, Perl, etc., I have to look up the
> books just for syntax.

Perl is a special case of going overboard in wild inconsistency.
Python's tasteful use of little touches of it here and there is
just my thing (except only for the horrid 'print >> bah' blotch).


> > Guido does.  Now what?
> >
> Guido does, but shouldn't we take also the opinions of the community?  I

No!  You want a language designed by committee?  Take your pick,
there's a wild abundance of THOSE around.

> don't like to make this analogy, but in the history, there were/are
> communities ruled by kings, and there communities ruled by
> democracy.  Which one works better for the welfare of the community?

In a hypothetical situation where anybody is totally free to choose
which community to "be born" in, and vote with his or her feet by
moving to another with minimal inconvenience, a maximal variety of
regimes might be optimal -- give everybody more choice.

Would you rather listen to a symphony composed by one great musician,
or to one made up, one note at a time, each the fruit of some democratic
political compromise?  The answer is pretty clear to ME -- particularly
since there are so many examples of the latter in the software world.

As long as some designer enjoys AND USES practical veto-right, a
committee can work to enhance a language -- the ANSI/ISO committees
for C did, and Ritchie showed the need for veto rights in his famous
"Noalias must go. This is non-negotiable" communication.  The
Python community (or at least its Guido-selected inner core, the
membership of the 'Python-developers mailing list') plays a similar
role -- they take straw polls and everything.  But our Benevolent
Dictator For Life (BDFL) is the one who decides in the end -- he
lets *good arguments* sway his decisions, but NOT opinions, even
widespread ones.


> The problems with C++ is a lot of software construction is based on
> conventions, and not part of the language itself.  But because it is based
on
> conventions, people can abuse them.

Python is even MORE convention-based than C++ -- for example, if you
design a class with some attribute X that is only meant for subclasses'
use, in Python you just document that *and rely on people's respect for
this design-convention*, while in C++ you can try to enforce it with
the 'protected' keyword (one of the few specific design decisions that
the language designer explains he now thinks were bad mistakes in his
excellent book on "Design and Evolution of C++" -- but, fortunately for
his language's actual use if not for its theoretical elegance, he, too,
respected the installed codebase too much to ever take something away).

If you want a language where everything is strictly enforced, you may
try Eiffel, but I think even that will fail to live up to that ideal.


> I'm sorry, Tim, my background is not too strong in this parsing
department.
> My question is, can you just detect the first word of the line?  I think
> there are only a finite number of them, such as class, def, if, for, etc.
If
> the first word is any of these, then indent the next line, without regard
of

That would be bad -- after I type
    if chatty: print "OK so far"
I most definitely do NOT want the next line indented (that would be a
Python syntax error unless I undid the indent manually!).


Alex





From spahievi at vega.bg  Tue Dec  5 15:39:07 2000
From: spahievi at vega.bg (Niki Spahiev)
Date: Tue, 5 Dec 2000 22:39:07 +0200
Subject: "inherited" keyword in Python?
In-Reply-To: <90gevs02uu1@news2.newsguy.com>
References: <3A2B9EFA.3A64C13B@schlund.de> <90gevs02uu1@news2.newsguy.com>
Message-ID: <1742065242.20001205223907@vega.bg>

04.12.2000, 17:58:18, Alex Martelli wrote:

AM> "Carsten Gaebler" <cg at schlund.de> wrote in message
AM> news:3A2B9EFA.3A64C13B at schlund.de...
>> Hi there!
>>
>> Does Python know something like the "inherited" keyword from Borland
>> Pascal? In Pascal it is used to call a method of a class's superclass,
>> i.e. instead of saying "superclass.method()" you say "inherited
>> method()". Can be quite helpful if you change the name of the superclass
>> or just forgot it. :-)

AM> Python has multiple inheritance, so there is (in general) no single
AM> 'superclass' that you could use that way.

OK but consider this:

class A:
  def foo( self ):
    print 'A.foo'

class B:
  def foo( self ):
    print 'B.foo'

#class C(A,B):
#...

class C_help(A,B):
  pass

class C(C_help):
  def foo( self ):
    C_help.foo( self )

I would like to have same functionality without writing X_help class
whenever i need it.

-- 
Best regards,
 Niki Spahiev





From Gareth.McCaughan at pobox.com  Fri Dec 29 18:31:51 2000
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: Fri, 29 Dec 2000 23:31:51 +0000
Subject: Interesting behaviour of the assignment
References: <92f4gk$qk6$2@news.nuri.net> <92fhsr01u88@news1.newsguy.com> <mailman.978059501.14500.python-list@python.org> <92j4n1$mhu$0@216.39.162.232>
Message-ID: <slrn94q7n7.905.Gareth.McCaughan@g.local>

Michael Esveldt wrote:

> This is a somewhat ignorant question, but if 'x == y' checks for 'x is 
> y' first, why use 'is' at all? If you only get a tiny bit of speed out 
> of using 'is' is it worth it to have this in the language at all? (Not 
> that it would ever change, just curious.)

1. You get a bigger increase in speed from "is" when it returns
   false.

2. You use "is" when you actually want to check object identity.
   For instance, consider this nasty code:

       class Foo:
         ...
         def copy_thing_from(other_foo):
           self.thing.destroy()
           self.thing = other_foo.thing.copy()

   It's nasty for several reasons. One of them is that if
   you have a Foo -- call it "foo", say -- and do foo.copy_thing_from(foo),
   then you'll lose. So you need to check that the thing you're
   destroying isn't the same thing you're about to copy.

       class Foo:
         ...
         def copy_thing_from(other_foo):
           if self.thing is not other_foo.thing:
             self.thing.destroy()
             self.thing = other_foo.thing.copy()

   You might still want to do the destroy-and-copy if the two
   "thing"s were == but not identical.

   An example of somewhere where you really might want to do this
   sort of thing: suppose you're working with very large numbers
   (for cryptography or something). You might have a "Bignum" class
   for them, and for reasons of efficiency you might want to make
   some arithmetic operations able to happen in place. But it might
   be very painful to make a.add_to(a) work completely in-place, so
   you'd need to check for the possibility that that was happening.

3. Another use for object identity: the ability to make unique
   objects. A random example:

       class Foo:
         def __init__(self):
           self._wibble = []
         def do_something(self, optional=self._wibble):
           if optional is self._wibble:
             return "optional not given"
           else:
             return "optional %s" % optional

   Unless the user deliberately screws around, this enables you
   to tell reliably whether an optional argument was given. If
   you use None (for instance) as the default, this isn't so.

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
sig under construc


From grante at visi.com  Thu Dec 28 12:50:52 2000
From: grante at visi.com (Grant Edwards)
Date: Thu, 28 Dec 2000 17:50:52 GMT
Subject: Reality check PYZOMA = python / zope for  Mac ?
References: <t44t20nmcgena7@corp.supernews.com>
Message-ID: <04L26.3810$4c.282264@ruti.visi.com>

In article <t44t20nmcgena7 at corp.supernews.com>, Jason Cunliffe wrote:

>I am wondering what the prospects are for Python and Zope on Macintosh:

I would guess that Python/Zope would run fine under Linux on a
Mac, but don't know firsthand of it being done.

-- 
Grant Edwards                   grante             Yow!  I have seen these
                                  at               EGG EXTENDERS in my
                               visi.com            Supermarket... I have read
                                                   theINSTRUCTIONS...


From erno-news at erno.iki.fi  Sat Dec  9 10:11:47 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 09 Dec 2000 17:11:47 +0200
Subject: Python wrapper for C++ STL?
References: <3A31760E.D157EB54@erols.com> <C0gY5.126062$U46.4232859@news1.sttls1.wa.home.com> <90s3fk$j8p$1@nnrp1.deja.com> <fUgY5.126218$U46.4241063@news1.sttls1.wa.home.com> <90s9l9$njd$1@nnrp1.deja.com> <KVjY5.126920$U46.4261650@news1.sttls1.wa.home.com> <90t9n902p8r@news2.newsguy.com>
Message-ID: <kur93h8w98.fsf@lasipalatsi.fi>

In article <90t9n902p8r at news2.newsguy.com>, "Alex Martelli"
<aleaxit at yahoo.com> writes:

| I consider it pretty sad that I still have to code Python extensions in C,
| rather than C++, to ensure they're "taken seriously" by the Python
| community, actually.  I'll do it when I must (as I'm doing in gmpy), but,
| it *IS* a sad reflection on the state of the Python world:-).  Oh well,
| at least it can be 'full ISO C' (commonly still mis-called ANSI C:-) these
| days -- meaning the standard of 1989, of course (we wouldn't want to
| be TOO modern, now would we?-).

there's a running discussion on Advogato <URL:http://www.advogato.org/> on
this C++ issue. i found it very interesting.

  -- erno


From krimsly at home.com  Mon Dec 11 23:07:40 2000
From: krimsly at home.com (North Krimsly)
Date: Tue, 12 Dec 2000 04:07:40 GMT
Subject: Executable python program?
References: <jKYY5.54278$df5.1795043@news1.crdva1.bc.home.com> <3a34a653@news.xtml.co.uk>
Message-ID: <gwhZ5.137408$U46.4588203@news1.sttls1.wa.home.com>

Or, check out Gordon McMillan's excellent Windows EXE maker at
http://mcmillan-inc.com/index.html

Gordon is a real magician.

"Nick Trout" <nick at videosystem dot co dot uk> wrote in message
news:3a34a653 at news.xtml.co.uk...
> > I am developing a python/Tk program in linux.  I want Windows users to
be
> > able to use this program as well.  Now...the situation I am in it will
> > make it inconvenient to get the windows people to install python/tk to
run
> > things.  Is it possible to get a python "Executable" per say that isnt
> > interpreted but is a "static" executable persay?  By this I mean static
> > in the same form as a linux executable that is "static" by making no
> > external library calls (or none other than system calls if that is
> > needed).
>
> You could look at PySol. It has everything you need to run a Python app in
> Tk.
>
> http://google.yahoo.com/bin/query?p=pysol&hc=0&hs=0
>
>
>
>




From garachop at club-internet.fr  Tue Dec  5 15:51:59 2000
From: garachop at club-internet.fr (G P)
Date: Tue, 05 Dec 2000 21:51:59 +0100
Subject: win32evtlogutil: How does it works ?
Message-ID: <3A2D556F.6240C661@club-internet.fr>

Hello,
I'am trying to write a script under windows NT.
I want it to logs events.
I use AddSourceToRegistry and ReportEvent

first :
AddSourceToRegistry( appName, dllName)
then :
ReportEvent( appName, 1063, eventType = warning, strings = [ 'Test.'])

My  problem is about dllName.

The code works, but if I try to know more about the event, the box
display : 
"La description pour l'ID d'?v?nement (1063) dans la source (DTO) n'a pu
?tre trouv?e. Elle contient la ou les cha?nes d'insertion suivantes
:Test."
I.E.
Description for event ID ( 1063 ) in the source (DTO) can't  be found.
It holds the following insertion's strings :Test.

If I use any specified dll I can found in the registry for other events,
it's the same,  except the norton antivirus dll : s32alogn.dll whith an
event ID number 4098.

Any ideas ?
Thanks.


From tim.one at home.com  Mon Dec 18 22:20:15 2000
From: tim.one at home.com (Tim Peters)
Date: Mon, 18 Dec 2000 22:20:15 -0500
Subject: Helpme "Corrupt installation detected"
In-Reply-To: <3a3ec297.51392347@news.neti.ee>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEHOIEAA.tim.one@home.com>

[posted & mailed]

[Teet Tarno]
> I have a dumb problem for all you nice people out there.

And we have a dumb answer, I'm afraid.  Thanks for the clear explanation of
the problem!

> One day I downloaded the Beopen-Python-2.0.exe and tried to install
> it on W95. Well, I ran the installer and it ran out of diskspace. No,
> not the 18M install, but FAT16 and 1800 (a lot) files. I used
> Quarterdeck's Cleansweep, uninstalled the thing, freed some more
> space and now when I try to run it, it does nothing and simply says :
> "Corrupt installation detected" and exits.
> To the best of my knowledge, Cleansweep should have removed all
> traces of the install (HD, *.ini, registry). I did search in the
> registry and took a quick look around the disk. Nothing seemingly
> left.

It probably did -- the install isn't very fancy (for example, it doesn't
create or modify any .ini files, or indeed *any* files except under the
\Python20 tree, plus a DLL in the System directory).

> Where does the BeOpen installer save the info about the failed
> installation ?

It doesn't.  Well, there was an install.log file created under \Python20,
but that's probably gone now.

> I could download the installer again, but 5M over slow modem is not
> THAT much fun. But if it has to Be...

I'm afraid it has to be.  The "corrupt installation" msg comes from the
installer scanning *itself* and getting a bad checksum.  The .exe file
itself is corrupted now.  It's a bad idea to run out of disk space on Win95
<wink>.

the-good-news-is-it-will-work-next-time-ly y'rs  - tim




From scarblac at pino.selwerd.nl  Wed Dec 27 06:08:46 2000
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 27 Dec 2000 11:08:46 GMT
Subject: Just starting out!
References: <92buip$kcp$1@nnrp1.deja.com> <u5i26.194888$U46.5994443@news1.sttls1.wa.home.com> <3a49c756.8528401@news.skynet.be>
Message-ID: <slrn94jj88.gfb.scarblac@pino.selwerd.nl>

Daniel <Daniel.Kinnaer at AdValvas.be> wrote in comp.lang.python:
> Where can one find a tutorial with lots and lots of example code?
> Reading about theory is fine when one already is proficient at a
> language... I learn fastest when examples are given, and that is what
> I'm looking for. Is there such a book/tutorial available?

There is an electronic book called "(the eff bot guide to) The Standard
Python Library" that is based on over 2500 newsgroup posts by Fredrik
Lundh, it has examples for every standard Python library.

http://www.secretlabs.com/people/fredrik/librarybook.htm

It's $12 (Windows only I think).

> As a newbie to Python, I always compare with Delphi and how I would go
> about solving a problem.  Very often, I can't find an answer in the
> ref-guide nor in the tutorials (or I am looking at the wrong places
> :-)  E.g.  I was trying to convert a number to a string and add
> another string to it. In Delphi i would write s:= IntToStr(myNumber) +
> ' is my number';  How is this done in Python? 

Either

s = str(myNumber)+" is my number"

or

s = "%d is my number" % myNumber

str(), int(), float(), list(), tuple() are the obvious convertion functions.
(there are probably a few I didn't list).

-- 
Remco Gerlich


From moshez at zadka.site.co.il  Fri Dec 29 18:06:13 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Sat, 30 Dec 2000 01:06:13 +0200 (IST)
Subject: One Python 2.1 idea
In-Reply-To: <018a01c071a5$e52abde0$102b2bc1@cadlab.it>
References: <018a01c071a5$e52abde0$102b2bc1@cadlab.it>, <92gjnu02q3r@news2.newsguy.com>, <mailman.978038049.481.python-list@python.org> <20001229190243.78478A84F@darjeeling.zadka.site.co.il>
Message-ID: <20001229230613.13868A84F@darjeeling.zadka.site.co.il>

On Fri, 29 Dec 2000 15:44:43 +0100, "Alex Martelli" <aleaxit at yahoo.com> wrote:

> I'm not claiming it's a _frequent_ application need, to
> change the behavior of a few methods in certain instances
> only -- but, I do claim it IS a real need, and I have the
> scars (and frameworks of Coplienesque C++ templates, because 
> that's the language I was using then) to prove it...:-)

OK, now we're getting somewhere: can you give an example (perhaps
simplified) of a problem this helps to solve?

this-will-probably-another-fun-and-fruitless-discussion-ly y'rs, Z.
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From aleaxit at yahoo.com  Tue Dec 12 09:52:00 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 12 Dec 2000 15:52:00 +0100
Subject: A question.
References: <9152a0$acb$1@panorama.wcss.wroc.pl> <9156h002uh2@news2.newsguy.com> <915a7j$fe0$1@panorama.wcss.wroc.pl>
Message-ID: <915elr0681@news2.newsguy.com>

"DeHa" <deha at poczta.onet.pl> wrote in message
news:915a7j$fe0$1 at panorama.wcss.wroc.pl...
>
> Alex Martelli napisa?(a) w wiadomo?ci: <9156h002uh2 at news2.newsguy.com>...
> >"DeHa" <deha at poczta.onet.pl> wrote in message
> >news:9152a0$acb$1 at panorama.wcss.wroc.pl...
> >
> >first = range(1,100001)
>
> I try this, but it doesn't work

Of course; there are many other errors, some of which I pointed out.
But this was a circuitous (although not erroneous) way to work


> >> for i in first:
> >>     if i != []:
> >> #until end of the list
> >
> >i will never equal an empty-list, as first's items are just integers.
>
> Well, I don't konw why it happend, but when there were no such line the
> program doesn't work.

Your program doesn't work anyway (with or without that line)
because of several other issues.  But, again, this specific
line plays no role.


> >Dynamical, interpreted languages such as Python are
> >not *mostly* about performance -- but hundreds or even
> >thousands of times of slowdown/speedup can easily be
> >had in simple cases by paying a _little_ attention to
> >performance issues.  I think it can be worth it...
> >
>
> Thanks for advice. Next time I will try to find better performance
issues:)

Correctness-first: you'll notice I _first_ produced a
working version of your original idea (modifying the
list itself, physically deleting items known not to be
primes), and only later introduced the key optimization
(not alter the list itself as we go, but rather keep
a separate 'parallel' list of flags; which lets us
remove the list itself, as it happens, since the list
of flags contains the same information...).

The key _simplification_ that made both versions easier
to write was looping on indices rather than on values.

Sometimes looping on values is simpler, sometimes you
need indices, sometimes it's ideal to have both, and in
the latter case it may be worth doing stuff like:
    for index, value in zip(range(len(seq)),seq):
        # whatever
(it WOULD be nice to have this frequent idiom be
more naturally expressed in Python -- Lemburg's
"new builtins" include this one and quite a few
other handy ones -- but, oh well).


Alex





From peter at schneider-kamp.de  Sun Dec 24 05:43:36 2000
From: peter at schneider-kamp.de (Peter Schneider-Kamp)
Date: Sun, 24 Dec 2000 11:43:36 +0100
Subject: Some Python 2.1 ideas
References: <005f01c06d01$d8114820$74eb0b18@C322162A> <20001223212159.H5515@xs4all.nl>
Message-ID: <3A45D358.CBEC055A@schneider-kamp.de>

Thomas Wouters wrote:
> 
> On Sat, Dec 23, 2000 at 09:00:25AM -0800, Bob Alexander wrote:
> 
> > Variations on "for"
> > Here are some possible additional forms of "for" that seem appealing:
> 
> Funnily enough, I also implemented the 'indexing-for' almost-PEP (which was
> written by Just van Rossum, but not accepted as a PEP) which was rejected
> because of uhm, one reason or another. Mainly the Guido-line above. I have
> since considered proposing
> 
>  for i in <x> .. <y>:
> 
> (or perhaps three dots instead of two) as an alternative to the
> 'indexing-for' syntax, but I wanted to catch Guido in a sufficiently good
> mood to try it. And perhaps I should first have a public fight with Just,
> just so show that I'm not conspiring with him :-)

Can you detail the 'for i in <x> .. <y>' proposal? Maybe it's
something I could add as an alternative in PEP-0212 (loop
counter iteration).

Anyway, have a look at
  http://python.sourceforge.net/peps/pep-0212.html

merry-xmas-for-the-last-time-ly y'rs
Peter



From badassgeek at earthlink.net  Mon Dec 11 06:40:43 2000
From: badassgeek at earthlink.net (David Humphreys)
Date: Mon, 11 Dec 2000 03:40:43 -0800
Subject: variable variables
In-Reply-To: <Pine.LNX.4.21.0012111127460.5044-100000@relax.dk>
Message-ID: <LIEMKHFIDGKFBCHAGODBEEADDLAA.badassgeek@earthlink.net>

You could:

exec "print test%d" % (var1)

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of pyth at relax.dk
> Sent: Monday, December 11, 2000 2:35 AM
> To: python-list at python.org
> Subject: RE: variable variables
> 
> 
> > I guess you want to be able to have multiple references to the same
> > variable, and that you want to be able to change the value of that
> > variable in such a way that all references refer to the same new value.
> 
> > /Mikael
> 
> Sorry Mikael, for using your time... I should have been more specific...
> What i was thinking of, was something like this:
> 
> test1 = 'test text 1';
> test2 = 'test text 2';
> 
> var1 = 1
> var2 = 2
> 
> print test{var1};
> 
> This should print: 'test text 1'
> This example doesn't work, but is close how you would do it in php :o)
> 
> Thanx for tryin' Mikael :o)
> 
> /Dan Larsen
> 
> 
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 



From jim.vickroy at noaa.gov  Wed Dec  6 16:41:47 2000
From: jim.vickroy at noaa.gov (j vickroy)
Date: Wed, 06 Dec 2000 14:41:47 -0700
Subject: wither PyFITS ?
Message-ID: <3A2EB29B.EBB3CCB5@noaa.gov>

Hello,

The following link

http://monet.astro.uiuc.edu/adass98/Proceedings/barrettpa/

claims there is a FITS module (PyFITS) available, but I have not been
able to find it.

Any help would be appreciated.

Thanks,

-- jv


From vogl at mavo.de  Sat Dec 30 18:42:45 2000
From: vogl at mavo.de (Markus Vogl)
Date: Sat, 30 Dec 2000 23:42:45 GMT
Subject: Probs with classes: cannot instantiate
Message-ID: <20001230.23424514@snafu.btx.dtag.de>

Hi all!

Sorry for asking a [stupid] question:

Building and instantiating a class in the interpreter works pretty well.
Nevertheless in the file modus it produces error messages like:
?call of non function?

Example:

This is the interpreter stuff:
[d:\]python
Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)] on win32
>>> class DirParser:
...     def display(self):
...             print "This is the display method!"
...
>>> x = DirParser()
>>> x.display()
This is the display method!
>>>

This is the file [ DirParser.py] stuff:
class DirParser:
	def display(self):
		print "This is the display method."

To test this i use classtester.py in the same directory [which is in 
path]:

import DirParser
x = DirParser()
x.display()

The error messages:
	x = DirParser()
...
TypeError: call of non-function ( type module )
...

Any help will be appreciated!!
Thank you!!!

TIA
Markus














From MarkH at ActiveState.com  Mon Dec  4 23:54:51 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Tue, 05 Dec 2000 04:54:51 GMT
Subject: Python programs (or scripts) as drop targets
References: <hv5k2tg9cdf33j8v333qpkk7jo45ht596n@4ax.com> <90d9hi029hm@news1.newsguy.com> <3A2ACF21.4060406@ActiveState.com> <90fl26022in@news2.newsguy.com>
Message-ID: <3A2C73CE.5060506@ActiveState.com>

Alex Martelli wrote:

> If saving a trivial file is indeed the only advantage of implementing
> the shell extension, then you may be right (though quibbling about the
> "registry additions" needed to _register_ a shell extension seems pretty
> peculiar to me).

Hrm - I hope you didn't think I was quibbling about that!  You won't 
find me arguing against that.  I apologize if that is what I was saying!

While we are at it, the registry information needed to the shell 
communicates to a running Pythonwin via DDE (instead of a new Pythonwin 
startng and _it_ using DDE) would be nice too ;-)

Mark.



From grey at despair.rpglink.com  Thu Dec  7 15:16:34 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Thu, 07 Dec 2000 20:16:34 -0000
Subject: GNU tools in a library?
Message-ID: <slrn92vrnb.1v5.grey@teleute.rpglink.com>

    Has anyone created a lib for python which reimplements the standard GNU
shell/text tools in Python?

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From grey at despair.rpglink.com  Tue Dec 12 18:39:45 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Tue, 12 Dec 2000 23:39:45 -0000
Subject: Spinning off time intensive procedures to speed CGI
References: <916bpn$a6i$1@nnrp1.deja.com>
Message-ID: <slrn93ddpv.eum.grey@teleute.rpglink.com>

On Tue, 12 Dec 2000 23:19:26 GMT, leejor at my-deja.com <leejor at my-deja.com>
wrote:
>How do I accomplish this? I have played with fork() and
>os.system(), but have not achieved the results I need. In
>my searches of the documentation and online forums, I
>have only found methods that _wait_ for the spun off
>proceedures to return a result.

>I don't want to wait.

    Won't one of the exec's work?

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From cg at schlund.de  Mon Dec  4 08:41:14 2000
From: cg at schlund.de (Carsten Gaebler)
Date: Mon, 04 Dec 2000 14:41:14 +0100
Subject: "inherited" keyword in Python?
Message-ID: <3A2B9EFA.3A64C13B@schlund.de>

Hi there!

Does Python know something like the "inherited" keyword from Borland
Pascal? In Pascal it is used to call a method of a class's superclass,
i.e. instead of saying "superclass.method()" you say "inherited
method()". Can be quite helpful if you change the name of the superclass
or just forgot it. :-)

Regards
Carsten.


From grey at despair.rpglink.com  Mon Dec 18 13:58:21 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Mon, 18 Dec 2000 18:58:21 -0000
Subject: while true: !!!
References: <mailman.977143081.15264.python-list@python.org>
Message-ID: <slrn93snid.tvg.grey@teleute.rpglink.com>

On Mon, 18 Dec 2000 13:36:27 +0100 (CET), Carsten Geckeler <nospam at no.spam>
wrote:
>Anyway it's a very common idiom.  But the point is that using true and false
>would make the code more readable in cases where it's not clear, that the
>value is used as a boolean value later.  E.g:

[Example of if conditions snipped]

    You're mixing examples.  I doubt that anyone has a problem with storing a
value in an appropriately named variable.  However, how does that address
having true equate to false?  That would be the requirement of having "while
true:" and then terminating the loop with assigning "true = 0" thus making
that consition false.  Meanwhile the very valid break would go unused for the
exact purpose that it was designed for, breaking out of a loop.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From josh at open.com  Fri Dec  8 01:35:19 2000
From: josh at open.com (Joshua Muskovitz)
Date: Fri, 8 Dec 2000 01:35:19 -0500
Subject: Attention NYC Shoppers & Visitors - Beware of Rip-off SY Stores!!! ...... kHP2pr9I
References: <66FF28553F3.hlfeano@wmwnkhkpvfv.gov> <3a2f1968_4@corp.newsfeeds.com> <3a2f30aa$1@news.iprimus.com.au> <3a2fdb88_4@corp.newsfeeds.com> <3a302348@news.iprimus.com.au>
Message-ID: <3a307f7a_4@corp.newsfeeds.com>

> "Well, duh" ??
> Arrggggg! stabbed thru the heart.

Hee hee!  Sorry, thought that my original post made my thinking clear, but
apparently that was only for those within proximity of my highly localized
"Duh" field.  So I had to extend it out to all of usenet for a minute there.
Don't worry, the effects wear off quickly.

:-)

-- josh




-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----


From billtj at y.glue.umd.edu  Fri Dec 22 16:00:57 2000
From: billtj at y.glue.umd.edu (William Djaja Tjokroaminata)
Date: 22 Dec 2000 21:00:57 GMT
Subject: A Suggestion for Python Dictionary/Class
Message-ID: <920fe9$c20$1@hecate.umd.edu>

Hi all,

After using Python for a while, a have a wish for another type of
dictionary.  If you ever used Tcl and Perl, you know that all Python, Tcl,
and Perl basically provides three different fundamental data types:

    - scalar (Perl-speak) (which includes numbers and strings)
    - list (or sequence in Python-speak)
    - dictionary

But in Python, we have two different kinds of lists: lists and tuples.  I
used Tcl and Perl before, so initially I was puzzled by why Python
provides a 'read-only' list.  But it is fine.

I also use C and C++ a lot.  Their power and large-scale capabilities come
from, among other things, the ability to write struct or class.  Among the
three scripting languages, only Python provides the most natural form of
struct/class like in C/C++/Java.  That's why I love Python so much; I can
think in Python much like I think in C/C++/Java.

However, there is a fundamental difference with Python class.  The members
are implemented using dictionary.  The danger with dictionary is, even
though a key does not exist, we still can make an assignment to it.  When
we make some typo, this can be hard to debug.  For example:

    class MyClass:
        def __init__ (self):
            self.member = 0

    myClass = MyClass ()
    myClass.menber = 1    # This is a bug!

    print myClass.member  # Expected 0, but found 1

In C/C++/Java, if a member does not exist and we try to access or assign
it (because of some typo, perhaps), it will be caught by the compiler
simply as syntax error.  Not so with Python class.

Therefore, is there any way for me to define a class, with some definite
members just like in C/C++/Java, so that no new members can be defined?  I
am trying to do it, but using some conventions, not using the Python
language features.

Because Python provides a read-only list in terms of tuple, I think it
will be nice if Python provides another class type with fixed members,
just like a dictionary with read-only keys.  Any opinion, anyone?

Regards,

Bill


From slhath at home.com  Sat Dec  9 11:04:37 2000
From: slhath at home.com (Scott Hathaway)
Date: Sat, 09 Dec 2000 16:04:37 GMT
Subject: corba, soap, com, etc...
Message-ID: <pKsY5.58762$yR4.1496678@news1.rdc1.tx.home.com>

Hello Everyone,

I need some advice.  I want to write an api engine in python that will be
used as the back end to a gui.  The gui will not be written in python.  I
need a way to expose the api I create that is cross platform.  On Windows, I
would use COM.  Since I need cross platform, what are my options and what
are the advantages and disadvantages.  I know about CORBA, but I do not know
if CORBA can be used from python.  I have also heard about SOAP, but is it
only web based?

If there is a corba or soap interface to python, can you please point me to
it?

Your input is greatly appreciated!

Thanks!
Scott




From mfletch at tpresence.com  Thu Dec 21 14:38:06 2000
From: mfletch at tpresence.com (Mike Fletcher)
Date: Thu, 21 Dec 2000 14:38:06 -0500
Subject: Fatal Python error: PyThreadState_Get
Message-ID: <CC59868A0A76D311B1F50090277B321C1F3767@VRTMAIL>

Me too (would have taken me ages to track down), someone posted a hint to
comp.lang.python when I asked, and in retrospect, I guess I should have been
able to puzzle it out (if the Python DLL isn't initialised for your app,
there's no main thread created for your app), but something like: 
	...no current thread, possible conflicting Python library versions.

Might have been useful.  Enjoy yourself,
Mike

-----Original Message-----
From: Gordon Williams [mailto:g_will at cyberus.ca]
Sent: Thursday, December 21, 2000 1:15 PM
To: Mike Fletcher
Subject: Re: Fatal Python error: PyThreadState_Get


Thanks Mike,

It was an old Python module that I thought I had recompiled, but evidently
had not.

"Fatal Python error: PyThreadState_Get: no current thread" seems like a
funny error to indicate that there is a problem like this.  It would have
taken me hours/days to track this one down.

Regards,

Gordon Williams
...



From lvirden at cas.org  Sat Dec 30 07:24:10 2000
From: lvirden at cas.org (lvirden at cas.org)
Date: 30 Dec 2000 12:24:10 GMT
Subject: Python legalities
References: <3A2FC63D.652101A@accessone.com> <3A308881.7FBEBAD4@engcorp.com> <3A3120DF.9A9F3725@san.rr.com>
Message-ID: <92kk5a$c65$2@srv38.cas.org>

According to Darren New  <dnew at san.rr.com>:
:Peter Hansen wrote:
:> Somehow I doubt the distinction between "free" and commercial software
:> is really that clear.  Is it definitely, legally, contractually the case
:> that you are completely safe licensing software for money from a third
:> party? 
:
:Generally, that's one of the things that goes into the contract, yes. Not
:shrinkwrap contracts, mind, but real contracts where both parties agree.
:Someone can still come after you, but it's a much cheaper response to say
:"No, go talk to Joe" than to fight it yourself.
:
:The other choice, of course, is to buy your own insurance. I would imagine
:someone would be glad to insure you against that sort of stuff for a price.


Another path would be for free software to come with the same kind of
assurances as commercial software - that the software is judged free and
clear of such future legal issues and that, if not, the software identify
some agency who will 'take the heat'...

-- 
-- 
"See, he's not just anyone ... he's my son."  Mark Schultz
<URL: mailto:lvirden at cas.org> <URL: http://www.purl.org/NET/lvirden/>
Even if explicitly stated to the contrary, nothing in this posting


From fredrik at effbot.org  Wed Dec 13 12:15:24 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 13 Dec 2000 17:15:24 GMT
Subject: Tkinter newbie question
References: <slrn932i2t.b96.kc5tja@garnet.armored.net>
Message-ID: <M8OZ5.3816$Kd1.434902@newsb.telia.net>

Samuel A. Falvo II wrote:
> I've set up a basic user interface using Tkinter.  One of the controls is a
> listbox.  However, I cannot find any documentation on how to read the
> current state of the listbox (e.g., what, if any, items are currently
> selected).

selection = list.curselection()

for more info, see

http://www.pythonware.com/library/tkinter/introduction/listbox.htm
=> patterns (and methods)

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->





From thaiwey at my-deja.com  Wed Dec  6 20:35:18 2000
From: thaiwey at my-deja.com (thaiwey at my-deja.com)
Date: Thu, 07 Dec 2000 01:35:18 GMT
Subject: SSL socket in 2.0
References: <90lpjm$edp$1@nnrp1.deja.com> <90mcv1$9c$1@nnrp1.deja.com>
Message-ID: <90mpgm$avg$1@nnrp1.deja.com>

In article <90mcv1$9c$1 at nnrp1.deja.com>,
  thaiwey at my-deja.com wrote:
> In article <90lpjm$edp$1 at nnrp1.deja.com>,
>   thaiwey at my-deja.com wrote:
> > hi,
> >
> > i have been trying to compile the socket module with SSL support on
my
> > nt without much success. i was using the pcbuild.dsw in MSVC 6.0.
can
> > anyone advice me as to how i can set the necessary flags in the
> project
> > options/settings to turn on SSL.
> >
> > i have already compiled openssl-0.9.6 for my machine. i have also
> added
> > "USE_SSL" in the preprocessor definitions, paths to openssl in the
> > include directories, library path to openssl, and "libeay32.lib
> > ssleay32.lib" to the Object\library modules section. when
compiling, i
> > keep getting the following errors:
> >
> > Compiling...
> > socketmodule.c
> > G:\Python-2.0\Modules\socketmodule.c(2112) : error C2099:
initializer
> > is not a constant
> > G:\Python-2.0\Modules\socketmodule.c(2114) : warning C4047:
> > 'initializing' : 'int ' differs in levels of indirection from 'char
> [4]'
> > G:\Python-2.0\Modules\socketmodule.c(2115) : warning C4047:
> > 'initializing' : 'char *' differs in levels of indirection from
> > 'unsigned int '
> > G:\Python-2.0\Modules\socketmodule.c(2118) : warning C4047:
> > 'initializing' : 'int ' differs in levels of indirection from 'void
> > (__cdecl *)(struct _object *)'
> > G:\Python-2.0\Modules\socketmodule.c(2120) : warning C4047:
> > 'initializing' : 'int (__cdecl *)(struct _object *,struct _iobuf
*,int
> > )' differs in levels of indirection from 'struct _object *(__cdecl
*)
> > (struct _object *,char *)'
> > Error executing cl.exe.
> >
> > thanks.
> >
> > tw
> >
> > Sent via Deja.com http://www.deja.com/
> > Before you buy.
> >
>
> actually, i found out that i can change some code in socketmodule.c so
> that it compiles. i found the answer in faq
> (http://www.python.org/doc/FAQ.html). but after that, when i try to
use
> it, i keep getting this error:
>
> Traceback (most recent call last):
>   File "httplib.py", line 819, in ?
>     test()
>   File "httplib.py", line 805, in test
>     hs.connect(host)
>   File "httplib.py", line 660, in connect
>     self._conn.connect()
>   File "httplib.py", line 616, in connect
>     ssl = socket.ssl(sock, self.key_file, self.cert_file)
> TypeError: ssl, argument 1: expected socket, instance found
>
> does anyone know why?
>
> tw
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
>

great. found the problem. not sure if this solution is correct. but it
seems that socket.ssl expect a socket as the first argument. so instead
of passing in sock (a _socketobject) in line 616 of httplib.py, we
should be passing in sock._sock. after changing that line to:
    ssl = socket.ssl(sock._sock, self.key_file, self.cert_file)
the problem seemed fixed. hope this will be useful to someone else on
the net. thanks.

tw


Sent via Deja.com http://www.deja.com/
Before you buy.


From flip at sbox.tugraz.at  Tue Dec 12 06:02:43 2000
From: flip at sbox.tugraz.at (Philipp Jocham)
Date: 12 Dec 2000 11:02:43 GMT
Subject: HPUX install
References: <913ett$u0b$1@nnrp1.deja.com>
Message-ID: <9150kj$ipe$1@fstgss02.tu-graz.ac.at>

Brad <blosh at tsbrass.com> wrote:
> We are running HPUX 10.20 and I was wondering why I cant get python to
> install.

> ./configure seems to work

> make does not...it gives the following error

>   cd Parser ; make OPT="-O" VERSION="2.0" \  prefix="/usr/local"
> exec_prefix="/usr/local" all  cc -Ae -O -I./../Include -I.. -DHAVE_CONFIG_H
> -c pgenmain.c (Bundled) cc: warning 480: The -A option is available only with
> the C/ANSI C pro duct; ignored. (Bundled) cc: warning 480: The -O option is
> available only with the C/ANSI C pro duct; ignored. cpp: "pymem.h", line 7:
> error 4036: Can't open include file 'pyport.h'. *** Error exit code 1

> Stop.
> *** Error exit code 1

You are using the stock c-compiler. Try installing the
	HP C/ANSI C Developer's Bundle for HP-UX 10.20 (S800) 
or the latest gcc version. (http://hpux.connect.org.uk/)

Also deactivate thread-support with
	./configure --without-threads

if python crashes.

Philipp




From aleaxit at yahoo.com  Mon Dec 11 11:46:52 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 11 Dec 2000 17:46:52 +0100
Subject: Python2 distutils question: how to best "autoconfig"...?
Message-ID: <9130e60obc@news2.newsguy.com>

An issue came up in the GMPY project that looks like it
would need to be solved through some kind of 'automatic
configuration' -- compile a small auxiliary C program
(under exactly the same flags/options/whatever as those
Python has been compiled on, on this installation), run
it, check the results, and determine compilation flags
accordingly.

The specific issue is "how many bits of significance
does a Python float have".  (Or is there somewhere in
the Python headers a #define for this that I missed...?)
I first tried a rather naive algorithm (halving a
number that starts at 1.0, and keep adding it to 1.0
until the result equals 1.0), which, on Visual C++,
gives the expected result -- 53 bits of precision.

People using gcc (on Intel hardware) meanwhile seemed
to be measuring by this same algorithm '64 bits' of
precision -- which aren't really there for _most_
float computations... apparently gcc is sometimes
using the 80-bit "temporary" (extended precision)
format, which does have 64 bits of precision... but
most places it's down to 53, of course.

This matters (most particularly) when I'm trying to
get the 'heuristically best' rational number to
match a given floating-point number.  If I know the
real precision of the float is 53 or thereabouts, I
can get very nice results through a Stern-Brocot
tree (as suggested by Pearu Peterson) -- the resulting
rational number equals the 'exact fraction' that was
used for all 'small' numbers i and j, when I turn
float(i)/j into a rational (gmpy.mpq).  But if I
run under the misleading impression that the float
has 64 bits of precision, well, I end up with a huge,
'unreadable' rational-number.

Keith Briggs has very kindly supplied a simple C
implementation of Linnainmaa's algorithm to determine
a float's precision -- it *does* seem to work, but
he suspects it might depend on the optimization
setting and other flags of gcc, _unless_ one inserts
appropriate inline assembler-code to defeat gcc's
"helpful" (!) attempts to use 80-bits when it can.

Some "automatic-configuration" idea would appear to
be the best way out of this mess -- but now, I'm not
sure of how (if at all) the distutils are meant to
allow this.  Any guidance or suggestions would be
appreciated... thanks!


Alex





From tbradshaw at orchestream.com  Fri Dec 22 12:39:43 2000
From: tbradshaw at orchestream.com (t o b e)
Date: Fri, 22 Dec 2000 17:39:43 -0000
Subject: Embedding Python, PyImport_* etc etc...
Message-ID: <977506990.851496@lave>

Hi,

-=-

First of all Merry Xmas, everybody.. and a very Happy New Year to you all...
also apologies to the SWIG guys to whom this doesn't *directly* apply but
who I figured would be very likely to know or be interested in the answers
to this problem..

-=-

Well, I'm nearing the end of my embedding project and am now trying to tie
up all the loose ends here.. One of my primary goals is that we should ship
1 executable file only and to this end I have been trying to find a way to
bind those files normally residing externally in .py files into our
project...

>From this work arises a couple of questions:

My first take on the problem was to add the .py files to my project as
strings and then do this to them (after calling Py_Initialise(), of course
*):

    pMyModule = Py_CompileString(strMyModule, "import.tmp", Py_file_input);
    if (pMyModule) PyImport_ExecCodeModule ("MyModule", pMyModule); else
PyErr_Print()

For modules with no dependencies this seems to work fine.. even for modules
that import other modules this worked as long as I was careful about the
order in which I did things. However,  I then started running into circular
dependencies (e.g. os.py and ntpath.py which import each other) for which
there was no right order in which to make the calls. I found that calling
PyImport_AddModule() for everything before I began importing alleviated this
problem but now I'm having problems with sre_constants.py (dependency of
sre) which doesn't seem to be exporting all of its contstants properly
(could this be becuase it is imported using the from * syntax)...

Anwyay.. the long and the short of it is.. does it look like I'm going about
this the right way and if not, which do you think would be the correct
approach ?

Thanks...

t o b e
Orchestream Plc,
London, UK.





From dsh8290 at rit.edu  Wed Dec 27 20:54:33 2000
From: dsh8290 at rit.edu (D-Man)
Date: Wed, 27 Dec 2000 20:54:33 -0500
Subject: testing
Message-ID: <20001227205433.B11725@westsidecnc.rh.rit.edu>

testing



From tim.one at home.com  Wed Dec 20 16:02:12 2000
From: tim.one at home.com (Tim Peters)
Date: Wed, 20 Dec 2000 16:02:12 -0500
Subject: ANNOUNCE: soundex.py
In-Reply-To: <3A4119C3.2E5E54E1@uswest.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCCENFIEAA.tim.one@home.com>

[posted & mailed]

[Bjorn Pettersen]
> Thanks Skip!  I was actually using the soundex module in one of my less
> frequently used apps so I hadn't realized that it was gone yet :-)
>
> Could I get you to put an explicit licence on it (it would be
> much more usable to me if it had a licence and if the licence wasn't
> GPL <wink>).

Note too that Sean True wrote a Python implementation of the
more-sophisticated NYSIIS algorithm, available at

    http://metagram.webreply.com/downloads/

(it's linked to from Parnassus if you forget that).  It's not encumbered.

Whatever stuff of mine Skip reused was released to the public domain.

if-soundex-is-the-answer-the-question-was-ill-advised<wink>-ly
    y'rs  - tim




From sbrunning at bigfoot.com  Mon Dec  4 05:02:18 2000
From: sbrunning at bigfoot.com (Simon B.)
Date: Mon, 04 Dec 2000 10:02:18 GMT
Subject: Programming with Python
Message-ID: <90fq39$jgq$1@nnrp1.deja.com>

I came across a Python book this weekend that I haven't heard of
before - Programming with Python, by Tim Altom
(see <http://www.prima-tech.com/book.asp?id=3942>). Anyone have this?
Any good?

--
Simon Brunning
I think... I think it's in my basement. Let me go upstairs and check. -
M.C. Escher


Sent via Deja.com http://www.deja.com/
Before you buy.


From aleaxit at yahoo.com  Mon Dec  4 10:43:45 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 4 Dec 2000 16:43:45 +0100
Subject: Printable time
References: <7m9n2tcp5ff10o97sr1d57qokakjqsa2m9@4ax.com>
Message-ID: <90ge3r02u1k@news2.newsguy.com>

"Dale Strickland-Clark" <dale at out-think.NOSPAMco.uk> wrote in message
news:7m9n2tcp5ff10o97sr1d57qokakjqsa2m9 at 4ax.com...
> Is there a more succinct way of getting the current local time as text
than:
>
> time.strftime('<format>', time.localtime(time.time()))

time.ctime(time.time())

will do, but you get no control over the format -- you do need strftime
if applying a format of your choosing is important, I think.


Alex





From quinn at dinar.ugcs.caltech.edu  Sat Dec 16 21:01:08 2000
From: quinn at dinar.ugcs.caltech.edu (Quinn Dunkan)
Date: 17 Dec 2000 02:01:08 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <r0pzoi11lhz.fsf@black.cs.washington.edu> <5jvZ5.3588$Kd1.418770@newsb.telia.net> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com> <kuu28971kp.fsf@lasipalatsi.fi> <3A380E4E.8537A0FB@san.rr.com> <91a1vr02b02@news2.newsguy.com> <Co9_5.25465$x6.13756579@news2.rdc2.tx.home.com> <91bbq102mtk@news1.newsguy.com> <rWd_5.26820$x6.14119694@news2.rdc2.tx.home.com> <91d5el0me7@news1.newsguy.com> <rOr_5.31120$x6.15241856@news2.rdc2.tx.home.com>
Message-ID: <slrn93o7j4.tu0.quinn@dinar.ugcs.caltech.edu>

On Fri, 15 Dec 2000 16:38:47 GMT, Rainer Deyke <root at rainerdeyke.com> wrote:
>Yes, in-place modification should be supported.  I had something like this
>in mind:
>
>class UserInt:
>  def __init__(self, value):
>    self.data = value
>  def __add__(self, other):
>    if isinstance(other, UserInt):
>      return self.__class__(self.data + other.data)
>    else:
>      return self.__class__(self.data + other)
>  __radd__ = __add__
>  def __iadd__(self, other):
>    if isinstance(other, UserInt):
>      self.data += other.data
>    else:
>      self.data += other
>  # Repeat for all other operations
>
>UserInt as written is more properly UserNumber, since it supports longs ints
>and floats.

Just as a matter of style, would it be superior to add a __coerce__ method
that knows how to coerce ints, longs, and floats into UserInts, or make an
explicit check on each method?  __coerce__ saves a lot of typing, but I'm
generally wary of implicit things like coercion.  When I implemented a class
like this (a vector, actually), I used __coerce__, but after trying to wrap my
brain around the control flow, I was thinking maybe I wouldn't next time, even
though it seems to work fine.

Is this fear unfounded?  Is there a better way?


From aivkov at my-deja.com  Wed Dec  6 11:09:58 2000
From: aivkov at my-deja.com (aivkov at my-deja.com)
Date: Wed, 06 Dec 2000 16:09:58 GMT
Subject: Python: what all the hype is about?
Message-ID: <90locf$d80$1@nnrp1.deja.com>

Dear all,

Excuse my ingnorance in the subject area. Can someone explaing me very
briefly, why one should choose Python over Java or is it better to
compare it with Javascript/ECMA?

I hope I am not trying to compare apples with oranges.

Thanks,
Anatoliy.


Sent via Deja.com http://www.deja.com/
Before you buy.


From thomas at xs4all.net  Thu Dec 28 17:14:07 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Thu, 28 Dec 2000 23:14:07 +0100
Subject: Interesting behaviour of the assignment
In-Reply-To: <92fhsr01u88@news1.newsguy.com>; from aleaxit@yahoo.com on Thu, Dec 28, 2000 at 03:12:57PM +0100
References: <92f4gk$qk6$2@news.nuri.net> <92fhsr01u88@news1.newsguy.com>
Message-ID: <20001228231407.D1811@xs4all.nl>

On Thu, Dec 28, 2000 at 03:12:57PM +0100, Alex Martelli wrote:

> Thus, limit your use of 'is' to cases in which you know
> "object-identity" is indeed meaningful: None, lists, tuples,
> dictionaries, object instances, type-objects, etc.
> (Actually, in the case of None, the semantics of is and
> == are identical, but 'is' can be a tiny little bit faster).
> 
> In practice I find myself using == almost by reflex, e.g.
>     if type(x)==type(0):
> when I KNOW I could safely and reliably write, instead:
>     if type(x) is type(0):
> and perhaps gain a 1/1000th of a second if this is at
> the heart of a heavily-executed loop... '==' is just a
> by-far-too-more-frequent case, except perhaps in the
> specific idion "x is None"!-)

Well, Guido-be-thanked, '==' does an identity check ('is') first. Consider:

>>> class X:
...     def __cmp__(self, other):
...             return -1
...
>>> x1=X()
>>> x1 == x1
1

In other words, 'x == y' first does 'x is y', and if it's true, the result
of the operation is true. Only if 'x is y' is false will it start the
(possibly very expensive) comparison-by-contents operation. The side effect
is that you can never build an object that compares as unequal to itself.
(as above.)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From MarkH at ActiveState.com  Wed Dec  6 07:13:03 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Wed, 06 Dec 2000 12:13:03 GMT
Subject: activex scripting and debugging with Python win32com
References: <90irub$24b$1@troll.powertech.no> <3A2D8317.20807@ActiveState.com>
Message-ID: <3A2E2C04.9020401@ActiveState.com>

Mark Hammond wrote:
...

Well, it appears a kind lurker out there named Grzegorz Makarewicz had 
the answer!  He mailed me the following patch...

I admit I haven't actually tested it yet, but the errors are so obvious 
I am a little embarrassed about posting it ;-)

Mark.


   RCS file: /home/cvsroot/PyWin32/com/win32comext/axdebug/adb.py,v
   retrieving revision 1.2
   retrieving revision 1.3
   diff -u -r1.2 -r1.3
   --- adb.py	2000/01/06 04:14:06	1.2
   +++ adb.py	2000/12/06 12:06:02	1.3
   @@ -68,7 +68,7 @@
    	def canonic(self, fname):
    		if fname[0]=='<':
    			return fname
   -		return bdb.Bdb.canonic(fname)
   +		return bdb.Bdb.canonic(self, fname)

    	def reset(self):
    		traceenter("adb.reset")
   @@ -115,7 +115,7 @@
    		if self.breakFlags==axdebug.APPBREAKFLAG_DEBUGGER_HALT:
    			self.breakReason = axdebug.BREAKREASON_DEBUGGER_HALT
    			return 1
   -		rc = abd.Adb.break_anywhere(self, frame)
   +		rc = bdb.Bdb.break_anywhere(self, frame)
    		tracev("break_anywhere",_dumpf(frame),"returning",rc)
    		return rc



From fluxent at yahoo.com  Tue Dec  5 10:18:54 2000
From: fluxent at yahoo.com (Bill Seitz)
Date: Tue, 05 Dec 2000 15:18:54 GMT
Subject: CGI File Upload
References: <p8ce0tcgtt2bksb3adrf9cfr2i8egr4ff5@4ax.com> <8u7fn7$jpf$1@la-mail4.digilink.net>
Message-ID: <90j10o$5p5$1@nnrp1.deja.com>

In article <8u7fn7$jpf$1 at la-mail4.digilink.net>,
  "Pete Shinners" <pete at visionart.com> wrote:
> "Simon Faulkner" <news at titanic.co.uk> wrote
> > I would like users to be able to upload .jpeg files via a web page
> > <input type=file>
> >
> > What python code do I need to be able to save the file to a
directory
> > on the hard disk?
>
> i have some code that does this.. you'll probably want to clean
> it to suit your own purposes, but here it is. at some point i
> wanted to add PIL support to resize the images to a certain
> resolution and convert them to be JPG, but that day hasn't come
> yet... heh. note it also allows the user to save a comment as a
> .TXT file. anyways, you can use it as a start for whatever you want

Hmm, this works for me with text data (though the file length is
slightly different: maybe line-end changes from Win to *nix?). With
binary it's totally off. Uploading a JPEG I consistently only get 109
bytes written to the file system. With GIFs I consistently get ~500
bytes (but not always exactly the same number). With an MsWord file I
get a bigger file written, but still not the whole thing (60kb out of
75kb).

What do I do next?


Sent via Deja.com http://www.deja.com/
Before you buy.


From aleaxit at yahoo.com  Tue Dec 26 13:40:14 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 26 Dec 2000 19:40:14 +0100
Subject: Python and Napster
References: <sx226.18970$RC1.511898@newsread1.prod.itd.earthlink.net>
Message-ID: <92ap0g12b6r@news1.newsguy.com>

"Doug Farrell" <dfarrell01 at earthlink.net> wrote in message
news:sx226.18970$RC1.511898 at newsread1.prod.itd.earthlink.net...
> Hi,
> Is there a way to build a 'Napster' like client in Python? What I'm really
> looking for is a way to have a cross-platform (Windows, Mac, Sun) way of
> having users share files, no necessarily music files, from machine to
> machine. I'd like to do this in Python as I'm the only Windows programmer
at
> my company and we'd like to support the other two platforms I mentioned.
> Maybe I should be looking at Java?

Python would probably let you be more productive, I suspect.  It sure has
all the network functionality you need; in your shoes, I'd look into pyro --
building file exchange on top of it should be easier than on sockets or on
protocols such as HTTP.


Alex





From gerson.kurz at t-online.de  Wed Dec 27 11:10:02 2000
From: gerson.kurz at t-online.de (Gerson Kurz)
Date: Wed, 27 Dec 2000 16:10:02 GMT
Subject: newbie : array initialization, anyone ?
Message-ID: <3a4a1352.9818548@news.t-online.de>

you can initialize a fixed-size array using something like

	array = [0] * 10

now, one would think that you can initialize a multidimensional array
like this:

	multidim = [ [0] * 10 ] * 3

but now, all 3 elements in multidim are references to the same single
array, so that if you type

	multidim[0][0] = 1
	print multidim 

you actually get three 1 in all those 0s. Where did my thinking go
wrong along these lines ?


From m.faassen at vet.uu.nl  Thu Dec 21 20:30:13 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 22 Dec 2000 01:30:13 GMT
Subject: Reality check PYZOMA = python / zope for  Mac ?
References: <t44t20nmcgena7@corp.supernews.com> <slrn945648.96k.scarblac@pino.selwerd.nl>
Message-ID: <91uar5$fj0$2@newshost.accu.uu.nl>

Remco Gerlich <scarblac at pino.selwerd.nl> wrote:
> Jason Cunliffe <jasonic at nomadicsltd.com> wrote in comp.lang.python:
>> I am wondering what the prospects are for Python and Zope on Macintosh:

> I don't know anything about this (I don't use Zope and I don't have a Mac)
> but I had never expected anyone to want to run Zope on a Mac. Are you really
> using a Mac as a web server?!

> I thought Macs were pretty cool (and certainly MacOS X) but I thought Zope
> was the kind of application that would be completely irrelevant for Macs.

Since MacOS X is really a BSD Unix, it should run Zope just fine. I think I
heard this mentioned in fact on some Zope mailing list. I don't know much 
about this either, though; check with the Zope mailing lists.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From shindich at my-deja.com  Fri Dec  8 22:41:28 2000
From: shindich at my-deja.com (shindich at my-deja.com)
Date: Sat, 09 Dec 2000 03:41:28 GMT
Subject: Python wrapper for C++ STL?
References: <3A31760E.D157EB54@erols.com> <C0gY5.126062$U46.4232859@news1.sttls1.wa.home.com> <90s3fk$j8p$1@nnrp1.deja.com> <fUgY5.126218$U46.4241063@news1.sttls1.wa.home.com>
Message-ID: <90s9l9$njd$1@nnrp1.deja.com>

In article <fUgY5.126218$U46.4241063 at news1.sttls1.wa.home.com>,
  "Greg Jorgensen" <gregj at pobox.com> wrote:
> <shindich at my-deja.com> wrote
>
> > With all due respect, what would be wrong with a definition like
> > std::map<PyObject*, PyObject*>? Obviously there is no type safety
here.
> > But Python is not type safe.
>
> How does that help the Python programmer? Using C++ and STL to write
Python
> modules is one thing, but the original question was about wrapping
the STL
> so it could be used from Python. STL is not a set of callable
libraries that
> can be wrapped--it's a collection of template definitions that are
> transformed into C++ by the compiler. Once the C++ module is compiled
the
> STL functionality is invisible and irrelevant to the Python user.
>
> In other words, writing a wrapper for a C library function like printf
(), or
> an OS function like TickCount(), makes sense. But writing a wrapper
around a
> compiler feature doesn't make sense. Take STL out of it and imagine
writing
> a wrapper around the C/C++ for loop. It's possible, but it doesn't
extend
> the C/C++ for loop into Python.
>
> --
> Greg Jorgensen
> Deschooling Society
> Portland, Oregon, USA
> gregj at pobox.com
>
I am not saying that there is any point in wrapping STL for Python. All
I'm saying is that STL can be used to write C++ extensions that
implement specific algorithms.


Sent via Deja.com http://www.deja.com/
Before you buy.


From MarkH at ActiveState.com  Wed Dec 20 23:52:29 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Thu, 21 Dec 2000 04:52:29 GMT
Subject: Events in win32com
References: <Y1g=Op6IfWSoJn1ISFtOyuvqwBb3@4ax.com> <FY0=OlZFLy71DtnK62u5TEPlhOts@4ax.com> <3A40060B.10504@ActiveState.com> <cYxAOmxi9FgNERgBo29zuCl5c2h1@4ax.com>
Message-ID: <3A418B5B.5010001@ActiveState.com>

Paul Moore wrote:

> Am I doing something wrong, or is this not expected to work?

It is expected to work, and I can't see what you are doing wrong.

If you can work out a very simple case that fails (ie, suiltable for 
inclusion in the test suite) I would look at it, but without knowing the 
exact signatures in question, or knowing if Python is finding the sigs 
in the makepy file etc, I really have no idea...

Mark.



From aahz at panix.com  Wed Dec  6 21:26:40 2000
From: aahz at panix.com (Aahz Maruch)
Date: 6 Dec 2000 18:26:40 -0800
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu>
Message-ID: <90msh0$rca$1@panix2.panix.com>

In article <3A2EE2AF.AD8A51E2 at caltech.edu>,
Nathan Gray  <nospam at caltech.edu> wrote:
>
>What a rude response to a polite and legitimate question.

Well, you're one-third right on your adjectives.  Rude, yes; polite and
legitimate, no.  Perhaps you haven't been on Usenet long enough (or in
the wrong places) to see the hordes of students who expect other people
to do their work for them.  Laziness of the highest order, is what it
is, and many long-term Usenetters see it as rudeness.

I'm sure you find it quite shocking to discover that rudeness is the
frequent response to rudeness on Usenet.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

What if there were no rhetorical questions?  --Aahz


From cg at schlund.de  Tue Dec  5 09:20:54 2000
From: cg at schlund.de (Carsten Gaebler)
Date: Tue, 05 Dec 2000 15:20:54 +0100
Subject: "inherited" keyword in Python?
References: <3A2B9EFA.3A64C13B@schlund.de> <90gevs02uu1@news2.newsguy.com>
Message-ID: <3A2CF9C6.F73794CC@schlund.de>

Alex Martelli wrote:

> "Carsten Gaebler" <cg at schlund.de> wrote in message
> news:3A2B9EFA.3A64C13B at schlund.de...

> > Does Python know something like the "inherited" keyword from Borland
> > Pascal? 

> class SomeClass(ABaseClass,AnotherOne,YetAnother):
>     __super = AnotherOne
>     # etc, etc
>     def somemethod(self, foo, bar):
>         print "before"
>         __super.somemethod(self, foo, bar)
>         print "after"


Yes, that'll do, thanks!

cg.


From dsh8290 at rit.edu  Thu Dec 21 23:49:18 2000
From: dsh8290 at rit.edu (D-Man)
Date: Thu, 21 Dec 2000 23:49:18 -0500
Subject: compiling BeOpen-Python-2.0-1.src.rpm
In-Reply-To: <v04220809b66843de8ce6@[192.168.11.33]>; from Blair.Lowe@compeng.net on Thu, Dec 21, 2000 at 04:34:02PM -0700
References: <v04220809b66843de8ce6@[192.168.11.33]>
Message-ID: <20001221234918.B20586@westsidecnc.rh.rit.edu>

I think that you are missing expat, or some part of it.  expat is the
xml parser written by James clark in C and is what pyexpat wraps.  It
is distributed separate from python.

HTH,
-D

On Thu, Dec 21, 2000 at 04:34:02PM -0700, Blair Lowe wrote:
> Hi,
> 
> gcc -fpic  -I/usr/local/include/xmlparse -g -O2 -Wall 
> -Wstrict-prototypes -I./..
> /Include -I.. -DHAVE_CONFIG_H -c ./pyexpat.c
> ./pyexpat.c: In function `newxmlparseobject':
> ./pyexpat.c:639: warning: implicit declaration of function `XML_ParserCreateNS'
<snip>



From tdelaney at avaya.com  Sun Dec 17 19:03:31 2000
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 18 Dec 2000 11:03:31 +1100
Subject: while true: !!!
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B0FBCBB@natasha.auslabs.avaya.com>

This is precisely the reason this thread exists ...

The idiom is actually

while 1:

	do something

	if condition:
		break

	do something else

Basically, it's for dealing with situations where you can't tell at the
beginning or end of a loop if the loop is completed - only somewhere in the
middle. The canonical example is

while 1:

	try to read a line from a file

	if we didn't get a line:
		break

	do something with the line

There are many ways of dealing with this, but they all have compromises - in
most cases either duplication of code, or an assumption of memory
availability.

Tim Delaney
Avaya Australia
+61 2 9352 9079

> -----Original Message-----
> From: D-Man [mailto:dman at res137a-039.rh.rit.edu]
> Sent: Sunday, 17 December 2000 4:33 AM
> To: morpheus at here.not.there
> Cc: python-list at python.org
> Subject: Re: while true: !!!
> 
> 
> On Sat, Dec 16, 2000 at 04:45:07PM +0000, Steve Lamb wrote:
> > 
> > while 1:
> >   if cond:
> >     break
> >   do something
> > 
> 
> why not
> 
> while ( !cond ) :
> 	do something
> 
> It involves less typing and more importantly shows the next reader
> when the loop should end.  I strongly dislike the fake infinite loops
> like "while 1 :".
> 
> -D
> 
> 
> PS.  I haven't been following most of this thread, so forgive me if
> this has already come up.
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 



From gregj at pobox.com  Fri Dec  8 21:35:55 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Sat, 09 Dec 2000 02:35:55 GMT
Subject: Python wrapper for C++ STL?
References: <3A31760E.D157EB54@erols.com> <C0gY5.126062$U46.4232859@news1.sttls1.wa.home.com> <90s3fk$j8p$1@nnrp1.deja.com>
Message-ID: <fUgY5.126218$U46.4241063@news1.sttls1.wa.home.com>

<shindich at my-deja.com> wrote

> With all due respect, what would be wrong with a definition like
> std::map<PyObject*, PyObject*>? Obviously there is no type safety here.
> But Python is not type safe.

How does that help the Python programmer? Using C++ and STL to write Python
modules is one thing, but the original question was about wrapping the STL
so it could be used from Python. STL is not a set of callable libraries that
can be wrapped--it's a collection of template definitions that are
transformed into C++ by the compiler. Once the C++ module is compiled the
STL functionality is invisible and irrelevant to the Python user.

In other words, writing a wrapper for a C library function like printf(), or
an OS function like TickCount(), makes sense. But writing a wrapper around a
compiler feature doesn't make sense. Take STL out of it and imagine writing
a wrapper around the C/C++ for loop. It's possible, but it doesn't extend
the C/C++ for loop into Python.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From zeitlin at seth.lpthe.jussieu.fr  Fri Dec 22 18:50:28 2000
From: zeitlin at seth.lpthe.jussieu.fr (Vadim Zeitlin)
Date: 22 Dec 2000 23:50:28 GMT
Subject: A Suggestion for Python Colon Syntax
References: <mailman.977446446.16022.python-list@python.org>
Message-ID: <slrn947q64.jkg.zeitlin@seth.lpthe.jussieu.fr>

On Thu, 21 Dec 2000 19:53:52 -0500, Tim Peters <tim.one at home.com> wrote:
>[William Djaja Tjokroaminata]
>> I saw many replies that give the reasons why the colons are
>> *needed*.
>
>Not mine:  it told you why the colons are *there*.  It's a human-factors
>thing, not a tech thing.  Desirable, not necessary.

 Isn't it desirable (although not necessary) to attract more people to Python?
IMveryHO any simplification of the syntax without any bad consequences (are
there any in this case?) would be welcome from this point of view.

 BTW, what was the real rationale for adding composed assignment operators
(+=) to Python 2.0 (which I'm really extremely glad to have now)? Wasn't it
meant to make Python more attractive to the people coming from C-lie
languages? And, if by chance it was so, why not continue making it even more
simple to use to the very same people?

 Regards,
VZ

--
def AreColonsReallyNeeded()
SyntaxError: invalid syntax


From pbjorn at uswest.net  Sun Dec 31 16:11:11 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Sun, 31 Dec 2000 14:11:11 -0700
Subject: Recursive inner function... How?
Message-ID: <3A4FA0EF.BF94F4BD@uswest.net>

Ok, this is what I want to do:

    def outer():
        def inner():
            inner()
        inner()

    outer()

i.e. I call outer, outer has an auxiliary sub function called inner that
recurses to infinity (that is not the problem). The above gives the
expected result that inner is not defined inside inner. Inconvenient,
but we know the drill:

    def outer():
        def inner(inner=inner):
            inner()
        inner()

    outer()

however, in 2.0 that gives an "UnboundLocalError: Local variable 'inner'
referenced before assignment". The following works of course, but is
rather repetitive...

    def outer():
        def inner(inner):
               inner(inner)
        inner(inner)

    outer()

Please tell me I'm doing something wrong?

-- bjorn




From knu at iDaemons.org  Fri Dec 15 10:52:27 2000
From: knu at iDaemons.org (Akinori MUSHA)
Date: Fri, 15 Dec 2000 15:52:27 GMT
Subject: comp.lang.ruby (Re: Smalltalk and Python)
In-Reply-To: <3A3A3A58.3917FA57@gte.net>
References: <3A2FFB8B.2030408@zoologic.com>
	<915qlh$89fa$1@nntp6.u.washington.edu>
	<917g5o$qh$1@news.rz.uni-karlsruhe.de>
	<55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com>
	<917r2b$4n7$1@news.rz.uni-karlsruhe.de>
	<Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org>
	<mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com>
	<86d7ewrqat.wl@archon.local.idaemons.org>
	<tcmay-19C054.14031113122000@news.got.net>
	<tcmay-FCA22F.14073713122000@news.got.net>
	<918v7a02scn@news1.newsguy.com>
	<tcmay-04E80C.16181313122000@news.got.net>
	<91a1ed02a4j@news2.newsguy.com>
	<3A389B1A.4F33FFD4@cfmu.eurocontrol.be>
	<91ae8s02llf@news2.newsguy.com>
	<3A397778.2E6D84D3@cosc.canterbury.ac.nz>
	<91c7r4$7a0$1@slb1.atl.mindspring.net>
	<3A3A3A58.3917FA57@gte.net>
Message-ID: <86lmthvg07.wl@archon.local.idaemons.org>

Hi guys,

Please post your messages on this thread to comp.lang.ruby if you want
to talk with Ruby folks all over the world in English.  comp.lang.ruby
is so active that you could well hear what Ruby language designers and
developers think.


Note that fj.comp.lang.ruby is a Japanese language newsgroup and thus
not suitable to discuss issues in English language.


Thank you for understanding.

-- 
                     /
                    /__  __            Akinori.org / MUSHA.org
                   / )  )  ) )  /     FreeBSD.org / Ruby-lang.org
Akinori MUSHA aka / (_ /  ( (__(  @ iDaemons.org / and.or.jp

"We're only at home when we're on the run, on the wing, on the fly"



From samschul at pacbell.net  Wed Dec  6 16:04:27 2000
From: samschul at pacbell.net (Sam Schulenburg)
Date: Wed, 06 Dec 2000 21:04:27 GMT
Subject: win32api SetComputerName ?
References: <3A2E90BF.4020204@pixelsystems.com>
Message-ID: <90m9km$t73$1@nnrp1.deja.com>

I have also found missing win32ui, and win32api calls i.e. GetDosDevice
(). I have used calldll, and windll.py to access any missing api
functions. The only problem is that calldll has to be recompiled for
Python20.

Sam Schulenburg

In article <3A2E90BF.4020204 at pixelsystems.com>,
  Patrick Blanchette <pblanchette at pixelsystems.com> wrote:
> I,
>    It seems that the function SetComputerName doesn't exist in the
> win32all-135.exe release but the function GetComputerName exist.  I'd
> like to know if the method SetComputerName will be added soon?
>    thanks
>
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From grante at visi.com  Tue Dec  5 12:15:34 2000
From: grante at visi.com (Grant Edwards)
Date: Tue, 05 Dec 2000 17:15:34 GMT
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Dec  4)
References: <8C419A8040EAFDCB.1774103991763D9E.2F034C54E78FF82C@lp.airnews.net> <90iqs8$n6k$1@newsreaderm1.core.theplanet.net>
Message-ID: <Wo9X5.2487$%Q3.342377@ptah.visi.com>

In article <90iqs8$n6k$1 at newsreaderm1.core.theplanet.net>, Franz GEIGER wrote:

>>     Alex Martelli shows a clever way to overload a method based on
>>     number of arguments.
>>         http://deja.com/=dnc/getdoc.xp?AN=975947338
>
>This link leads to "Invalid document request". How is the right one?

My bad. 

I screwed up and copied the wrong number from Deja for three of
the links.  Sorry about that.  The corrected links are:

Andrew H. Chatham announces Python bindings for Ogg/Vorbis (MP3
patent/licensing free replacement for MP3) -- and requests help
with dlopen().
    http://deja.com/=dnc/getdoc.xp?AN=699163645

Alex Martelli shows a clever way to overload a method based on
number of arguments.
    http://deja.com/=dnc/getdoc.xp?AN=700230993

Johannes Nix releases Python wrappers for ALSA (Advanced Linux
Sound Architecture) that enable real-time audio processing.
    http://deja.com/=dnc/getdoc.xp?AN=699878280

-- 
Grant Edwards                   grante             Yow!  RHAPSODY in Glue!
                                  at               
                               visi.com            


From clarence at netlojix.com  Wed Dec 13 10:54:37 2000
From: clarence at netlojix.com (Clarence Gardner)
Date: Wed, 13 Dec 2000 07:54:37 -0800
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <WntZ5.3566$Kd1.414873@newsb.telia.net>
Message-ID: <976722991.1660122408@news.silcom.com>

On Tue, 12 Dec 2000, Fredrik Lundh wrote:
>
>All Python objects has the following:
>
>    -- some content
>    -- an unique identity (a integer, returned by id(x))
>    -- a type (returned by type(x))
>
>You cannot change the identity.
>
>You cannot change the type.
>
[snip]

Nice summary, Fredrik!

I have no idea if it would have been such an enjoyable read if I didn't
already understand Python :)


From peter at engcorp.com  Sun Dec 31 00:35:45 2000
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 31 Dec 2000 00:35:45 -0500
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <05C6FUhLDNUU-pn2-tuTnzOCokxem@bill> <Sgw36.208288$U46.6385684@news1.sttls1.wa.home.com> <92m9pm$d19$1@nnrp1.deja.com> <B3z36.122$9s1.18804@e420r-atl2.usenetserver.com>
Message-ID: <3A4EC5B1.8912E571@engcorp.com>

Kragen Sitaker wrote:
> 
> The only time I wish I wasn't subject to Python's indentation rules is
> when I can't outdent my "if finished(): break" statements like I do in
> C.

But why would you want to outdent it?  It's *part* of the loop, not a
following statement.  Identation should match the block structure of the
code, and that statement lies _within_ the block.


From qdlaty at wielun.dhs.org  Thu Dec 28 08:04:20 2000
From: qdlaty at wielun.dhs.org (QdlatY)
Date: Thu, 28 Dec 2000 14:04:20 +0100
Subject: Is this a BUG?
References: <3A4B0C5E.1E7976D2@wielun.dhs.org> <3a4b314d@news.edunet.ru>
Message-ID: <3A4B3A54.7DA3711C@wielun.dhs.org>

Hello!


Like always, You're right.

Thanks for trouble.;)


Yours
QdlatY


From aahz at panix.com  Mon Dec 18 16:58:52 2000
From: aahz at panix.com (Aahz Maruch)
Date: 18 Dec 2000 13:58:52 -0800
Subject: [Q] Are Exceptions used that much in practice?
References: <3A35D546.A4731B5B@enteract.com> <3A36C6E3.26181E3F@enteract.com> <FJBZ5.19773$x6.11272365@news2.rdc2.tx.home.com> <Pine.GSO.4.21.0012122229470.19682-100000@y.glue.umd.edu>
Message-ID: <91m1as$e9j$1@panix2.panix.com>

In article <Pine.GSO.4.21.0012122229470.19682-100000 at y.glue.umd.edu>,
Roy Katz  <katz at Glue.umd.edu> wrote:
>
>If you enclose statements in a try/except clause, and assuming no
>exception is thrown, what are the runtime penalties?

Minimal, from what I've heard, but I don't feel like doing any
benchmarking right now.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

This is a signature anti-virus. 
Please stop the spread of signature viruses!


From tseaver at starbase.neosoft.com  Sun Dec 10 19:32:42 2000
From: tseaver at starbase.neosoft.com (Tres Seaver)
Date: 10 Dec 2000 18:32:42 -0600
Subject: sgmllib nit
References: <eyErVLAEHlM6EwA4@jessikat.fsnet.co.uk> <EAsY5.2808$O%.435892@newsc.telia.net> <JagDRIAEbuM6EwxY@jessikat.fsnet.co.uk> <y_IY5.2910$O%.457391@newsc.telia.net>
Message-ID: <2723C33941E68F6B.0B574026DB340C81.D1991EEB23A2AC3D@lp.airnews.net>

In article <y_IY5.2910$O%.457391 at newsc.telia.net>,
Fredrik Lundh <fredrik at effbot.org> wrote:
>
>...so if DTML expects
>
>    <dtml-var standard_html_header>
>
>to mean
>
>    <dtml-var standard_html_header=''>
>
>instead of
>
>    <dtml-var standard_html_header='standard_html_header'>
>
>it's neither SGML, HTML, nor XML.

In DTML,

  <dtml-var standard_html_header>

is actually a shortcut for

  <dtml-var name="standard_html_header">

>
>> Which parser should one use to analyse dtml which can contain tags like
>> <dtml-var standard_html_header>?
>
>dtmllib?
>
>I suggest asking the Zope folks if/why they're using an embraced
>and extended syntax, and what parser they recommend...

This version of DTML syntax was intended to be friendly to
"HTML tools", which can be taught to recognize the handful of
DTML tags.  The "entity-reference" DTML syntax allows cleaning
up some parsing issues:  instead of

  <a href="<dtml-var absolute_url>">

one can type

  <a href="&dtml-absolute_url;">

Note that we are as well-aware as anyone of the problematic nature
of DTML, which succeeds at its main goal (easing the generation of
HTML against an object-space), while imposing unfortanate burdens
on those who would integrate DTML with other tools.  We are working
directly with the folks from HiperLogica on their proposal to create
fully-XHTML-compliant templates as an eventual replacement for DTML:

  http://dev.zope.org/Wikis/DevSite/Projects/HiperDom/FrontPage

Tres.

Tres.
-- 
===============================================================
Tres Seaver                                tseaver at digicool.com
Digital Creations     "Zope Dealers"       http://www.zope.org


From erno-news at erno.iki.fi  Tue Dec 12 16:48:54 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 12 Dec 2000 23:48:54 +0200
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <r0pzoi11lhz.fsf@black.cs.washington.edu> <5jvZ5.3588$Kd1.418770@newsb.telia.net> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com>
Message-ID: <kuu28971kp.fsf@lasipalatsi.fi>

In article <3A368E11.20C17E92 at san.rr.com>, Darren New
<dnew at san.rr.com> writes:

| Having 
|  x = 2 + 2
|  print x
| result in anything except "4" is probably a *really* bad idea.
| AFAIK, only FORTH and really old Fortrans allow this.

agreed - but this is not something (sane) a mutable integer object
would break.

  -- erno


From peter at schneider-kamp.de  Sun Dec 24 05:21:03 2000
From: peter at schneider-kamp.de (Peter Schneider-Kamp)
Date: Sun, 24 Dec 2000 11:21:03 +0100
Subject: Speed of Python 1.5.x vs 2.0
References: <4vj54t4v0hvjsnstfmtmf36q9irotp59sj@4ax.com>
	 <mailman.977475665.3482.python-list@python.org>
	 <dud64ts8mb7mlc41kht410q5g87ekb2nc5@4ax.com> <91vfc6029r5@news2.newsguy.com>
Message-ID: <3A45CE0F.F92E256E@schneider-kamp.de>

Hei Alex!

If you used the code you posted, I'm not surprised by
the results. In the first module you have:

> print "Using map and lambda:     %f" % stopWatch(withMap)
> print "Using list-comprehension: %f" % stopWatch(withMap)

And in the second again:

> print "Using map and lambda:     %f" % stopWatch(withMap)
> print "Using list-comprehension: %f" % stopWatch(withMap)

merry-xmas-ly y'rs
Peter



From fredrik at effbot.org  Thu Dec 14 13:09:54 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Thu, 14 Dec 2000 18:09:54 GMT
Subject: what is a closure?
References: <3a39033b@news.xtml.co.uk>
Message-ID: <S18_5.3929$Kd1.455131@newsb.telia.net>

Nick Trout wrote:
> Sorry, this is probably not the place to ask this but you lot are pretty
> friendly :-)

So you don't want me to flame you for not knowing about
the free on-line dictionary of computing (foldoc)? ;-)

    http://www.foldoc.org/

> I have seen the term "closure" used... what is it exactly? Is it just a
> class member function?

    http://foldoc.doc.ic.ac.uk/foldoc/foldoc.cgi?query=closure

    a closure is a data structure that holds an expression
    and an environment of variable bindings in which that
    expression is to be evaluated. The variables may be
    local or global

In Python, you have to use explicity syntax to move variables
from an outer scope (environment) to the inner scope.

In the following example, the "a=a" default argument binding
binds add's local name "a" to the value of the "a" argument
from make_adder.

    def make_adder(a):
        def add(b, a=a):
            return a + b
        return add

    >>> m = make_adder(10)
    >>> m(20)
    30

In a language with full lexical scoping, you could leave it to the
compiler to figure this out.

But in today's Python, it's probably better to use a class instead.

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From syver at NOSPAMcyberwatcher.com  Fri Dec 15 21:29:13 2000
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Sat, 16 Dec 2000 03:29:13 +0100
Subject: For Python ASP developers
Message-ID: <91ek6l$vvk$1@troll.powertech.no>

I have been posting to this group a couple of times asking how to call
functions written in other ASP languages from Python.

I seem to have got it down now, the way to do it are as follows:

All functions written in other languages must be prefixed by
ScriptingNamespace.
    ex: ScriptingNamespace.VbScriptFunc() # calls a vb script function
defined in a runat=server script block

I didn't succeed at calling JscriptFunctions however until I found out that
you have to pass a dummy None parameter to
javascript functions without parameters. Like this:

<script language=Jscript runat=server>
function JscriptFunc()
{
    Response.Write("Jscript func called<br/>");
}
</script>

<%
ScriptingNamespace.JscriptFunc() # won't work
ScriptingNamespace.JscriptFunc(None) # will work
%>

The way I found out was the faq on perlscript (com enabled perl) on asp,
as they seem to run into the same issues as with Python.
http://faqs.net/perlscript/




From ronjeffries at acm.org  Sun Dec 17 11:02:24 2000
From: ronjeffries at acm.org (ron jeffries)
Date: Sun, 17 Dec 2000 11:02:24 -0500
Subject: writing a file on my web site
References: <1D793E572F1CCC75.4177A7CD821DE2C4.6A38056B7E83389E@lp.airnews.net> <91ie42$shk$1@bob.news.rcn.net>
Message-ID: <6BBCEDE25C4DD0D8.067F4BB38B52531A.4846B6E09FEA5EB2@lp.airnews.net>

On Sun, 17 Dec 2000 08:12:58 -0500, "Satheesh Babu" <vsbabu at erols.com>
wrote:

>With FTP tools, you probably have write permissions. But
>others might not have write permissions to that file. You'll
>need that so that web server process has permissions to
>write to it.
>
>chmod 666 should do it. Security alert - change the mode
>only to your roncount.txt

Thanks, I'll ask my ISP to change it. I don't seem to be able to do it
from here ...


From MarkH at ActiveState.com  Mon Dec  4 23:57:57 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Tue, 05 Dec 2000 04:57:57 GMT
Subject: PyThreadState_Get error again :(
References: <42+P6LAmqiK6Ewl1@jessikat.fsnet.co.uk>
	 <3A2AD10F.8040803@ActiveState.com> <uTVaDtAJ13K6EwUw@jessikat.fsnet.co.uk>
Message-ID: <3A2C7488.4040005@ActiveState.com>

Robin Becker wrote:

> Is there any way to get 135 without reinstalling 2.0 or perhaps there
> has been a fix in that? Is it worth pursuing further?

135 is available stand alone at 
http://www.activestate.com/Products/ActivePython/win32all.html

I would love to know if it is a problem there, as I can't repro on 98. 
Obviously, this is not to suggest that 95 has a problem while 98 doesnt. 
  I no longer have 95 installed here.

Mark.





From dsh8290 at rit.edu  Tue Dec 12 19:24:42 2000
From: dsh8290 at rit.edu (D-Man)
Date: Tue, 12 Dec 2000 19:24:42 -0500
Subject: [Q] Are Exceptions used that much in practice?
In-Reply-To: <915p3q$p1i$1@nnrp1.deja.com>; from ullrich@math.okstate.edu on Tue, Dec 12, 2000 at 13:00:32 -0500
References: <3A35D546.A4731B5B@enteract.com> <914mfd02ibt@news2.newsguy.com> <915p3q$p1i$1@nnrp1.deja.com>
Message-ID: <20001212192442.I16853@westsidecnc.rh.rit.edu>

Could you show what those reasons are?  I think that try-except-finally would be
really useful, but maybe there's some issues I haven't thought of.

-D

On Tue, 12 Dec 2000 13:00:32 David C.Ullrich wrote:
<snip>
 | I'm not going
 | to try to convince you he was right, but at least in Delphi
 | it wasn't just an oversight, they considered it and finally
 | decided against it.
 | 
 | --




From bryan at eevolved.com  Wed Dec 13 02:11:40 2000
From: bryan at eevolved.com (Bryan Mongeau)
Date: Wed, 13 Dec 2000 07:11:40 GMT
Subject: recommend DB modules?
Message-ID: <MiFZ5.2487$0U2.146712@news20.bellglobal.com>

One last thing:

Can anyone recommend a good DB module that must be completely platform 
independent?

-- 
<==================================>
Bryan Mongeau
Lead Developer, Director
eEvolved Real-Time Technologies Inc.
www.eevolved.com
<==================================>

"... the true utility function of life, that which is being maximized in 
the natural world, is DNA survival. But DNA is not floating free; it is 
locked up in living bodies and it has make the most of the levers of power 
at its disposal." -- Richard Dawkins



From olafb+usenet at pvv.org  Thu Dec 14 06:18:30 2000
From: olafb+usenet at pvv.org (Olaf Trygve Berglihn)
Date: 14 Dec 2000 12:18:30 +0100
Subject: rpmmodule.so and python 2.0
References: <dn1hf47mgu2.fsf@verden.pvv.ntnu.no>
Message-ID: <dn1d7evmet5.fsf@verden.pvv.ntnu.no>

Olaf Trygve Berglihn <olafb+usenet at pvv.org> writes:

> I've compiled rpmmodule against the new python api. But when I try to
> user the interface to the rpm database, I get the following error:
> 
> Python 2.0 (#1, Dec 13 2000, 12:05:47)
> [GCC 2.95.2 19991024 (release)] on linux2
> Type "copyright", "credits" or "license" for more information.
> >>> import rpm
> >>> db = rpm.opendb( 0, '/' )
> cannot open file //var/lib/rpm/nameindex.rpm: Invalid argument
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> rpm.error
> >>>
<cut>

Hmmm. It seems that the rpmmodule and the rpm binary uses different
versions of the Berkeley DB:

Before rebuilding with rpm.rebuilddb( '/' ):

$ file /var/lib/rpm/nameindex.rpm
/var/lib/rpm/nameindex.rpm: Berkeley DB 1.85 (Hash, version 2, native byte-order)

After rebuilding with rpm.rebuilddb.( '/' ):

$ file /var/lib/rpm/nameindex.rpm
/var/lib/rpm/nameindex.rpm: Berkeley DB (Hash, version 5, native byte-order)


Olaf
-- 
Olaf Trygve Berglihn <olafb+usenet at pvv.org>


From arturo_perez at wanadoo.es  Thu Dec  7 09:49:32 2000
From: arturo_perez at wanadoo.es (=?iso-8859-1?Q?Arturo_P=E9rez_Mulas?=)
Date: Thu, 07 Dec 2000 14:49:32 GMT
Subject: [Pythonwin] Pywin encoding (again)
Message-ID: <0sNX5.168$M47.3490@m2newsread.uni2.es>

Hello.

Sorry if I seem not to be able to understand this. I think I need some more help here. Let me first express my question, and then
some hints I have gotten from the group, to make sure I understand anything at all:

Question:

    - Is it possible to write Python scripts under Pythonwin using non-ascii characters (latin-1 or cp1252 in my case) in comments, docstrings, etc?

What I know/have read/guess, I would appreciate comments on these particular issues also:

- It is not the same to have non-ascii characters in the comments, etc (this depends on the file encoding) than output a certain encoding (which I can get using the appropriate encode).
    
- My WinNT box has a default encoding cp1252, and files edited with, for instance, "Notepad" are thus encoded.
   
- Pythonwin edits and reads files in "utf-8" encoding, so if I open in pythonwin a file edited in "notepad" I will get the file messed up, and the other way round.
    
- So, I guess, what I am trying to find is a way to have pythonwin reading/writing files using my choice of encoding

  To do this I browsed this group, and found a few threads on this, but I only got this all messed up; this is what I could make out:

            - I can change python's default encoding (which is ascii) in "site.py", but I really do not know what is the effect of this

            - Regarding pythonwin, Mark suggested installing the patch that can be downloaded for his site, and then a further patch, but:

                    1. The downloaded patch does not work for me. I get a "cannot import pywin.is_platform_unicode"
                    2. I followed another hint indicating that I should change "pywin.is_platform_unicode" to "pywin.scintilla.view.is_platform_unicode", but I get the same result
                    3. I got pythonwin to work using "pywin.scintilla.is_platform_unicode", but this did not solve my problem.

            - The two lines patch in "interact.py" did not work either.


Thank you all. I promise I will get all the answers on this topic sorted and posted in a website for future reference to everyone!

Arturo.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20001207/99d0366e/attachment.html>

From donn at u.washington.edu  Wed Dec 13 12:22:01 2000
From: donn at u.washington.edu (Donn Cave)
Date: 13 Dec 2000 17:22:01 GMT
Subject: running a pipe of commands from python
References: <yv24s08pz0c.fsf@lionsp093.lion-ag.de> <8ZFZ5.3701$Kd1.429028@newsb.telia.net> <yv2puiwoh6v.fsf@lionsp093.lion-ag.de>
Message-ID: <918b7p$3idm$1@nntp6.u.washington.edu>

Quoth Harald Kirsch <kirschh at lionbioscience.com>:
[ ... re module "pipes" ...]

| This also uses /bin/sh. And all this although python has all the
| necessary ingredients like fork/exec/wait/signal to do the plumbing
| itself. Strange.
|
| I think I'll stick to my own hack, the doc of which currently reads
| like this:
|
| def execpipe(cmds):
|   """
|   runs the commands given in the list cmds in a pipe with their stdout
|   and stdin plumbed together. Every element of cmds must itself be a
|   list giving the name of the command as the first element and the
|   rest of the list being its arguments.
|
|   If any of the elements of cmds is a pair, then the first must be the
|   command to run and the 2nd must be a file descriptor for output
|   which is then connected to the commands stderr. By default, the
|   stderr is the same as that of the calling script.
|   """
|
| The implementation seems to work for most cases. Not all signals which
| cause the parent to die are distributed properly to the childs.

You might in some cases want to dup unit 2 ("stderr") from the
output pipe on unit 1.  For example, if an upstream process in
the pipeline is something like "make" that generates a session
output somewhat indiscriminately on units 1 and 2.  That would
appear kind of hard to specify by supplying the descriptor in
advance.

Python processes in the pipeline would be feasible and maybe useful.
I.e., fork but no exec.

Occasionally you may want more control over argv[0] vs. the file
name to be exec'd.  Since execve() provides this control, it's
kind of a shame to design it out.  Environment variables likewise.

Remember to close pipe file descriptors accidentally inherited from
parent processes, lest they hold the pipe open when its intended
writer has exited.  Remember to trap all exceptions inside each
child fork.  Think about where to go from "can't exec".

It may be useful to allow pipelines to be components of a pipeline.

Generally speaking, the possibilities are numerous enough that
even UNIX shells, built to do this for a living, support only
the most popular options, so don't feel obligated to implement
anything unless you have a use for it.  Your objection to the
shell based functions like system() is legitimate, they can
be a hazard if the command data isn't exactly known in advance.
(I wouldn't be very concerned about the overhead, though.)

A process class would obviously make the system more flexible.

Signals are normally delivered to a process, but the tty
driver delivers them to the foreground process group, so
one might not be too surprised to see different results
from a kill vs. ctrl-C.  There might be situations where
this would be a problem, but I can't think of any right
off hand.  All the processes in your pipeline will be in
the same process group as the parent.

	Donn Cave, donn at u.washington.edu


From tbryan at python.net  Tue Dec 26 01:31:47 2000
From: tbryan at python.net (Thomas A. Bryan)
Date: Tue, 26 Dec 2000 01:31:47 -0500
Subject: Python Database Interfaces - Any Standard?
References: <RUo16.204045$RP.37382354@news1.rdc1.md.home.com> <kuk88pu0cd.fsf@lasipalatsi.fi> <Gwu16.205577$RP.37521696@news1.rdc1.md.home.com>
Message-ID: <3A483B53.ACB0C04D@python.net>

Brian Knox wrote:
> 
> The Perl DBI consists of two parts: the non database specific DBI module
> (DBI.pm). Then there is a DBD (DataBase Driver) module for each database.
> So, if I wanted to say work with Oracle, I'd use DBI.pm and the Oracle DBD.
> 
> use DBI;
> $dbh = DBI->connect( $dbi:Oracle:archaeo", "username", "password" );
> 
> To switch to say PostgreSQL, just change "Oracle" to "Pg" and make sure you
> have the DBD.

I haven't done much with Python and databases (yet), but I think that the 
idea is to be able to write 

import DCOracle
dbh = DCOracle.Connect( connect_string )
query = dbh.cursor()
query.execute( query_string )


To change to PostgreSQL, you'd have to change the import and the connection.

import pg
dbh = pg.connect( connect_string )
query = dbh.cursor()
query.execute( query_string )

Of course, this sort of change only works to the extent that you avoid 
any database-specific features of the database and the module that you're 
using.Some people achieve that goal, but it can be difficult in some 
applications to squeeze all of the performance you need without using some 
special feature that isn't portable.

---Tom


From andrew_dot_henshaw_at_earthling_dot_net  Sat Dec 30 22:21:12 2000
From: andrew_dot_henshaw_at_earthling_dot_net (Andrew Henshaw)
Date: Sat, 30 Dec 2000 22:21:12 -0500
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com>
Message-ID: <t4t9qv73vvvfff@corp.supernews.com>

<henry_crun3583 at my-deja.com> wrote in message
news:92jj3j$ff7$1 at nnrp1.deja.com...
> Can someone point me at an explanation of the logic for not having
> do:
> until
>
> or explain the elegant way to achieve this in Python.
>
> eg
> do :
>     R=Themistor.ReadResistance()
>     T=R_to_Temperature(R)
>   until (T >=25)
>

No one else seems to have brought up this point, so I'll attempt it ...
1.  How do you suggest implementing the indentation rules?  You seem to have
left your indentation ambiguous.
a)
do:
    R=Themistor.ReadResistance()
    T=R_to_Temperature(R)
until (T>=25)

or
b)
do:
    R=Themistor.ReadResistance()
    T=R_to_Temperature(R)
    until (T>=25)

If a), then this looks very unPythonic, because the indentation blocking is
backwards.  In other words "until (T>=25)" is a control statement for an
indented block that is above it (and it doesn't have a ":")
.
I understand that Python has statements like
if blah:
    do_something()
else:
    do_something_else()

but the else: doesn't control the statement above it.

if you suggest syntax b), then I wouldn't be greatly opposed, I suppose.
But then "do:" would simply be a synonym for "while 1:" and "until" would be
a synonym for "if : break".







From tmealy at my-deja.com  Fri Dec 15 10:02:48 2000
From: tmealy at my-deja.com (tmealy at my-deja.com)
Date: Fri, 15 Dec 2000 15:02:48 GMT
Subject: MS + Python = help???
References: <91an5e$qfm$1@nnrp1.deja.com>
Message-ID: <91dbqk$v02$1@nnrp1.deja.com>

In article <91an5e$qfm$1 at nnrp1.deja.com>,
  Arint? <jamarijr at my-deja.com> wrote:
> What is the address of the webpage that had the
python help in the form
> of one of the new micrsoft html help format
things.
>
> tia
>
> Sent via Deja.com
> http://www.deja.com/
>

www.activestate.com also has html-help for their
distribution of python.



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


From aahz at panix.com  Thu Dec  7 11:16:26 2000
From: aahz at panix.com (Aahz Maruch)
Date: 7 Dec 2000 08:16:26 -0800
Subject: OT: _Fifth Business_ (was Re: Python Origins)
References: <90k3pq$50p$1@nnrp1.deja.com> <5hGX5.68$E3.1197@uchinews> <90ncs0$mt5$1@panix2.panix.com> <m3lmts4k5x.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <90od4q$ont$1@panix2.panix.com>

In article <m3lmts4k5x.fsf at atrus.jesus.cam.ac.uk>,
Michael Hudson  <mwh21 at cam.ac.uk> wrote:
>aahz at panix.com (Aahz Maruch) writes:
>> In article <5hGX5.68$E3.1197 at uchinews>,
>> Ben Wolfson <wolfson at midway.uchicago.edu> wrote:
>>>
>>>_Fifth Business_ is the first book in the Deptford trilogy by Robertson
>>>Davies, a Canadian and a man with a very impressive beard (the other
>>>books being _Manticore_ and _World of Wonders_).
>> 
>> <confused look>  But what does that have to do with Andrew?
>
>http://www.amk.ca/davies/

Color me duh.  Should have checked there.  In my defense, Andrew didn't
have his URL in that post.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

What if there were no rhetorical questions?  --Aahz


From dale at out-think.NOSPAMco.uk  Sat Dec  9 11:12:28 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Sat, 09 Dec 2000 16:12:28 +0000
Subject: reloading modules question
References: <90scdf$hpe$1@troll.powertech.no>
Message-ID: <ouk43t494isd9m8fk92udjjbg2pb6crq7q@4ax.com>

"Syver Enstad" <syver at NOSPAMcyberwatcher.com> wrote:

>Is it safe to loop through sys.modules.items()
>reloading each module, as long as I catch the ImportError?
>
>Am I guaranteed to have all python modules reloaded if I do the loop twice?
>
>import sys
>for key, value in sys.modules.items():
>    if value:
>    try:
>         reload(value)
>    except ImportError, e:
>         print '%s\n' % e
>
>The reason I am asking is that IIS webserver keeps Python in memory so that
>it won't use the new versions of python modules I deploy if don't stop and
>start the webserver. I can't afford to stop the webserver all the time, so I
>am considering ways to reload the python modules while it's running.
>

I would suggest you are better off just stopping and starting the affected web sites. That's what we
do and it only takes a second.

You can automate the procedure using WMI or ADSI. I'm sure I used to have an example script but I'm
buggered if I can find it right now.

Try asking over in microsoft.public.scripting.wsh.


--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants


From joseamartin at venmex.com  Thu Dec  7 09:47:55 2000
From: joseamartin at venmex.com (Jose Martin)
Date: Thu, 7 Dec 2000 10:47:55 -0400
Subject: DCOracle Memory Leak?
References: <m3g0k1qih5.fsf@h0050047c4794.ne.mediaone.net>
Message-ID: <t2v9fq7v9ns565@corp.supernews.com>

Hi, sorry i cant help you, but , can you help me ?
I am trying to use DCOracle on Python 2.0 under Windows or Linux
but when i try to make my scritp :

The first line is
from DCOracle import *

when i try to run:
Python say me:

File "d:\python20\DCOracle\__init__.py", line 52, in ?
    import sys, Buffer
ImportError: No module named Buffer

I was searched my python path for a module called Buffer, but i dont have it
Whats happennig ?

Thanks...


Brad Howes <bradh at mediaone.net> escribi? en el mensaje de noticias
m3g0k1qih5.fsf at h0050047c4794.ne.mediaone.net...
> Has anyone seen a Python process using DCOracle grow and grow with each
fetch
> from an Oracle database? I can write a really simple loop that does a
fetch but
> does not save the value, and the Python process keeps growing. Are there
known
> bugs with Oracle 8i client libraries for Linux?
>
> --
>     "Were people this stupid before TV?" -- _White Noise_ by Don DeLillo




From piet at cs.uu.nl  Wed Dec  6 09:59:43 2000
From: piet at cs.uu.nl (piet at cs.uu.nl)
Date: 06 Dec 2000 15:59:43 +0100
Subject: CGI File Upload
References: <p8ce0tcgtt2bksb3adrf9cfr2i8egr4ff5@4ax.com> <8u7fn7$jpf$1@la-mail4.digilink.net> <90j10o$5p5$1@nnrp1.deja.com>
Message-ID: <ulmttsihs.fsf@cs.uu.nl>

>>>>> Bill Seitz <fluxent at yahoo.com> (BS) writes:

BS> Hmm, this works for me with text data (though the file length is
BS> slightly different: maybe line-end changes from Win to *nix?). With
BS> binary it's totally off. Uploading a JPEG I consistently only get 109
BS> bytes written to the file system. With GIFs I consistently get ~500
BS> bytes (but not always exactly the same number). With an MsWord file I
BS> get a bigger file written, but still not the whole thing (60kb out of
BS> 75kb).

BS> What do I do next?

Is your server on Windows? Then it probably is a matter of the standard
input being in text mode rather than binary mode.
So you have to set it to binary mode before anything is read from it.

See msvcrt.setmode()
-- 
Piet van Oostrum <piet at cs.uu.nl>
URL: http://www.cs.uu.nl/~piet [PGP]
Private email: P.van.Oostrum at hccnet.nl


From aleaxit at yahoo.com  Mon Dec 25 19:24:21 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 26 Dec 2000 01:24:21 +0100
Subject: Season's greetings (was Re: Some Python 2.1 ideas)
References: <mailman.977765348.13376.python-list@python.org>
Message-ID: <928otj02l39@news1.newsguy.com>

"Tim Peters" <tim.one at home.com> wrote in message
news:mailman.977765348.13376.python-list at python.org...
> [Bob Alexander]
> > P.S. Merry Christmas!
>
> [Aahz Maruch]
> > Bah, humbug.
>
> Sounds like you need something to cheer you up, Santa Aahz!  Elf Timmy has

Me, I had the best Christmas ever since I got my first electric
train... got to show Python to my son (Pascal and Assembler
his only previous programming experiences), and again and
again he got the jaw-drops-"but-it-CAN'T-be-THAT-powerful!"
effect which I well recall from my early Python days.  (And as
a cherry on top he kibitzed me for an OKbridge session where
I got close to 70%, but the Python part alone would have, by
itself, been quite enough...!-).

One thing I wasn't able to easily answer -- what's the
*easiest* way for a Python beginner to do function-plots,
with nicely labeled graphs, minor/major ticks, color,
auto-scaled, &c &c?  On-screen display the major need,
though nice printouts wouldn't hurt either.  Speed no
issue, as his new PC is twice as fast as mine!-)

I tried downloading DISLIN but it appears to embed its
own Python, need its own setting of PYTHONPATH; and
(from a superficial look at it) it doesn't seem to do auto
scaling.  Any suggestions?  I'd like to keep amazing him
with Python power and ease, as what dad wouldn't like
for his late-teenage son...:-).


Alex





From anton at vredegoor.doge.nl  Tue Dec 19 11:20:09 2000
From: anton at vredegoor.doge.nl (Anton Vredegoor)
Date: Tue, 19 Dec 2000 17:20:09 +0100
Subject: Changing IDLE's stdout color at runtime
References: <91isi7$hup$1@news.hccnet.nl> <u8zpdzusd.fsf@roundpoint.com>
Message-ID: <91o25g$piv$1@news.hccnet.nl>

On 18 Dec 2000 14:08:34 +0000, Ben Hutchings
<ben.hutchings at roundpoint.com> wrote:

>You'd need to work out whether standard output is connected to a
>terminal or terminal emulation; if so, what kind of terminal that is;
>whether that terminal supports colour; and if so, what the appropriate
>control codes are for changing the colour.  You *could* assume that
>any terminal supports the ANSI escape codes for colour (ESC [30n, ESC
>[31n, etc, if I remember correctly), but that would really annoy
>anyone using a terminal that doesn't.

Thanks for pointing this out. After inspecting the sourcefiles I have
the following interpretation: Idle is not a terminal but a Tkinter
window redirecting standard output. The text is colorized later
according to a syntax scheme. There seems to be no obvious way of
having two lines of "stdout" with different colors in the same window.

Below is a script to change Idle's stdout color. This only works for
the standard Idle thats delivered with the python20 distribution. It
doesn't work with the idlefork-0.7.1 thats on sourceforge. By the way
this idlefork won't run with python20 unless one gets the new
"protocol.py" or manually edits line 331 and 364 to :

331: s.bind(('', self.port))  
364: s.connect((ip,port or Server.default_port))

#stdoutcolor.py

"""
 Example program showing how to change
 the foregroundcolor of stdout in Idle
"""

import sys
import time

OldStdoutForeground = sys.stdin.color.tagdefs['stdout']['foreground']
sys.stdin.color.tagdefs['stdout']['foreground'] = 'Orange'
sys.stdin.color.config_colors()
print sys.stdin.color.tagdefs
time.sleep(3)
sys.stdin.color.tagdefs['stdout']['foreground'] = OldStdoutForeground
sys.stdin.color.config_colors()






From ullrich at math.okstate.edu  Tue Dec 12 13:00:32 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Tue, 12 Dec 2000 18:00:32 GMT
Subject: [Q] Are Exceptions used that much in practice?
References: <3A35D546.A4731B5B@enteract.com> <914mfd02ibt@news2.newsguy.com>
Message-ID: <915p3q$p1i$1@nnrp1.deja.com>

In article <914mfd02ibt at news2.newsguy.com>,
  "Alex Martelli" <aleaxit at yahoo.com> wrote:
>[...] it's a pity a 'try' can't have one
> or more 'except' clauses _followed_ by a 'finally' one, etc, etc,

    Python and Delphi both have try-except and try-finally
but no try-except-finally. I once saw the person responsible
for this in Delphi explain why there was no try-except-finally,
in reply to a complaint about its non-existence. I'm not going
to try to convince you he was right, but at least in Delphi
it wasn't just an oversight, they considered it and finally
decided against it.

--
Oh, dejanews lets you add a sig - that's useful...


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


From raymond at ibercom.com  Wed Dec 20 15:42:32 2000
From: raymond at ibercom.com (Raymond Biurrun)
Date: Wed, 20 Dec 2000 21:42:32 +0100
Subject: Python interpreter in python?
References: <mailman.977335880.30954.python-list@python.org>
Message-ID: <91r547$emg$1@talia.mad.ttd.net>

Thank you so much for your reply.

Regards -- Raymond




From paul at vanerk.demon.nl  Wed Dec  6 05:16:18 2000
From: paul at vanerk.demon.nl (Paul van Erk)
Date: Wed, 6 Dec 2000 11:16:18 +0100
Subject: Changing frames with a python script
Message-ID: <976097884.20994.0.pluto.c30bfd5c@news.demon.nl>

Hi,

I am new tp Python, but eager to learn the language. I have a little
problem. My website is divided in two frames (navigation and main). In the
main frame I have a login page which is connected to a database. People can
fill in their name an passwd there. Then a python-script will be started. If
the login and the passwd is correct the navigation frame has to changed into
a frame with more possibilities. How can I solve this problem.
My first solution was that the script would make a new html document, which
was also divided in frames. But then I get a document with three frames
(logical), because it would publish the html document only in the main
frame.

Has anyone a solution for this??

Thanks in advance

Paul van Erk
paul at vanerk.demon.nl




From mliss at exmail.de  Tue Dec 12 07:29:27 2000
From: mliss at exmail.de (Mirko Liss)
Date: Tue, 12 Dec 2000 13:29:27 +0100
Subject: compiled or interpreted?
References: <w8jZ5.74702$IP1.2487256@news1.giganews.com>
Message-ID: <3A361A27.F462DA91@exmail.de>

Matt Frey wrote:
> 
> Is Python a compiled language that can be run on a platform like c++, or is
> it interpreted like java, or is it in a class of its own?

It's compiled to virtual machine bytecode just like Java.
Python code can be developed very fast, but your computer needs 
a lot of time to run it.
I use Python to create software prototypes.

You might want to take a look at the list of frequently asked questions
at www.python.org for further details.

Mirko

-- 
Mirko Li? <mirko_liss at web.de>


From tim.one at home.com  Thu Dec 28 01:19:23 2000
From: tim.one at home.com (Tim Peters)
Date: Thu, 28 Dec 2000 01:19:23 -0500
Subject: Good introduction to functional programming with Python?
In-Reply-To: <921nd205si@news2.newsguy.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEKPIFAA.tim.one@home.com>

[Alex Martelli]
> We agree on far too much, and dwelling on agreements ain't no
> fun, so, let's pick on a difference:

Well, if we're still addressing the original fellow's question about a good
intro to FP, I think "Haskell or Scheme, depending" is as good an answer as
he's ever going to get -- so there's no real *need* to disagree beyond that
<wink>.

[Tim]
>> ... then if FP is "about" anything, it's about the art of using
>> higher-order functions, and I believe mere mortals *need* a
>> strong static typing system to keep those straight.  Maps returning
>> maps returning maps ... can get hard to follow real fast.

[Alex]
> Erlang designers and practitioners seem to disagree: and Erlang
> is arguably the FP around which most $$$ turn in today's real-
> world, what with its use in Ericsson and Bluetail's acquisition by
> a Nortel subsidiary.  I have no real experience with it, and thus
> no solid basis on which to judge it, but, sordid materialist that
> I am, I _am_ impressed when hundreds of megabucks are in play:-).
> And Erlang relies on strong dynamic (aka latent)  typing, and an
> emphasis on concurrency in an FP setting, to (allegedly) reach good
> levels of reliability, performance, ease of learning, programmer
> productivity.
>
> Static typing has its place, but so, it appears, does latent
> typing, even in FP.

I haven't used Erlang at all, so that makes it an excellent battlefield:  no
unpleasant haggling over facts <wink>.  I'll hazard a guess based on the
Erlang FAQ (http://www.erlang.org/faq/t1.html).  According to 10.9.1,

    10.9.1. Erlang should have a static type checker

    Erlang has a type checker.  Several man years of work have been
    put into the type checker by people who "should know" (Joe
    Armstrong, Phil Wadler, Simon Marlow, Thomas Arts) but the result
    is still not a "must use" part of the language.  The type checker
    is not widely used.

So they have a capable static type inference system but nobody cares,
neither the implementers nor the users.  There are two reasons for that I'm
guessing:

1. Erlang appears to be a nice implementation of a generalization of CSP,
and of a microthread flavor.  They have a crushing speed advantage over most
OS-level threads in their niche ("Context switching between Erlang processes
is typically one or two orders of magnitude cheaper than switching between
threads in a C program"), so don't care whether anything else "runs really
fast":

    1.4. What sort of problems is Erlang not particularly
    suitable for?

    Anything where constant-factors are crucial for performance,
    such as image processing, signal processing, sorting large
    volumes of data, device drivers and low-level protocol termination.

    Most (all?) large systems developed using Erlang make heavy use
    of C for low-level code, leaving Erlang to manage the parts
    which tend to be complex in other languages, like controlling
    systems spread across several machines and implementing complex
    protocol logic.

2. While the "complex protocol logic" may be complex indeed, protocols
typically sling a small number of very simple data types (like "Data is
passed as a sequence of bytes between the Erlang processes and the external
UNIX processes") in very rigid ways.  In fact, I got so suspicions at the
lack of *any* higher-order functions in any of the examples I saw, I
downloaded the Erlang book and read this ("Concurrent programming in
Erlang", ISBN 0-13-508301-X):

    Currying, higher-order functions, lazy evaluation, ZF
    comprehension, logical variables, deep guards, etc., add to
    the expressive power of a declarative programming language, but
    their absence is not a sign cant detriment to the programming of
    typical industrial control applications.

In the absence of higher-order functions, I can believe programmers don't
have much trouble reasoning about types without declarations.  They don't
use the static type checker because it's more trouble than it's worth!

Fine by me.  I must say, though, that Erlang appears to occupy an extreme
corner of the FP universe.

    9.2 Why can't I change the value of a variable?

    ... [because you can't <0.5 wink>] ...

    (This behaviour is called "single assignment", and it's considered
    to be a feature, not a bug. Most functional languages, including ML
    and Haskell, also behave this way. Among other things, It's nice
    being able to rely on a certain variable always having the same
    value.)

While not the most compelling defense of FP I've ever seen <ahem>, they do
know their audience!  That's all to their credit.

I *like* the attitude of the Erlang developers.  Looks like a very useful
language for its target domain.  Even better, according to FAQ 5.7.11
there's a Python binding (if you really need some higher-order functions
after all <wink>).

if-we-don't-act-fast-i'm-afraid-python-may-lose-the-industrial-
    control-market-ly y'rs  - tim




From aleaxit at yahoo.com  Thu Dec 14 04:39:59 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 14 Dec 2000 10:39:59 +0100
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net>
Message-ID: <91a4ht02e59@news2.newsguy.com>

"Tim May" <tcmay at got.net> wrote in message
news:tcmay-19C054.14031113122000 at news.got.net...
    [snip]
> Large, complex apps in LISP and Smalltalk did well without using MI.
>
> I don't want to start a religious war over MI vs. SI vs. SI + mix-ins,
> but I'd say that MI fans have not shown a compelling reason why MI is
> necessary.

I'd go further, and say that implementation-inheritance fans
have never shown a compelling reason why any inheritance AT
ALL is 'necessary' -- and we never will (I'll class myself among
such fans:-), because it is *NOT* _necessary_.  Many complex
LISP applications were programmed without actually using CLOS;
C was used to program complex, successful applications, and has
no inheritance at all; Visual Basic, up to version 6, did not
allow implementation inheritance.  Whatever you can do by
inheriting implementation, you can also do by encapsulating
and delegating, and in some languages it's almost as easy
(in Python, for example, a suitable __getattr__ method may
let you 'transparently delegate' almost as conveniently as
inheritance does).

(This is all about inheriting _implementation_ -- C++, Eiffel,
and other languages yet, merge/confuse the concepts that Java
separates as '<class> extends <class>' [implementation
inheritance] and '<class> implements <interface>' [interface
compliance] -- the issues with the two mechanisms are really
pretty separate, and implementation inheritance is more often
what is meant when unqualified 'inheritance' is discussed).

If inheritance itself is not _necessary_ (just _convenient_),
then, a fortiori, so will multiple inheritance be -- surely
not _necessary_, possibly more _convenient_.

It is not *necessary* for a family to own a car; bicycles,
public transportation, rental, etc, may (perhaps at some
inconvenience) substitute.  A fortiori, ownership of more
than one car by a family is not _necessary_ either; it's
pretty intuitive, in fact, that the added value of each
further car you add to a family's pool of vehicles goes
down as the number of vehicles already in the pool goes
up (the first car adds very significant convenience to the
family's ability to go around; the second one may add a
little less; the third one is probably less important
yet; etc).  Similarly, the added value of MI vs SI (in
terms of _convenience_ -- no _necessity_ involved) is no
doubt lesser than that of SI versus no implementation
inheritance at all.


Say that 2/3 of your objects need/want to delegate some
functionality to one other object; 1/3 of those want to
delegate to two other objects; 1/6 of those want to
delegate to three; and so on.  Say this covers only
those cases of delegation that are well covered by
implementation inheritance (delegating to a unique
and permanent 'base object').

With no inheritance at all, you have to take the (modest)
hit of contain-and-delegate for 66 objects out of 100.

With single inheritance, this goes down to 22 -- a net
gain in 44 cases; with multiple inheritance, it goes
down to 0 -- but the further net gain is only 22, half
of the convenience gain that SI brought over nothing.

(A hypothetical 'dual inheritance' system, where you
can inherit from 0, 1, or 2 bases, but no more than
that, would already cover 18 of those 22 cases...).


That the added-convenience is on a diminishing-returns
curve doesn't imply the ideal stopping point is at
allowing 1 inheritance-of-implementation, of course --
you might as well argue for 0, or, 2.  Or maybe, just
like it happens for family and cars, it would be best
to 'disperse' the design-decision -- let each family
determine how many cars best serve its needs, be it
0, 1, 2, whatever.  Similarly, the designer of each
class is perhaps best-placed to determine how many
other classes it is most convenient for it to inherit
from, rather than having the language designer fix
it in advance for all time...?


Alex





From arturo_perez at wanadoo.es  Sun Dec 10 05:34:11 2000
From: arturo_perez at wanadoo.es (=?iso-8859-1?Q?Arturo_P=E9rez_Mulas?=)
Date: Sun, 10 Dec 2000 10:34:11 GMT
Subject: Fault trees structure
Message-ID: <D_IY5.49$17.2785@m2newsread.uni2.es>

Hi,

I am creating a fault-trees (kind of binary trees, but with unlimited inputs in every gates and several different types of gates/inputs) management package; thouhg I have created a basic structure for the class, I am not so happy with it and I am certain that it can be improved greatly.

Any one knows of an Object Oriented description of a similar structure, or is interested in this so we can discuss further?

Regards,

Arturo.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20001210/8c965233/attachment.html>

From glenfant at equod.com.nospam  Mon Dec 18 10:07:40 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Mon, 18 Dec 2000 16:07:40 +0100
Subject: Stored Procedures (ODBC) from Python?
References: <91l6mq$lge$1@nnrp1.deja.com>
Message-ID: <91l8va$qa$1@reader1.imaginet.fr>

Depends on your DB engine.

With mxOdbc and an Informix server, it works for me.

import mxOdbc
mydatasource = 'DSN=adatasource;URS=myself;PWD=secret'
myconnection = mxOcbc.Windows.DriverConnect(mydatasource)
curs = myconnection.cursor()
curs.execute('Any SQL statement including stored procedure call or informix
specific SQL operation')

That's all

<steinarrune at my-deja.com> a ?crit dans le message news:
91l6mq$lge$1 at nnrp1.deja.com...
> I am developing in Pythonwin, and have been using
> the odbc module successfully for a while. Still I
> have not yet called Stored procedures over this
> interface.
>
> Is it possible ? Any examples out there ?
>
> All the best
>
> Steinar Eriksen
>
>
> Sent via Deja.com
> http://www.deja.com/



From cfelling at iae.nl  Mon Dec 11 07:33:47 2000
From: cfelling at iae.nl (Carel Fellinger)
Date: 11 Dec 2000 13:33:47 +0100
Subject: type vs. class
References: <3A34244A.415F4DFB@t-online.de>
Message-ID: <912hjb$2gi$1@animus.fel.iae.nl>

Thomas Thiele <tnt-audio at t-online.de> wrote:
> Hallo!

> simple typetest

>>>> type(3)
> <type 'int'>

>>>> class X:
> 	pass
> 	
>>>> x = X()
>>>> type(x)
> <type 'instance'>

> How can I realize that I will not get type 'instance' but type 'X'? I

have a look at the build-in functions isinstance() and issubclass()

> have different classes, more than one, and I want to check out which
> class it is.  Are all classes type instance?

No, classes are of type Class:)  You see, python doesn't do types like
say Pascal. It has two levels of type. The first level deals with integers,
floats, lists, functions, classes, etc. The second level gives a finer
grain of type distinction, but only for classes.  The type function only
deals with the first level of types.  The above mentioned isinstance and
issubclass functions deal with the second, class oriented level of types.

The language itself only uses the first level of types and ignores the
other.  The nice side of this dichotomy is that programs tend to focus
on interface instead of class type.  So as long as an instance provides
all the methods used al is swell.
-- 
groetjes, carel


From dougfort at downright.com  Sat Dec 30 07:39:47 2000
From: dougfort at downright.com (Doug Fort)
Date: Sat, 30 Dec 2000 07:39:47 -0500
Subject: httplib and '100 Continue' after POST
References: <UIi36.7084$Sl.370436@iad-read.news.verio.net>
Message-ID: <oIk36.7086$Sl.371260@iad-read.news.verio.net>

I believe there is a bug in httplib (I'll send a report to ActiveState,
where I got my distro).

IIS 4 and 5 are subject to send an unsolicited result code of '100 Continue'
with a couple of headers and a blank line before sending '302 Object Moved'.
The 100 response is totally worthless and should be ignored.  Unfortunately,
httplib.HTTPConnection is unwilling to go back and read more headers when it
already has a response object.

I was able to get past this with the following kludge:

                while 1:
                    response = self._client.getresponse()
                    if response.status != 100:
                        break
                    # 2000-12-30 djf -- drop bogus 100 response
                    # by kludging httplib
                    self._client._HTTPConnection__state =
httplib._CS_REQ_SENT
                    self._client._HTTPConnection__response = None

"Doug Fort" <dougfort at downright.com> wrote in message
news:UIi36.7084$Sl.370436 at iad-read.news.verio.net...
> I can't figure out how to persuade httplib to handle a response status of
> '100 Continue'.  Has anyone accomplished this?
>
> --
> Doug Fort
> Senior Meat Manager
> Downright Software
> http://www.dougfort.com/~dougfort
>
>
>




From sandj.williams at gte.net  Wed Dec 13 10:54:06 2000
From: sandj.williams at gte.net (Steve Williams)
Date: Wed, 13 Dec 2000 15:54:06 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com> <r0pwvd51gga.fsf@black.cs.washington.edu> <9165he$lon$1@panix3.panix.com> <917pvv02clr@news2.newsguy.com>
Message-ID: <3A379CB4.405992F8@gte.net>

Alex Martelli wrote:

> [snip]

> Maybe they took solidity _for granted_, because, in their
> (Renaissance) times and in their (Architecture) calling,
> compromises regarding solidity were simply unthinkable.
> Well, we're not so lucky, in the software field, today; the
> Firmitas of *by far* most software around is imperfect.

[snip]

I understand the Romans required an architect to stand under his arch as
the falsework was being removed.





From urner at alumni.princeton.edu  Wed Dec 20 21:29:42 2000
From: urner at alumni.princeton.edu (Kirby Urner)
Date: Wed, 20 Dec 2000 18:29:42 -0800
Subject: Python in "math through programming" curriculum
References: <2g2t3t06r2agrrjmfnr41eu4lc2u39a8qi@4ax.com> <3A3EFF47.67DE69CF@gte.net> <91n9i30pr7@news2.newsguy.com> <slrn9402bt.fa6.neelk@alum.mit.edu> <91qa3r01m78@news2.newsguy.com> <slrn942nln.mjs.neelk@alum.mit.edu>
Message-ID: <p4q24tsdp1387496c63reu0ehoo7pnkged@4ax.com>

neelk at alum.mit.edu (Neelakantan Krishnaswami) wrote:

>Yes -- the reference to Knuth's _Concrete Mathematics_ made me think
>that an introduction to CS was envisioned. For things like graphing
>and statistics, there's no reason not to use Python, since it has good
>graphing and statistical libraries. Static typechecking doesn't help
>here, I don't think, and in Haskell's case it becomes a positive
>nuisance (because of all the IO).

Certainly don't want to close the door to CS at all, no way.
Just trying to get computers used more in math class, instead
of just calculators all the time.  'Concrete Mathematics' is
mathematics too, not just CS (same with Knuth's 'Art of...'
-- Miller-Rabin derives from number theory).

Of course more computers in math class'd open more doors to 
CS.  But I'm fighting the prejudice that you don't get to 
use the computer for serious tasks until you do some CS 
elective -- as if math were some whole other domain that 
shouldn't be sullied with computers (but calculators OK 
for some reason).

One thing computers have over calculators is better spatial
geometry displays.  I graft Python to Povray, suggesting Povray
might be more fully exploited in its own right (independently
of Python) -- in art class for example.  In the meantime, I 
encourage Polyhedra as paradigm Objects in both a geometric
and in an OO sense.[1]

Your remarks re strongly typed FP languages are interesting 
and I still want to learn more in that domain -- glad it's 
being discussed.  

I think Python opens doors to languages like Haskell, vs 
closes them, because it has some of the FP syntax, even if 
it's not built around lazy evaluation, tail recursion etc.
etc.  The map, lambda, zip, filter and other syntax has an
FP flavor.  So I think if students pick up on Python first, 
they're not thereby ruined for Haskell or ML later (or Java
either, or any number of other worthy, useful languages).

Anyway, thanks for your intelligent contributions -- learned
some stuff.

Kirby

[1] http://www.inetarena.com/~pdx4d/ocn/trends2000.html




From barry at digicool.com  Fri Dec 22 11:55:13 2000
From: barry at digicool.com (Barry A. Warsaw)
Date: Fri, 22 Dec 2000 11:55:13 -0500
Subject: mail processing module
Message-ID: <web-395109@digicool.com>

I've been working on a new package for dealing with message
objects, including mime messages.  Mostly this is for use in
Mailman, but they will be generally useful.  I want to
eventually replace essentially what's in rfc822.py,
multifile.py, mimetools.py, and taking some of the best
ideas of stuff like mimecntl.py.

The basic idea is that I've separated the object model from
parsing and generating.  That way you can drop in different
parsers or generators, and you can manipulate the object
model independently.  Conceptually a message object has a
set of headers and a payload.  The payload can be just a
string or a message object, or a list of those.  Then you
can add/remove headers with semantics similar (but slightly
different) to rfc822, and you can add/remove payload
objects.

You pass a root message object to the generator, and it
spits out text suitable for delivery via smtplib.py.  It
knows that message objects w/ list payloads are multiparts.
The parser on the other hand, takes text and spits out a
message object tree.  parsing->object tree->generator is
idempotent.

For composition, there is a base class that knows about
content-type: headers, and subclasses for common subtypes
(image/*, text/*, etc.).  Cool things you can do with this
are: take a mime message, find the image subparts, rip them
out and post them to a webdav server, replacing the image/*
part with a text/* part that contains just a url to the
image; throw away all binary subparts, reordering the
containing multipart, etc. etc.

It's not complete right now and I won't be able to post
anything until I get back my internet connection, but stay
tuned.  It'll be distrib'd as a distutils package so it'l be
easy to install.  I'm just not sure what to call it (always
the most important thing to decide. :)  The two I've come up
with so far are "pantomime" and "mimeo".

Cheers,
-Barry



From greg at cosc.canterbury.ac.nz  Thu Dec 14 21:05:40 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Fri, 15 Dec 2000 15:05:40 +1300
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <915jcf0c6o@news2.newsguy.com> <9162p3$7t2$1@animus.fel.iae.nl> <MBWZ5.334$h67.37748@newsread1.prod.itd.earthlink.net>
Message-ID: <3A397C74.6C8906E1@cosc.canterbury.ac.nz>

Oldayz wrote:
> 
> I thought something like forever: would be even more lucid

If there is a break in the loop somewhere, it's
not really forever, though, is it?

So perhaps it should *really* be

   for a while:
      ...

:-)

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


From sbrunning at bigfoot.com  Mon Dec  4 12:23:05 2000
From: sbrunning at bigfoot.com (Simon B.)
Date: Mon, 04 Dec 2000 17:23:05 GMT
Subject: good newbie books?
References: <D_jU5.25243$xK.887693@news20.bellglobal.com> <3A227DB1.C3F90F8F@dmu.ac.uk>
Message-ID: <90gjtf$7s1$1@nnrp1.deja.com>

In article <3A227DB1.C3F90F8F at dmu.ac.uk>,
  danielneumann at gmx.net wrote:
>
> "The Quick Python Book"
> By Daryl Harms, Kenneth McDonald
> Manning Publications
> ISBN: 1-884777740
>

If know how to program, but you are new to Python, then I'd go along
with this entirly. I found this book very useful indeed.

If, however, you are new to programming altogether, then this book
isn't really for you IMHO.

--
Simon Brunning
Honest criticism is hard to take, especially from a relative, a friend,
an acquaintance, or a stranger. - Franklin P. Jones


Sent via Deja.com http://www.deja.com/
Before you buy.


From fra at bastardi.org  Thu Dec 14 06:51:24 2000
From: fra at bastardi.org (Fra)
Date: 14 Dec 2000 12:51:24 +0100
Subject: Close a tkinter window (not from win manager!)
Message-ID: <3A38B453.CA724F73@bastardi.org>

Hi all, i got a little problem.
I have to do some windows that explain to the user what a program is
doing, something like "copying file", "deleting file" and so on
I want that when the copy of the file (or whatever else) is done the
window automatically close itself, and that a second window appear.
(It's ok also if the text in the window change).
I couldn't find how to make that! Please help me.
Thanks a lot
		Francesca

-- 
Posted from mister.cdc.polimi.it [131.175.12.3] 
via Mailgate.ORG Server - http://www.Mailgate.ORG


From aahz at panix.com  Fri Dec 29 14:11:31 2000
From: aahz at panix.com (Aahz Maruch)
Date: 29 Dec 2000 11:11:31 -0800
Subject: Great New Book
References: <t4o1vr8hktk531@corp.supernews.com> <92ij47$n3p$1@bob.news.rcn.net> <slrn94pnrk.irm.grey@teleute.rpglink.com>
Message-ID: <92inl3$a5o$1@panix2.panix.com>

In article <slrn94pnrk.irm.grey at teleute.rpglink.com>,
Steve Lamb <morpheus at here.not.there> wrote:
>On 29 Dec 2000 17:54:15 GMT, Fred Yankowski <fcy at enteract.com> wrote:
>>
>>I actually prefer heavier paper since it allows me to wield my
>>highlighter with abandon without worrying about marring the backside
>>of the marked page.
>
>    Not to mention 2'5" isn't all that thick.  I have fictional paperbacks
>that are that thick.  :)

Really?  You have paperbacks more than two feet thick????  (No fair
counting all the Robert Jordans as a single book.)
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

Ninth Virtual Anniversary: 2 days and counting


From gervin at pd.jaring.my  Thu Dec 21 07:58:59 2000
From: gervin at pd.jaring.my (Kervin Goh)
Date: Thu, 21 Dec 2000 20:58:59 +0800
Subject: Editing files
References: <Bdd06.5591$K6.674078@juliett.dax.net>
Message-ID: <91svba$8rc$1@news4.jaring.my>

I am not sure if this is what you wanted but I had to write a text to html
program for my assignment and I used regex object to match the pattern and
change it. I haven't done Python in a while so I'm a bit rusty but try
looking at the regex object and read up on Python regular expressions. Hope
this helps.

dv <lars at stea.nu> wrote in message
news:Bdd06.5591$K6.674078 at juliett.dax.net...
> I'm trying to automate some configuration tasks with python. Is it
possible
> to do a "search and replace" or specify where (which line) in a file
python
> will write?
>
> Thanks in advance
>
> Lars
>
>




From aleaxit at yahoo.com  Tue Dec 19 09:00:57 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 19 Dec 2000 15:00:57 +0100
Subject: [Q] Are Exceptions used that much in practice?
References: <3A35D546.A4731B5B@enteract.com> <3A36C6E3.26181E3F@enteract.com> <FJBZ5.19773$x6.11272365@news2.rdc2.tx.home.com> <Pine.GSO.4.21.0012122229470.19682-100000@y.glue.umd.edu> <91m1as$e9j$1@panix2.panix.com>
Message-ID: <91npnh018bc@news2.newsguy.com>

"Aahz Maruch" <aahz at panix.com> wrote in message
news:91m1as$e9j$1 at panix2.panix.com...
> In article <Pine.GSO.4.21.0012122229470.19682-100000 at y.glue.umd.edu>,
> Roy Katz  <katz at Glue.umd.edu> wrote:
> >
> >If you enclose statements in a try/except clause, and assuming no
> >exception is thrown, what are the runtime penalties?
>
> Minimal, from what I've heard, but I don't feel like doing any
> benchmarking right now.

Depends on how much you're doing.  If the amount of
work is minimal, the overhead can easily get as big
as 10% or so...:

import time

def clean(n):
    tot = 0L
    for i in range(n):
        tot += i
    return tot

def safe(n):
    tot = 0L
    for i in range(n):
        try: tot += i
        except ValueError: pass
    return tot

start = time.clock()
tot = clean(100000)
stend = time.clock()
print 'clean:', stend-start

start = time.clock()
tot = safe(100000)
stend = time.clock()
print 'safe:', stend-start


D:\PySym>python ddd.py
clean: 0.650930034144
safe: 0.724842480024

D:\PySym>python ddd.py
clean: 0.645219254062
safe: 0.726402175024

D:\PySym>python ddd.py
clean: 0.651437081686
safe: 0.726812003533

D:\PySym>


So, putting a try/except around (say) each single
addition *does* have measurable impact.  A grain
a _tad_ less fine than this might be better, if
performance is an important issue.


Alex





From tim.one at home.com  Tue Dec 12 22:42:23 2000
From: tim.one at home.com (Tim Peters)
Date: Tue, 12 Dec 2000 22:42:23 -0500
Subject: [Q] Are Exceptions used that much in practice?
In-Reply-To: <Pine.GSO.4.21.0012122229470.19682-100000@y.glue.umd.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEJIIDAA.tim.one@home.com>

[posted and mailed]

[Roy Katz]
> ...
> If you enclose statements in a try/except clause, and assuming no
> exception is thrown, what are the runtime penalties?

This varies widely across languages and language implementations, but in
CPython the penalties are minimal, in large part because CPython performs no
optimizations that make unwinding exceptions difficult.




From KatMouse at gmx.de  Thu Dec 28 12:08:50 2000
From: KatMouse at gmx.de (Eduard Hiti)
Date: Thu, 28 Dec 2000 18:08:50 +0100
Subject: sniffing btw my web browser and a server
References: <92aush$rhc$1@news.nuri.net>
Message-ID: <92fs3l$5gg$01$1@news.t-online.com>

You can try 'Proxomitron' as a proxy. It's basically an ad filtering
(shonenware!?) software (but capable of much more), and has a very nice log
window, in which you can view all HTTP headers sent.

The URL: http://members.tripod.com/Proxomitron/


There's also a Python only solution: 'WebDebug'. It also acts as a proxy,
logging all transfers and generating a nice and comprehensive HTML report.

The URL: http://www.cyberclip.com/webdebug/download.html

Good luck,

    Eduard Hiti


"June Kim" <junaftnoon at nospamplzyahoo.com> wrote in message
news:92aush$rhc$1 at news.nuri.net...
| I'm doing some sort of web client programming, and have to observe the
| messages that my browser and web-servers transmit, being tired of guessing
| what the messages'd be and looking up the references and the
specifications.
|
| Anyone has done it in Python or knows how it could be done?
|
|
| Best regards,
|
| June
|
| ps. using TCP/IP packet sniffers made for mere hacking seems too primitive
| or low-level in this case.
|




From SBrunning at trisystems.co.uk  Wed Dec  6 11:29:21 2000
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Wed, 6 Dec 2000 16:29:21 -0000 
Subject: Python: what all the hype is about?
Message-ID: <31575A892FF6D1118F5800600846864D5B155A@intrepid>

> From:	aivkov at my-deja.com [SMTP:aivkov at my-deja.com]
> Excuse my ingnorance in the subject area. Can someone explaing me very
> briefly, why one should choose Python over Java or is it better to
> compare it with Javascript/ECMA?
 
See <http://www.python.org/doc/Comparisons.html>.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From pbjorn at uswest.net  Sun Dec 17 14:03:35 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Sun, 17 Dec 2000 12:03:35 -0700
Subject: Error evaluating numbers starting in zero
References: <3A3D07BD.85EFEB3D@wag.caltech.edu>
Message-ID: <3A3D0E07.EC5604C3@uswest.net>

How about:

   >>> time.strptime('01/08/99', '%m/%d/%y')
   (1999, 1, 8, 0, 0, 0, 0, 0, 0)

or

   >>> string.atoi('09')
   9

or

   >>> int('09')
   9

-- bjorn

"Richard P. Muller" wrote:

> I'm sure people already know about this, but I just got hit with a
> really screwy error.
>
> I have a python checkbook program that evaluates date strings, such as
> "01/31/99". Most dates are fine, but if I have a date such as
> "01/08/99", python runs into an error when it tries to eval("08").
>
> I can see how this might be a feature and not a bug, in that it
> interprets numbers starting in "O" as octyl.
>
> My workaround is the function:
>
> def chop_leading_zeros(str):
>     # removes a single leading zero from a string to be evaluated
>     # I think it interprets these numbers as Octyl
>     if str[0] == '0':
>         str = str[1:]
>     return str
>
> Does anyone have a simpler or more elegant workaround?



From kalle at gnupung.net  Wed Dec 20 19:32:29 2000
From: kalle at gnupung.net (Kalle Svensson)
Date: Thu, 21 Dec 2000 01:32:29 +0100
Subject: Customizing IDLE
In-Reply-To: <3A4081C1.DDBBE08C@et.dtu.dk>; from be@et.dtu.dk on Wed, Dec 20, 2000 at 10:54:09AM +0100
References: <3A4081C1.DDBBE08C@et.dtu.dk>
Message-ID: <20001221013229.A11084@gnupung.net>

Sez Brian Elmegaard:
> I would like to customize IDLE and use with another language. I mainly need to
> change the keywords list and the compile command.
> 
> Can this be done easily, i.e., which files to change, and may I do it?

It should be easy to change the keywords, they are defined in the standard
library module keyword.  Change all 'import keyword' to 'import mykeyword',
which should be a module following the same format, but with your keywords.

Changing compile commands should not be too hard either, but I don't know
how to go about that.

I believe you are free to do whatever modifications you want to IDLE, as
long as you preserve copyright and stuff, but I don't know for sure.  There
should be some kind of licence distributed with IDLE, but I'm too lazy to
look for it.

You might be interested in http://idlefork.sourceforge.net/

HTH,
  Kalle
-- 
Email: kalle at gnupung.net     | You can tune a filesystem, but you
Web: http://www.gnupung.net/ | can't tune a fish. -- man tunefs(8)
PGP fingerprint: 0C56 B171 8159 327F 1824 F5DE 74D7 80D7 BF3B B1DD
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 240 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20001221/d8c2c6b0/attachment.sig>

From joconnor at mail.cybermesa.com  Fri Dec  8 19:57:22 2000
From: joconnor at mail.cybermesa.com (Jay O'Connor)
Date: Sat, 09 Dec 2000 0:57:22 GMT
Subject: Python wrapper for C++ STL?
Message-ID: <200012090057.RAA09005@mail.cybermesa.com>

An embedded and charset-unspecified text was scrubbed...
Name: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20001209/7939106c/attachment.ksh>

From gregj at pobox.com  Wed Dec 20 01:19:39 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Wed, 20 Dec 2000 06:19:39 GMT
Subject: while true: !!!
References: <mailman.977180344.10242.python-list@python.org> <91noj60178k@news2.newsguy.com>
Message-ID: <%bY%5.172262$U46.5440607@news1.sttls1.wa.home.com>

This new learning amazes me, Sir Bedevere. Explain again how sheep's
bladders may be employed to prevent earthquakes.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From scarblac at pino.selwerd.nl  Thu Dec 28 12:26:00 2000
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 28 Dec 2000 17:26:00 GMT
Subject: Python vs PHP (& Java?)
References: <mailman.977897649.4877.python-list@python.org>
Message-ID: <slrn94mtng.i3d.scarblac@pino.selwerd.nl>

Ben Ocean <zope at TheWebsons.com> wrote in comp.lang.python:
> Plus, I'm curious as to why Python wasn't written in such a way that
> it could be written directly into HTML documents like PHP.

One reason I can think of is that Python is older than HTML...

-- 
Remco Gerlich


From amyk at foretec.com  Wed Dec 20 17:10:59 2000
From: amyk at foretec.com (Amy Katherine)
Date: Wed, 20 Dec 2000 17:10:59 -0500
Subject: ***The Ninth International Python Conference***
Message-ID: <4.2.2.20001219134444.00a8c460@colossus.foretec.com>

Greetings!

The Ninth International Python Conference will be held on March 5-8, 2001, 
at the Hilton Long Beach, in Long Beach, California.  The conference will 
consist of two days of refereed paper and application tracks  - including a 
Zope Track and a multi-technology Python Applications Track - as well as 
keynote speakers, a poster display, and a vendor exhibit (new!).  The 
conference will be preceded by a day of tutorials and will be followed by a 
developers' day.  For information on conference events and activities, 
please go to http://www.python9.org.

Registration for Python 9 is now open.  To register for the conference, 
please go to http://www.python9.org/p9-reginfo.html.  Early Bird 
registration closes on February 2, 2001.  On-line registration closes on 
February 25, 2001.

A limited block of rooms has been reserved at the Hilton Long Beach for 
Python 9 attendees.  To learn more about the hotel, and to reserve a 
sleeping room at the special conference rate, please go to 
http://www.python9.org/p9-hotelinfo.html.

If you are interested in sponsoring and/or exhibiting at the conference, 
information is available at http://www.python9.org/p9-sponsors.html. 
Alternately, you may send a message to sponsors at python9.org, or 
exhibitors at python9.org.

To receive periodic emails with new information about conference events and 
activities, please subscribe to the Python 9 mailing list at 
http://www.python9.org/p9-announce.html.

---------------------------------------------------------------------------- 
---------
Python 9 is Sponsored by ActiveState (www.activestate.com). 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20001220/26581bbf/attachment.html>

From glenfant at equod.com.nospam  Fri Dec 22 09:40:37 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Fri, 22 Dec 2000 15:40:37 +0100
Subject: __init__ like in java
Message-ID: <91vosi$19h$1@reader1.imaginet.fr>

Hi,

Is there an *elegant pythonic*  way to have several constructors for one
class ?

TIA



From thomas at xs4all.net  Wed Dec 27 05:00:37 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Wed, 27 Dec 2000 11:00:37 +0100
Subject: Python vs. PHP (& Java?)
In-Reply-To: <LOBBJMHOIFOLFIPFANHJMEMHCFAA.beno@TheWebsons.com>; from beno@TheWebsons.com on Tue, Dec 26, 2000 at 10:03:05PM -0800
References: <LOBBJMHOIFOLFIPFANHJMEMHCFAA.beno@TheWebsons.com>
Message-ID: <20001227110037.G25820@xs4all.nl>

On Tue, Dec 26, 2000 at 10:03:05PM -0800, Ben Ocean wrote:

> I?m an intermediate-level programmer that?s new to Python (via Zope) and I
> like it a lot. However, it appears to me that, while Python is a good
> replacement for Perl, it cannot replace much of the functionality of PHP,
> which can be scripted into HTML (whereas Python, it appears, must be
> called). So, I have to wonder about the speed of Python vs. PHP for
> Internet-based applications. Python has the advantage of byte-code and
> compilation, but it?s server-side. PHP might not be as fast on the server,
> but it doesn?t have to make as many trips back and forth between server
> and client. Plus, I?m curious as to why Python wasn?t written in such a
> way that it could be written directly into HTML documents like PHP.

You don't need roundtrips with Python either, you just need to start it as a
CGI script (which is some additional overhead, granted.) And note that there
are several extentions (PSP, ASP and if-I-recall-correctly mod_snake for
apache, and Zope's DTML) that allow embedding Python into HTML.

As for why Python isn't PHP, well, because it isn't. Python is a fair bit
older than PHP, and simply wasn't intended as an 'html-embedded-language'.
It was intended as a proper, easy to learn, easy to read, efficient (from
the programmers view) language, and not as a quick-and-dirty-hack to allow
code inbetween HTML :) The WWW and HTML did exist, but were only just
starting, when Python emerged.

And then there's two other points. I do a reasonable amount of PHP, and I
think it sucks :) I don't see a terribly good reason to mix HTML and script
code, for one. It looks sexy but there isn't a real advantage, and it only
serves to clutter the code, IMHO. And besides that I've been bitten by
irregularities in PHP code way too often, like how you can pass environment
variables to the html page which end up variables inside the script --
that's been the cause of several huge security issues for me. So even if you
really want to embed code into HTML, I'd go for Python (or even Perl!)
rather than PHP.

> On a similar subject, someone once informed me that Java was by far the
> best language for scripting communication programs over the Internet (like
> chat, IRC, etc.). How does Python compare? Your comments are appreciated.

Well, not suprising (since this is a Python newsgroup/list, not a Java one)
I don't agree. Python is easier to write and easier to read (and
understand), and way easier to learn. Python is (supposedly :) slower than
Java in execution, but especially in networked environments, where you spend
most of your time waiting for the network anyway, you aren't going to notice
that.

And besides, Python is so easy, you can learn it pretty quickly (in a few
hours if you're already a somewhat-experienced programmer) and I doubt
you'll ever regret it. If anything, it'll be a good programming learning
experience. And it's just a fraction of the time you'll spend learning Java
:)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From syver at NOSPAMcyberwatcher.com  Fri Dec  8 23:25:54 2000
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Sat, 9 Dec 2000 05:25:54 +0100
Subject: reloading modules question
Message-ID: <90scdf$hpe$1@troll.powertech.no>

Is it safe to loop through sys.modules.items()
reloading each module, as long as I catch the ImportError?

Am I guaranteed to have all python modules reloaded if I do the loop twice?

import sys
for key, value in sys.modules.items():
    if value:
    try:
         reload(value)
    except ImportError, e:
         print '%s\n' % e

The reason I am asking is that IIS webserver keeps Python in memory so that
it won't use the new versions of python modules I deploy if don't stop and
start the webserver. I can't afford to stop the webserver all the time, so I
am considering ways to reload the python modules while it's running.




From tseaver at starbase.neosoft.com  Sun Dec 10 19:38:21 2000
From: tseaver at starbase.neosoft.com (Tres Seaver)
Date: 10 Dec 2000 18:38:21 -0600
Subject: corba, soap, com, etc...
References: <pKsY5.58762$yR4.1496678@news1.rdc1.tx.home.com> <3QzY5.9825$dg6.637046@pouncer.easynews.com>
Message-ID: <1E4B6DCBE02D9689.190C3191326E0E6B.806861993926703B@lp.airnews.net>

In article <3QzY5.9825$dg6.637046 at pouncer.easynews.com>,

> Check out fnorb.  http://www.fnorb.org  It's an excellent open source
> CORBA implementation for python.

While I like fnorb, and use it, its license doesn't (quite) fit the
Open Source Definition (http://www.opensource.org/osd.html).

It also appears to be in limbo at the moment;  DSTC can't quite
figure out whether to continue funding support for it, but they
won't relax the licensing for it, either.

Tres.
-- 
===============================================================
Tres Seaver                                tseaver at digicool.com
Digital Creations     "Zope Dealers"       http://www.zope.org


From mwh21 at cam.ac.uk  Tue Dec 12 18:59:44 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 12 Dec 2000 23:59:44 +0000
Subject: Python-like compiled language
References: <913jr3$dmv$1@sylvester.vcn.bc.ca> <3A35835B.AAA92005@bioeng.ucsd.edu> <kuzoi178kg.fsf@lasipalatsi.fi> <3A36B288.FD0524DF@bioeng.ucsd.edu>
Message-ID: <m37l552ntb.fsf@atrus.jesus.cam.ac.uk>

Curtis Jensen <cjensen at bioeng.ucsd.edu> writes:

> If C/C++ isn't the closest, then what is?

The Lisps of the world are very similar in some ways (variables being
bindings, dynamic typing, interactive, introspective (well, CL, at
least)), but very different in others (syntax, the macro system).

Cheers,
M.

-- 
  Premature optimization is the root of all evil in programming.  
                                                       -- C.A.R. Hoare


From dsh8290 at rit.edu  Tue Dec 12 23:31:35 2000
From: dsh8290 at rit.edu (D-Man)
Date: Tue, 12 Dec 2000 23:31:35 -0500
Subject: [Q] Are Exceptions used that much in practice?
In-Reply-To: <3A36C6E3.26181E3F@enteract.com>; from goose@enteract.com on Tue, Dec 12, 2000 at 19:46:27 -0500
References: <3A35D546.A4731B5B@enteract.com> <3A36596A.89EEA210@alcyone.com> <3A36C6E3.26181E3F@enteract.com>
Message-ID: <20001212233135.X17575@westsidecnc.rh.rit.edu>

I would go for python because it has more features (ie classes) and is much
easier to read/maintain.  As others have stated, Python doesn't enforce
exception catching as Java does, and doesn't even allow you to declare (as part
of the code) what exceptions a funciton might throw.  However, exceptions are
very useful.  For example, here is some C code that I wrote for a course not too
long ago:

==============================

int
main( int argc, char** argv )
{
	FunctionData data ;
	FILE* input_file ;
	float result ; /* the root of the function */
	int i ; /* a loop counter */
	int status ; /* the return value from scanf */

	/* to begin with, register the function to call when the program ends */
	atexit( &exit_func ) ;


	/* check the number of arguments */
	if ( argc != 2 )
	{
		/* print a usage message */
		puts( "Invalid arguments." ) ;
		puts( "Usage:  roots <filename>" ) ;
		exit( EXIT_FAILURE ) ;
	} /* if ( argc != 2 ) */


	/* open the file */
	input_file = fopen( argv[1], "r" ) ;

	/* check for errors */
	if ( errno != 0 )
	{
		/* print an error message */
		perror( "" ) ;
		exit( EXIT_FAILURE ) ;
	} /* if ( errno != 0 ) */


	/* get the endpoints from the file */
	status = fscanf( input_file, "%f %f", &data.endpoints[0], 
			&data.endpoints[1] ) ;

	/* check for errors */
	if ( status != 2 )
	{
		fprintf( stderr, "Missing endpoint.\n" ) ;
		exit( EXIT_FAILURE ) ;
	} /* if ( errno != 0 ) */


	/* get the number of terms in the polynomial */
	status = fscanf( input_file, "%d", &data.func.num ) ;

	/* check for errors */
	if ( status != 1 )
	{
		fprintf( stderr, "Bad file: couldn't read number of terms.\n" );
		exit( EXIT_FAILURE ) ;
	} /* if */


	/* check for validity of input */
	if ( data.func.num < 3 || data.func.num > 10 )
	{
		fprintf( stderr, "The number of terms must be bewteen 3 and " );
		fprintf( stderr, "10, inclusive.\n" ) ;
		exit( EXIT_FAILURE ) ;
	} /* if ( ! 3 < data.func.num < 10 ) */

	polynomial_init( &data.func ) ;

	/* now read the coefficients into the polynomial */
	for( i = 0 ; i < data.func.num ; i++ )
	{
		status = fscanf( input_file, "%f", &data.func.coeffs[i] ) ;

		/* check for error reading file */
		if ( status != 1 )
		{
			fprintf( stderr,
				"There are too few values for coefficients.\n");
			exit( EXIT_FAILURE ) ;
		}
	} /* for */


	/* now read the exponents into the polynomial */
	for( i = 0 ; i < data.func.num ; i++ )
	{
		status = fscanf( input_file, "%d", &data.func.exp[i] ) ;
		/* check for error reading file */
		if ( status != 1 )
		{
			fprintf( stderr,
				"There are too few values for exponents.\n" ) ;
			exit( EXIT_FAILURE ) ;
		} 
	} /* for */


	/* find the root */
	result = bisect( data, EPSILON ) ;

	/* print the information */
	printf( "The given interval is [%4.2f, %4.2f].\n\n", \
			data.endpoints[0], data.endpoints[1] ) ;
	printf( "The given equation is:\n" ) ;
	print_polynomial( stdout, data.func ) ;
	putchar( '\n' ) ;
	printf( "Approximate value of the root is: %8.6f\n", result ) ;

	polynomial_clear( &data.func ) ;

	return EXIT_SUCCESS;
} /* main */

==============================

Since C doesn't have exception handling, I need to clutter the code with if
statements to check for errors after each operation.  If you're not paying
attention to the code, or I hadn't commented it you wouldn't be able to
determine what it should do very easily.  If I write the same funcitonality
(only for the beginning though) using Python and assuming that all functions
will throw an exception if there is an error it would look like this:

==================================
EXIT_SUCCESS = 0 
EXIT_FAILURE = 1

# check the number of arguments
if ( len( sys.argv != 2 ) :
	# print a usage message
	puts( "Invalid arguments." )
	puts( "Usage:  roots <filename>" )
	sys.exit( EXIT_FAILURE )
# end if


try :
	# open the file
	input_file = open( sys.argv[1], "r" )

	# get the endpoints from the file
	# hmm, guess I'm not that familiar with input in python, I'll just leave
the C here
	input_file.fscanf( "%f %f", &endpoints[0], &endpoints[1] )

	# get the number of terms in the polynomial
	input_file.fscanf( "%d", &num ) 

	# check for validity of input
	if ( not ( 3 < num < 10 ) ) :
		print >> sys.stderr, "The number of terms must be bewteen 3 and
10, inclusive."
		sys.exit( EXIT_FAILURE ) ;
	# end if

	# now read the coefficients into the polynomial
	for i in range( num ) :
		coeffs[i] = input_file.fscanf( "%f" )
	# end for

	# rest of processing here
except :
	input_file.close()
	print >> sys.stderr , "There was an error processing the input."

# now the program terminates
==================================

I think you can see how much easier it is to read the flow of execution.  This
program is supposed to get a filename on the commandline then read a polynomial
from the file and find a root of it.

Of course, most of this would be written as support functions that would be
executed as part of the ctor (__init__ function) in a class called Polynomial. 
I choose not to do a complete rewrite here, and also gives a closer comparison
to the original C.


In answer to your question, exceptions are very useful and convenient and
improve the readiabilty of the programs/functions greatly.  They also provide an
easy way to terminate a function/program if some bad error condition occurs. 
(ie, if no one catches an exception the program will terminate).  I would pick a
language that has exceptions over a language that doesn't any day.

I also feel that I should point out a flaw in my example:  it seems to indicate
that you can't make useful error messges using exceptions.  That is not so, I
simply didn't put the time in to do so.  When you catch exceptions, you can
specify the types of exceptions to catch.  Each type of exception can have it's
own 'except' block that can produce different results..

HTH

-D


On Tue, 12 Dec 2000 19:46:27 Jerome Mrozak wrote:
 | I believe I've been misunderstood.  The problem is not "can I write code
 | without handling exceptions" but rather "are exceptions used
 | voluntarily, rather than by compulsion".
 | 
 | An example of compulsion is java.lang.Integer.parseInt(), which throws a
 | NumberFormatException you *must* deal with or the program won't
 | compile.  I'm aware of examples of Python such as (faking it here):
 | 
 | while 1:
 |   try:
 |     input = inputFileObject.readline()
 |   catch EOF:
 |     break
 |   else:
 |     process(input)
 | 
 | The point of the question is this -- is code like this sample a way of
 | exercising the try-catch, or is this an example of something I end up
 | doing all of the time because it is so much easier than the alternative?
 | 
 | Other posters to this question provided a fair sampling of examples.
 | 
 | Some, I'm disappointed to say, were, shall we say, not so supportive.  
 | 
 | The long-range goal of my question is to determine for my satisfaction
 | whether it is worth while for me to spend my limited time pursuing
 | Python as a compliment to (whatever other skills I might have) or to go
 | with the apparently much greater flow and use those resources on Perl. 
 | The 'exception' thing could be a determiner, esp. with nothing similar
 | in Perl, but only if I can use it in a *positive* way instead of just to
 | test a return value.  Heck, Perl has 'die' for that kind of testing.
 | 
 | Thanks for your concern anyways.
 | 
 | Jerome.
 | 




From neelk at alum.mit.edu  Tue Dec 26 18:30:00 2000
From: neelk at alum.mit.edu (Neelakantan Krishnaswami)
Date: 26 Dec 2000 23:30:00 GMT
Subject: One Python 2.1 idea
References: <mailman.977771949.17911.python-list@python.org> <9292tn$jka$1@nnrp1.deja.com> <929n9c013m3@news1.newsguy.com> <92ab9u$dou$1@nnrp1.deja.com> <92ap0i22b6r@news1.newsguy.com>
Message-ID: <slrn94ie9k.anf.neelk@alum.mit.edu>

On Tue, 26 Dec 2000 19:43:52 +0100, Alex Martelli <aleaxit at yahoo.com> wrote:
><rturpin at my-deja.com> wrote in message news:92ab9u$dou$1 at nnrp1.deja.com...
>>
>> Most work in a Python program involves object methods
>> or attributes. (At least, this is the case in my
>> programs. ;-) Increasing the speed of reference to
>> these should have appreciable benefit, across the
>> board. Yeah, it would make the code a bit more
>> complex to maintain. But this isn't a special case --
>> it is a general mechanism central to how Python works.
>
> It does seem likely that caching lookups is where most
> across-the-board performance might be gained, yes.  The trick is
> finding out a way to invalidate the cache every time some dynamic
> manipulation may need lookups to happen afresh.

I'm not sure this would help very much. Tim posted a link to an
attempt to add method caches, and reported inconclusive results.  As
of late last week, I'm not surprised; the Python interpreterseems to
be spending most of its time in bytecode instruction dispatch. Logic
suggests that that's where performance enhancements would have the
most impact.

Fortunately, I don't have to put my money where my mouth is, since
someone has already done it for me. :) Vladimir Marangozov has
modified the Python interpreter (as of 1.5) to use the GNU computed
labels extension and implemented direct threading.

  http://starship.python.net/crew/vlad/archive/threaded_code/

Bringing that up to date with 2.0 would be an interesting experiment. 


Neel


From tismer at tismer.com  Mon Dec 11 11:29:26 2000
From: tismer at tismer.com (Christian Tismer)
Date: Mon, 11 Dec 2000 18:29:26 +0200
Subject: corba, soap, com, etc...
References: <pKsY5.58762$yR4.1496678@news1.rdc1.tx.home.com> <3QzY5.9825$dg6.637046@pouncer.easynews.com> <912ls3$bn8$1@violet.singnet.com.sg>
Message-ID: <3A3500E6.7C8802FC@tismer.com>


Ng Pheng Siong wrote:
> 
> According to Jiva DeVoe  <jiva at no-spam.com>:
> > Check out fnorb.  http://www.fnorb.org  It's an excellent open
> > source CORBA
> > implementation for python.
> 
> Just downloaded Fnorb. According to the docu, it supports
> BOA. BOA has been superceded by POA.
> 
> See the CORBA FAQ here:
> 
>     http://www.javacommerce.com/CorbaFaq/toc-exhaustive.html
> 
> OmniOrb3 supports POA:
> 
>     www.uk.research.att.com/omniORB/omniORBpy/

I have been using FNORB as well for a while, but now have
switched to omniORB. I must say the product is just
*excellent* ! The IDL compiler is correct and smart even
with large projects like ours, where Fnorb simply gave up,
without an error message, just nothing.
Also, onmiORB allows you to run more than one ORB in one
process, since it is multithreaded from the ground up.
We even had the effect that a client was talking to my ORB,
which was started in PythonWin, and I thought it was
stopped. In fact it was still active, as a thread.
OmniORB comes with a GPL license, btw.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From sholden at holdenweb.com  Thu Dec 21 08:42:04 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 21 Dec 2000 08:42:04 -0500
Subject: Editing files
References: <Bdd06.5591$K6.674078@juliett.dax.net>
Message-ID: <oOn06.2898$PP2.31556@e420r-atl2.usenetserver.com>

dv <lars at stea.nu> wrote in message
news:Bdd06.5591$K6.674078 at juliett.dax.net...
> I'm trying to automate some configuration tasks with python. Is it
possible
> to do a "search and replace" or specify where (which line) in a file
python
> will write?
>
> Thanks in advance
>
> Lars
>
Sounds like you want to update files in place rather than read the old one,
write a new one and replace the old one with it.  Am I right?

It's certainly possible to use the tell() and seek() file primitives to find
out where you are in a file and direct output to a particular location in
that file, if that's what you want.

The documentation should give you the rest of the story.  For Python 2.0 see
section 2.1.7.9 - File Objects.  Hope this helps.

regards
 Steve





From lists at TheWebsons.com  Sat Dec 23 15:18:07 2000
From: lists at TheWebsons.com (Ben Ocean)
Date: Sat, 23 Dec 2000 12:18:07 -0800
Subject: Why Won't This Work?
In-Reply-To: <slrn949gag.mr.amk@207-172-111-138.s138.tnt1.ann.va.dialup.rcn.com>
Message-ID: <LOBBJMHOIFOLFIPFANHJGEGACFAA.lists@TheWebsons.com>

Hmmm. *python* at a command prompt, sans *./* is a command. This worked when
I configured 2.0, which I had to rip out because it's not compatible with
Zope. But you're the *second* person to recommend this! What am I not
seeing? (Blind, you know ;-)
BenO

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of A.M. Kuchling
Sent: Saturday, December 23, 2000 7:01 AM
To: python-list at python.org
Subject: Re: Why Won't This Work?


On Sat, 23 Dec 2000 06:31:38 -0800, Ben Ocean <lists at TheWebsons.com> wrote:
>and everything appears to be in order. But when I type
>python
>at a command prompt, I get a bash error. What have I not done?

Probably it's very simple; do you have '.', the current working
directory, in your PATH variable?  Try ./python.

--amk
--
http://www.python.org/mailman/listinfo/python-list




From olczyk at interaccess.com  Sat Dec 23 20:17:05 2000
From: olczyk at interaccess.com (Thaddeus L. Olczyk)
Date: Sun, 24 Dec 2000 01:17:05 GMT
Subject: Problem with python mode
References: <3a44bcf1.73260781@nntp.interaccess.com> <3A44BF14.F7721A06@is.elta.co.il>
Message-ID: <3a464e89.3641015@nntp.interaccess.com>

On Sat, 23 Dec 2000 17:04:53 +0200, in comp.emacs you wrote:

>"Thaddeus L. Olczyk" wrote:
>> 
>> When i try loading pdb.el in NTemacs I keep getting an error--
>> Symbol's function defiition void: toolbar-make-button-list.
>> Any idea what the problem is?
>
>This is an XEmacs function, it doesn't exist in NTEmacs.
>
>Are you sure pdb.el works with Emacs?
I seem to remember it doing so.
It comes after the line:
(if (featurep 'xpm)

so I suspect that xpm is not supposed to be present in NTEmacs.
The question then is why does it test as being present?


From aleaxit at yahoo.com  Tue Dec 19 08:22:04 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 19 Dec 2000 14:22:04 +0100
Subject: embedding help
References: <3A3F0EFC.D2F2BAE9@ic.sunysb.edu>
Message-ID: <91nnek0166d@news2.newsguy.com>

"Jp Calderone" <jcaldero at ic.sunysb.edu> wrote in message
news:3A3F0EFC.D2F2BAE9 at ic.sunysb.edu...
>
>
>   I'm trying to embed python in a much larger project (a MUD)
> as a scripting language, but I'm stuck, essentially right at
> the begining.  The docs at python.org on this section seem,
> ahh, incomplete (unless I'm missing something).  What I'm

They _are_ a bit incomplete -- some of the Python C API
docs just list relevant functions &c without any info
about them.

> trying to do ATM is something along the lines of let python
> code access and modify the values of C++ objects outside of
> the interpreter.  I believe the correct way to go about this
> is to create a new python object type that understands how
> to find the C++ objects in the C++ portion of the program,
> providing it with get and set attr methods.  However, I'm

Yes, that makes sense.  If you're _embedding_ Python,
rather than just _extending_ it, you'll still create a
Python module for the purpose, but 'pump it into Python'
just before or just after you start it, rather than
have it loaded when needed via import statements.

> at a loss as to how to actually do this.  I found the
> PyTypeObject struct, and I see it has a lot of fields for
> functions that do things, but no actual data storage, so
> I don't see any way to keep track of which python object
> instances points to which C++ object instances.  Can anyone
> point me to any examples of how to do this, or explain it?
> Or am I barking up the wrong tree entirely?

Since you're C++-bound anyway, I think that much the
easiest way for you to work is via the Boost Python
library, www.boost.org; once named py_cpp -- like all
of Boost, it does require a C++ compiler with halfway
decent templates (maybe CXX or other such projects
may still help you even with older C++ compilers...),
but gcc and Visual C++ 6, for example, work with it.

For C-level work, the ideal read would be "How to
write a Python extension",
http://starship.python.net:10080/crew/arcege/extwriting/pyext.html
except that accessing anything on Starship is always
a bother these days (I wonder if mirrors of this may
be found anywhere else...?).  Worst case, Google IS
fortunately caching it (bless it!)...: enter the
following very long URL on ONE single line...:
http://www.google.com/search?q=cache:starship.python.net:10080/crew/arcege/e
xtwriting/pyext.html

If you do use Google's cache, remember you have to force
its use manually for links in the page that also refer to
the Starship (assuming the Starship is still down, that is).


Anyway, to address your specific question, the C-level
structure may be something like...:

typedef struct oneVariable {
    const char* name;
    int* pValue;
} oneVariable;

typedef struct {
  PyObject_HEAD
  oneVariable variables[NUM_VARIABLES];
} myObject;

that is, your object are defined by _following_ the
'Python head' with your own data -- a sort of C level
emulation of C++'s simplest kind of inheritance (which
is part of why the C++ encapsulations of Python
interfacing are just SO much handier than the C level
ones... but, enough C++ haters lurk around here that
I'm sure to get flamed for this assertion!-).


Alex





From moshez at zadka.site.co.il  Tue Dec 26 01:59:04 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Tue, 26 Dec 2000 08:59:04 +0200 (IST)
Subject: os.exec?
In-Reply-To: <jwbnews-82EF1A.10532625122000@news.olympus.net>
References: <jwbnews-82EF1A.10532625122000@news.olympus.net>, <LXw16.8561$KY1.23035@news1.rivrw1.nsw.optushome.com.au> <mailman.977727428.19738.python-list@python.org>
Message-ID: <20001226065904.584EDA84F@darjeeling.zadka.site.co.il>

[deadmeat]
> How to I execute an external program without replacing the Python
> interpreter? I want control to return to Python when the external 
> program is completed. I still want stdio to go to the console not to a pipe 
> if possible.

[Moshe Zadka]
> Quite possibly, you'd want os.system, and os.popen.
> Or, you might want to work on a lower level with os.fork/os.exec/os.open.

[John W. Baxter]
> Or, moving to a higher level, there is the commands module, which wraps 
> those things mentioned above.  I tend to be lazy and use that.

And while we're on a higher level, the pipes module does deserve an
honourable measure.
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From jh at web.de  Mon Dec  4 18:01:59 2000
From: jh at web.de (Jürgen Hermann)
Date: Tue, 5 Dec 2000 00:01:59 +0100
Subject: Python 2.0  libreadline.so.3 unresolved dependency
References: <3A28405C.EB87CB6C@kahala.net> <3A2AD4D6.B441EA09@monitor.net>
Message-ID: <90h7pd$hl2$02$1@news.t-online.com>

"Stephen R. Figgins" <fig at monitor.net> schrieb im Newsbeitrag
news:3A2AD4D6.B441EA09 at monitor.net...
> I don't have an answer for your original question, but I was wondering
> where the Wiki Python 2.0 FAQ can be found.

http://www.python.org/2.0/ ==> Post-release news

or directly

http://www.python.org/cgi-bin/moinmoin (not available right now, Guido
installing the new release?)




From yelled at yahoo.com  Sun Dec 10 14:33:13 2000
From: yelled at yahoo.com (yelled at yahoo.com)
Date: Sun, 10 Dec 2000 19:33:13 GMT
Subject: _oleobj_ and dir
Message-ID: <910lpp$p2g$1@nnrp1.deja.com>

Hi,

 I am trying to get available functions and attributes of COM obj:
	adoConn = win32com.client.Dispatch('ADODb.connection')
	adoConn.Open(connStr('roux'))
	sqlStr = "SELECT ticker, cusip FROM security WHERE ticker in (%
s)" %StkList
	if dbg: sqlStr
	(adoRst, TF) = adoConn.Execute(sqlStr)
 if I do dir(adoRst):
     ['_oleobj_']
 and dir(rst._oleobj_)
['GetIDsOfNames', 'GetTypeInfo', 'GetTypeInfoCount', 'Invoke', 'InvokeTy
pes', 'QueryInterface']

   I really want attributes like GetRows and others associated with ADO.
How do I get that?
  Thanks,

Hoon,


Sent via Deja.com http://www.deja.com/
Before you buy.


From francois.granger at free.fr  Sun Dec 10 06:11:22 2000
From: francois.granger at free.fr (=?ISO-8859-1?Q?Fran=E7ois_Granger?=)
Date: Sun, 10 Dec 2000 11:11:22 GMT
Subject: Tkinter prob in Python 2.0, Macintosh
References: <muvol.remove-this-ABD013.16350709122000@news.knology.net>
Message-ID: <1eledzv.iqxaj9154sl7mN@paris11-nas9-37-96.dial.proxad.net>

Mike <muvol.remove-this at knology.net> wrote:

> Am I missing something particular to Tk on the Mac?  I would appreciate any
> suggestions.  

As far as I know, tk does not works well with the IDE. Someone will be
more precise on this.

-- 
"il faut bien r?fl?chir avant de prendre un animal quel qu'il soit"
___________________________________________________________________
- Francesca Binet-Giannoni


From amk at mira.erols.com  Sun Dec 24 10:55:56 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 24 Dec 2000 15:55:56 GMT
Subject: PEP 222 draft
References: <slrn947bj1.1rs.amk@207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com> <3A43AFEE.AC8F7BD@stroeder.com>
Message-ID: <slrn94c7ti.m4.amk@207-172-57-128.s128.tnt2.ann.va.dialup.rcn.com>

On Fri, 22 Dec 2000 20:47:58 +0100,
	Michael Str?der <michael at stroeder.com> wrote:
>This method should either take a parameter or use an object
>attribute to check the currently used cipher used against a list of
>accepted ciphers to enforce a local policy or be renamed to
>something like isSslConn() which does not mean much...

I'd prefer the renaming; the criteria for accepting a particularly
cipher could be complicated, and I'd rather such methods be added by
subclasses.

(I hope the lack of response to the PEP is because everyone's off for
the holidays.)

--amk








From jasonic at nomadicsltd.com  Thu Dec 28 22:44:40 2000
From: jasonic at nomadicsltd.com (Jason Cunliffe)
Date: Thu, 28 Dec 2000 22:44:40 -0500
Subject: Great New Book
Message-ID: <t4o1vr8hktk531@corp.supernews.com>

This morning I wes very happy to see in my local Barnes&Noble today a
prominent stack of an excellent new book:
'Core Python Programming' by Wesley Chun.

http://www.amazon.com/exec/obidos/ASIN/0130260363/o/qid=978061249/sr=8-1/ref
=aps_sr_b_1_3/002-1140792-5636848

IMHO This is #1 the best written book on Python I have yet set eyes on.
Wesley Chun has a deceptively clear prose style, and strikes a clean balance
between the obvious but essential, yet carefully elucidating famous pythonic
topics.

I really thingk this is the most suitable entry text for Python which
presently exists. It is writen with grace and skill by an author who
obviously wnows his topic. It lays a great foundation.

I truly look forwards to any more volumes he may produce in this series.

The only shame is not the author's fault - it is the still crazy high price
$45 and thick paper which bulks the 760pages to 2.5" thick.
I wish american publishers would stop this telphone book web publishing
madness & follow the example set by Korean technical books I have seen -
thin paper, larger comfortable format, plenty of color and excellent step by
step illustrations.



- Jason
___________________________________________________________
Jason CUNLIFFE = NOMADICS['Interactive Art and Technology']




From thomas at xs4all.net  Mon Dec 25 17:32:40 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Mon, 25 Dec 2000 23:32:40 +0100
Subject: One Python 2.1 idea
In-Reply-To: <9289mq$5vq$1@panix6.panix.com>; from aahz@panix.com on Mon, Dec 25, 2000 at 12:12:10PM -0800
References: <mailman.977771949.17911.python-list@python.org> <9289mq$5vq$1@panix6.panix.com>
Message-ID: <20001225233240.B25820@xs4all.nl>

On Mon, Dec 25, 2000 at 12:12:10PM -0800, Aahz Maruch wrote:

> (I'm highlighting it just in case there's someone stupid out there who
> doesn't read every Tim Peters post top-to-bottom.)

Now *there's* an urban myth if I ever saw one :)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From tim.one at home.com  Mon Dec  4 19:31:47 2000
From: tim.one at home.com (Tim Peters)
Date: Mon, 4 Dec 2000 19:31:47 -0500
Subject: Python interpreter hangs
In-Reply-To: <3A2B6853.2B0F7909@ilt.fhg.de>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEGHICAA.tim.one@home.com>

[posted & mailed]

[Rolf Wester]
> when I'm running the script that is listed below on one of our
> DEC-Alpha stations, python hangs after the print statement in
> the function test0.

How long did you wait for it to finish?

> The output "sum = 125000.0" is printed on the screen but the
> print statement at the end of the script is not executed.

So it reached the end of the function but hasn't returned from the function.
Between those two times, Python has to run around all over memory
decrementing refcounts on, and subsequently freeing, the approximately half
million temp array elements the function created.  That *could* take a long
time, depending on details of memory allocation, the platform libc, and the
platform's VM implementation.

> The process is not idle but is running and demands 100% of CPU-time.

Consistent with the above.

> Python does not react on Ctr C, I have to explicitely kill it.

Ditto, ditto.

> Curiously this script works on a NT-PC, on another DEC-Alpha
> station and it already worked on the same machine it's now hanging.

Wait longer before killing it; my guess is it will eventually finish if you
leave it alone.

> This is probably a problem with the DEC-Alpha station but  I
> don't have any idea what the reasons could be.  Did anybody have\
> a similar problem or  does anybody have an idea concerning my problem?

I don't think any of the Python developers use Alphas.  We've had weird
reports of supernaturally slow memory behavior on them before.  Until an
Alpha expert volunteers to dig into this stuff, though, it's likely to
remain A Mystery.

> import string
>
> def test0():
>  p  = []
>  iz = []
>  ix = []
>  iy = []
>  sum = 0.0
>  nn = 125001
>  heiz_dat0 = "12 23 33 1.0"
>  for i in range(1,nn):
>   zeile = string.split(heiz_dat0)
>   iz.append(zeile[0])
>   ix.append(zeile[1])
>   iy.append(zeile[2])
>   pf  = string.atof(zeile[3])
>   sum = sum + pf
>   p.append(pf)
>  print "sum = ", sum
>  return sum
>
> sum1 = test0()
> print "ende", sum1

One-at-a-time appending *may* be causing massive memory fragmentation.  You
could try preallocating the arrays and then indexing into them; e.g.,

iz = [None] * nn  # outside the loop
iz[i] = zeile[0]  # inside the loop

and similarly for the others.




From financejobNY at yahoo.com  Mon Dec 11 12:00:13 2000
From: financejobNY at yahoo.com (financejobNY at yahoo.com)
Date: Mon, 11 Dec 2000 17:00:13 -0000
Subject: Financial programming position in Manhattan
Message-ID: <91316t+sisv@eGroups.com>

A financial trading firm in midtown Manhattan seeks to hire a quality
applications programmer in Python and C/C++. The primary job is to 
develop tools to run the trading system in real time. The programmer 
would have some experience in real time systems, data acquisition, 
and simple database and spreadsheet development.

No financial experience is necessary. Compensation is competive
and weighted towards a significant annual bonus.

Applicants should send an e-mail describing their programming 
experience and job history to financejobNY at yahoo.com.




From joconnor at mail.cybermesa.com  Fri Dec  8 12:26:31 2000
From: joconnor at mail.cybermesa.com (Jay O'Connor)
Date: Fri, 08 Dec 2000 17:26:31 GMT
Subject: Refactoring browser
Message-ID: <200012081726.KAA12259@mail.cybermesa.com>

An embedded and charset-unspecified text was scrubbed...
Name: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20001208/9192402a/attachment.ksh>

From aleaxit at yahoo.com  Wed Dec 27 12:32:45 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 27 Dec 2000 18:32:45 +0100
Subject: regex, split and ()
References: <slrn94k8rq.18t.jp@localhost.localdomain>
Message-ID: <92d94u11av7@news2.newsguy.com>

"Jacek Pop?awski" <jp at ulgo.koti.com.pl> wrote in message
news:slrn94k8rq.18t.jp at localhost.localdomain...
    [snip]
> >>> w=re.compile(r'(<[^<>]*(".*")?>)')
> >>> s="<html> one <br> two <img src=\"<blah\"> three </html>"
> >>> w.split(s)
> ['', '<html>', None, ' one ', '<br>', None, ' two ', '<img src="<blah">',
> '"<blah"', ' three ', '</html>', None, '']
>
> works good, but why it double everything? probably becouse I used second
(),

Right.

> how to fix it?

Use non-grouping parentheses (?: ... ) for the inner group, for example:

>>> w=re.compile(r'(<[^<>]*(?:".*")?>)')
>>> s="<html> one <br> two <img src=\"<blah\"> three </html>"
>>> w.split(s)
['', '<html>', ' one ', '<br>', ' two ', '<img src="<blah">', ' three ',
'</html>', '']



Alex





From wester at ilt.fhg.de  Mon Dec  4 04:48:03 2000
From: wester at ilt.fhg.de (Rolf Wester)
Date: Mon, 04 Dec 2000 10:48:03 +0100
Subject: Python interpreter hangs
Message-ID: <3A2B6853.2B0F7909@ilt.fhg.de>

Hi,

when I'm running the script that is listed below on one of our DEC-Alpha
stations,
python hangs after the print statement in the function test0. The output
"sum = 125000.0"
is printed on the screen but the print statement at the end of the
script is not executed.
The process is not idle but is running and demands 100% of CPU-time.
Python does not react on Ctr C, I have to explicitely kill it. Curiously
this script
works on a NT-PC, on another DEC-Alpha station and it already worked on
the same
machine it's now hanging. This is probably a problem with the DEC-Alpha
station but  I
don't have any idea what the reasons could be.
Did anybody have a similar problem or  does anybody have an idea
concerning my problem?
I would be very appreciative for any help.

--Rolf


import string

def test0():
 p  = []
 iz = []
 ix = []
 iy = []
 sum = 0.0
 nn = 125001
 heiz_dat0 = "12 23 33 1.0"
 for i in range(1,nn):
  zeile = string.split(heiz_dat0)
  iz.append(zeile[0])
  ix.append(zeile[1])
  iy.append(zeile[2])
  pf  = string.atof(zeile[3])
  sum = sum + pf
  p.append(pf)
 print "sum = ", sum
 return sum

sum1 = test0()
print "ende", sum1



From aleaxit at yahoo.com  Sun Dec 24 02:55:22 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 24 Dec 2000 08:55:22 +0100
Subject: About WebBrowser
References: <91vbr9$eo1q2@SGI3651ef0.iddeo.es> <9235bj$6e6$1@troll.powertech.no> <9236f3$6rb$1@troll.powertech.no> <3a45260d.52012638@news.skynet.be>
Message-ID: <924a8b0ckb@news1.newsguy.com>

"Daniel" <Daniel.Kinnaer at AdValvas.be> wrote in message
news:3a45260d.52012638 at news.skynet.be...
    [snip]
> >import win32com.client
    [snip]
> Traceback (innermost last):
>   File "C:/Python20/AProject/test1/browzer.py", line 1, in ?
>     import win32com.client
> ImportError: No module named win32com.client

You probably have not installed the win32all extensions from ActiveState
(or the ActiveState distribution of Python, that comes with them); or else
something has gone awry in the installation process (seems unlikely).


> I suppose I need this  makepy-utility you write about? What do I have
> to do to make this sourcecode work for me?

Download the win32all extensions from ActiveState and install them.
http://www.activestate.com/Products/ActivePython/win32all.html
If you have Python 2, you need to get build 135 of win32all specifically
(a single URL, possibly split by news-handling software...):
http://www.activestate.com/download/ActivePython/windows/win32all/win32all-1
35.exe


Alex





From sholden at holdenweb.com  Wed Dec 13 18:25:15 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 13 Dec 2000 18:25:15 -0500
Subject: A question.
References: <9152a0$acb$1@panorama.wcss.wroc.pl> <9156h002uh2@news2.newsguy.com> <915a7j$fe0$1@panorama.wcss.wroc.pl> <915elr0681@news2.newsguy.com> <9166uv$8rs$1@animus.fel.iae.nl>
Message-ID: <cCTZ5.705$4y4.23160@e420r-atl1.usenetserver.com>

Carel Fellinger <cfelling at iae.nl> wrote in message
news:9166uv$8rs$1 at animus.fel.iae.nl...
> Alex Martelli <aleaxit at yahoo.com> wrote:
>
> ...excelent tutorial stuf skipped
>
> now that the tim-bot is exposed as a fake (he has a *face* !:)
> and stopped work^H^H^H^Hposting and C4E still the BDFL's pet
> I just wondered:
>
>   have they sprouted a new timbot, more geared towards newbies,
>   more polite and friendly maybe, with a touch of human fallibility
>   (hence the occasional slip of the keyboard) and named it Alex?
>
No, you are confusing two entirely different things.  The Martellibot Mark 1
has a completely European flavour to it, and adds a cosmopolitan touch of
linguistics to its output, sprinkling foreign language references in.  It is
similar to the timbot in its overall erudition, but can be distinguished
from it by its tendency to indulge in flamewars (which, I believe, ot does
mostly to convince us it is human).

regards
 Steve






From frc.removethistextfirstplease at eznet.net  Wed Dec 20 09:31:22 2000
From: frc.removethistextfirstplease at eznet.net (David Fisher)
Date: Wed, 20 Dec 2000 14:31:22 GMT
Subject: audio playback
Message-ID: <3a40c236.3935118@news.frontiernet.net>

The Library Ref lists:

"audiodev 
     -- Platform-independent API for playing audio data. "

but, apparently, nothing else on how to play .wav files in Python.
Where can I find more info on this?

Thanks!

	- David Fisher



From sholden at holdenweb.com  Thu Dec 14 10:06:37 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 14 Dec 2000 10:06:37 -0500
Subject: Changing Procesor Flags and word state From Python
References: <t3hg1fpdpvkg7d@corp.supernews.com>
Message-ID: <Oo5_5.7521$4y4.192209@e420r-atl1.usenetserver.com>

Jose Martin <joseamartin at venmex.com> wrote in message
news:t3hg1fpdpvkg7d at corp.supernews.com...
> Is there any way to work with CPU registers and the flags directly from
> python?
>
> ...?
>
No.  Python is a high-level language which can be implemented on many
different processor types, so it wouldn't make sense to provide hardware
abstractions of that nature.  People even complain that library modules are
specific to Windows or Unix!

regards
 Steve





From gregj at pobox.com  Tue Dec 19 01:38:27 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Tue, 19 Dec 2000 06:38:27 GMT
Subject: while true: !!!
References: <mailman.977127721.4398.python-list@python.org>
Message-ID: <DnD%5.166633$U46.5342944@news1.sttls1.wa.home.com>

"Blair Lowe" <Blair.Lowe at compeng.net> wrote:

> This is exactly what a python program should be NOT doing. "while 1"
> is one of the most cryptic statements in Computer Science.

That's a bit of an exaggeration. I thought we decided that Knuth's MIX
renditions of common algorithms were the most cryptic statements in Computer
Science.

> other languages are known for this stuff, but I do not know why python
> doesn't auto define true and false, so that we don't have to write
> unreadable stuff.

Adding a new boolean type to Python would add needless complexity to solve a
non-problem. Even a simple new type that consisted of only true and false
would require conversion rules that would affect every existing Python
library.

You can define true and false yourself, as many C/C++ header files do, but
then you run the risk of stuff like this:

    true = 1
    false = 0
    ...
    a = len(s)
    if a == true:  # fails unless a is 1

It's much harder to track this kind of thing down than learning once in your
career what "while 1" means.

> While "while 1" (pardon the pun) is a standard fixture that anyone
> with any experience knows what is going on, the whole point is to
> make all our code readable so that we, and others can maintain it and
> reuse it easily.

Exactly--that's why learning idioms is as important as learning syntax.

> Although the definition is cryptic, the resulting code is
> very readable. That is what we want.

Until it doesn't work. Encountering "while true:" in a program, a newbie may
assume that true is a Python keyword. I remember a C++ programmer I worked
with who though that true and false were keywords until he discovered this
(typical) definition in a header file:

    enum { false = 0, true = -1 };

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From jblazi at vipsurf.de  Tue Dec 26 13:17:39 2000
From: jblazi at vipsurf.de (Janos Blazi)
Date: Tue, 26 Dec 2000 19:17:39 +0100
Subject: Pari for Python
Message-ID: <3a48e0c7_5@news.newsfeeds.com>

Years ago I read about a Pari module for Python. Does such a module still
exists and where can it be downloaded from?

Janos Blazi




-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----


From gregholmes at my-deja.com  Mon Dec 18 06:23:46 2000
From: gregholmes at my-deja.com (gregholmes at my-deja.com)
Date: Mon, 18 Dec 2000 11:23:46 GMT
Subject: Tool name (posted here, can't recall), documentation generation
References: <slrn93r69b.ggs.jblaine@kickflop.ne.mediaone.net>
Message-ID: <91ks40$e30$1@nnrp1.deja.com>

DocBook?

That is not simple; actually it is more complex than HTML.  But it is
the only thing that comes to mind.


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


From amk at mira.erols.com  Sun Dec 10 10:37:41 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 10 Dec 2000 15:37:41 GMT
Subject: coding standards in cpython
References: <8vvvmp$93f$1@news.kth.se>
Message-ID: <slrn9379is.h4.amk@207-172-146-162.s162.tnt3.ann.va.dialup.rcn.com>

On Tue, 28 Nov 2000 11:02:48 +0100, 
	Oivvio Polite <oivvio at cajal.mbb.ki.se> wrote:
>I guess that there's a c coding standard used in the implementation of
>cpython.

I'm not sure it's written down in a document anywhere, If you're using
cc-mode in Emacs or XEmacs, c-set-style "python" will use the right
style -- 8-space indents, and blocks in this style:

        if (tstate == NULL) {
                PyInterpreterState_Delete(interp);
                return NULL;
        }

I forget what that style of brace placement is called.

--amk


From jcm at bigskytel.com  Tue Dec 26 05:09:03 2000
From: jcm at bigskytel.com (David Porter)
Date: Tue, 26 Dec 2000 03:09:03 -0700
Subject: Duplicated List...
In-Reply-To: <929lq7$ds1$1@hiline.shinbiro.com>; from dyo@shinbiro.com on Tue, Dec 26, 2000 at 05:54:25PM +0900
References: <929lq7$ds1$1@hiline.shinbiro.com>
Message-ID: <20001226030903.A30109@bigskytel.com>

* dyo <dyo at shinbiro.com>:

> >>> a=[1,2,3]
> >>> b=[a,10,20,30]
> >>> b[0].append('x')
> >>> b
> [[1, 2, 3, 1, 2, 3, 'x'], 10, 20, 30]
> >>> a
> [1, 2, 3, 1, 2, 3, 'x']
> >>> a.append('xxx')
> >>> a
> [1, 2, 3, 1, 2, 3, 'x', 'xxx']
> >>> b
> [[1, 2, 3, 1, 2, 3, 'x', 'xxx'], 10, 20, 30]

a and b[0] are references to the same object (try id() to test this).
So in both cases you are altering in place the same object. Since both a and
b[0] point to the same object they are both altered.

> But in this case..
> 
> >>> a=[1,1000]
> >>> a
> [1, 1000]
> >>> b
> [[1, 2, 3, 1, 2, 3, 'x', 'xxx'], 10, 20, 30]
> >>>
> 
> Why do not change List 'b' ?
> What happen List 'b' ?

You did not alter a in place this time. A new object was created with a
pointing to it. b[0] still points to the old object. Since we know from
above that altering in place will not create a new object, you can do this
instead:

>>> a[:]=[1,1000] 
>>> a
[1, 1000]
>>> b
[[1, 1000], 10, 20, 30]


David




From jml at ids.org.au  Wed Dec 13 06:06:58 2000
From: jml at ids.org.au (Jonathan Lange)
Date: 13 Dec 2000 22:06:58 +1100
Subject: Direct pixel-by-pixel drawing (with Tk)
Message-ID: <6izoi0bmwd.fsf@ids.org.au>

Hello,

I'm relatively new to Python (but not to programming), and I'd like to
muck around by doing some fractal stuff. However, alot of this stuff
needs (wants?) setting pixels directly.

Now, of course, one solution is to draw a bunch of lines of length 1,
but TK seems to want to remember each bloody line its drawn, which
is a bit of overkill and a memory hog.

I'm doing this on Linux and Windows, so a platform-independent
solution would rock.

Thanks,
jml

PS. I'm not afraid of reading documentation :)

-- 
Jonathan Mark Lange
http://www.flashkit.com - For God's sake Jim, I'm a programmer, not a designer!


From urner at alumni.princeton.edu  Wed Dec 20 16:04:34 2000
From: urner at alumni.princeton.edu (Kirby Urner)
Date: Wed, 20 Dec 2000 13:04:34 -0800
Subject: Python in "math through programming" curriculum
References: <2g2t3t06r2agrrjmfnr41eu4lc2u39a8qi@4ax.com> <3A3EFF47.67DE69CF@gte.net>
Message-ID: <kj624toonmgugtrgrpkea9b01a56hgogam@4ax.com>

"Clifford J. Nelson" <cnelson9 at gte.net> wrote:

>I don't think Python should be a standard at all, because of
>my experience with it on the Mac. Keep the options open for
>computer programming languages.
>
>Here are two:
>
>Clean at: http://www.cs.kun.nl/~clean/

I went to this website.  It doesn't work for the Mac at all --
Windows only.  So what's your beef with Python on the Mac?  
At least it exists for the Mac.

Also:  "Allthough Clean is not easy to learn (it's very 
different from imperative languages, such as Pascal and C), 
once you can program in Clean, you can easily create games 
with the Clean Game Library."

Do you think this emphasis on game programming is better
than my "math through programming" approach?  It seems to
me that writing games is (a) a rather specialized goal
(vs. learning generic math ideas) and (b) a domain in 
which it's difficult to compete with already-existing 
commercial products (i.e. if your goal is to become a
game programmer, vs. competent in basic math, you have a 
long road ahead of you).

That being said, I'm intrigued by Clean and might tackle
it.  This looks like a more user-friendly environment for
kids than Hugs (a free Haskell).

>Caml at:  http://pauillac.inria.fr/caml/
>
>Both of those languages are OOP too, and easier to use than Python.

What do you base this "easier to use" verdict on?  Personal
experience (see below)?  I've been studying the docs and 
don't, off hand, see anything to suggest that either Caml
or Clean are "easier" than Python.  The websites in question
make no such claims, that I can find.

Your post doesn't make it clear the Caml is actually a 
"dialect of ML".

To me, OCaml looks like a research language for computer 
science majors to study, not a "first language" for 8th 
graders (see my initial post to this thread).  You'd have
to make some pretty persuasive arguments for me to want to
switch gears and do my curriculum writing in such a language.
So far, you've offered none.

My guess is you've done approximately zero using any of
the languages under discussion (Python, Clean, Caml).  If
I'm wrong, I hope you will show me otherwise.  I don't
have a lot of time for completely uninformed discussion.

Kirby



From cfelling at iae.nl  Sun Dec 31 08:23:20 2000
From: cfelling at iae.nl (Carel Fellinger)
Date: 31 Dec 2000 14:23:20 +0100
Subject: os module - reliable?
References: <92m38n$88f$1@nnrp1.deja.com>
Message-ID: <92nc08$3qr$1@animus.fel.iae.nl>

curtissol at my-deja.com wrote:

> if __name__ == "__main__":
>     ShowDirContents(os.getcwd()+"\Doc")

better to use:

    ShowDirContents(os.path.join(os.getcwd(), "Doc"))

as explained just the other day in the thread with the title:

    Newbie question ... reading modules from directories in Windows
-- 
groetjes, carel


From be at et.dtu.dk  Wed Dec 20 15:13:17 2000
From: be at et.dtu.dk (Brian Elmegaard)
Date: Wed, 20 Dec 2000 21:13:17 +0100
Subject: Running pdflatex from Python
Message-ID: <3A4112DD.E8B59F3F@et.dtu.dk>

This is posted to comp.lang.python and miktex at dsts.dk

When I try to run pdflatex (miktex 1.20e) from a Python script with
popen("pdflatex "+filename) under win98 SE I get a fault in kernel32.dll. 

This does not happen for pure latex.

I would like to know if this also happens on other machines or only on mine.



-- 
Brian Elmegaard (be at et.dtu.dk) http://www.et.dtu.dk/energysystems 
Dept. of Energy Engineering, Technical University of Denmark, 
Phone +45 4525 4169  Fax +45 4593 5215
:) http://www.rk-speed.dk http://fiduso.dk http://sunsite.auc.dk/dk-tug


From borrini at libero.it  Thu Dec 21 10:30:30 2000
From: borrini at libero.it (Giulio)
Date: Thu, 21 Dec 2000 16:30:30 +0100
Subject: Python Image Resize Problem
Message-ID: <002b01c06b62$f6be2f40$01c8a8c0@it.tin.it>

> Hi guys,
>
> this is my first message..
> I've a problem with the resize of a jpeg.
> The original image is very big with a resolution of 150 pixel per inch.
> I use this to resize:
> _______________________________________________
> pip = pip.resize((dimenx,dimeny),Image.BICUBIC)
> enhancer = ImageEnhance.Sharpness(pip)
> pip = enhancer.enhance(fact)
> _______________________________________________
>
> The problem is that the resulting image is very bad
> (i can see the pixel or is too blur depending of "fact" ) quality and have
a
> resolution of 72 pixel per inch.
>
> How can i manain the original resolution and have a good quality even on
> image of 1/10 size of the original???
>
> Thanks in advance.
>
> Regards
>
> Giulio Borrini
> Web Programmer
> Netbuilder Snc
>
>
>




From joconnor at cybermesa.com  Sun Dec 17 12:09:59 2000
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Sun, 17 Dec 2000 17:09:59 +0000
Subject: Python for REAL Projects ?
Message-ID: <3A3CF367.40708481@cybermesa.com>

>Not to waste a whole post just to say "absolutely" - I'm
>curious whether this one will go through.

It didn't go through, you might want to try again...

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

"God himself plays the bass strings first when He tunes the soul"



From aleaxit at yahoo.com  Fri Dec 22 07:02:04 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 22 Dec 2000 13:02:04 +0100
Subject: Basic MAPI
References: <nEG06.1008$Qb7.95625@newsb.telia.net>
Message-ID: <91vfsb02cn6@news2.newsguy.com>

"Henrik M?rtensson" <henrik.martensson at frontconsulting.se> wrote in message
news:nEG06.1008$Qb7.95625 at newsb.telia.net...
> Hi
>
> I'm trying to run a basic MAPI initialisation, but I get stuck with the
> dispatcher.
    ...
> >>> s=Dispatch("Mapi.Session")

There is no creatable coclass (that I ever heard of) with progId
'Mapi.Session'.


> Below I have pasted the code I'm running. Is there something that you have
> to do
> to make the MAPI interface available to you? I'm not running this in a MS
> Exchange office environment.

You have to instantiate a coclass that will let you get at a
relevant MAPI object model.  The progId you will use will
depend on what MAPI-supplying product you want to use --
Exchange Server, Outlook, Eudora, Netscape, whatever.  With
Outlook, for example, the Session property of the Outlook
application object (which is a creatable coclass, with a
progId of 'Outlook.Application') might serve your needs.


Alex





From emile at fenx.com  Tue Dec 19 11:40:24 2000
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 19 Dec 2000 08:40:24 -0800
Subject: while true: !!!
References: <mailman.977180341.10206.python-list@python.org> <slrn93t7bv.le.grey@teleute.rpglink.com> <KTz%5.48644$x6.22165257@news2.rdc2.tx.home.com> <91nc7h0s0p@news2.newsguy.com> <mailman.977231769.15324.python-list@python.org> <AwL%5.80470$15.17054718@news1.rdc1.az.home.com>
Message-ID: <3A3F8F78.1040805@fenx.com>


Tim Hochberg wrote:

> "David Porter" <jcm at bigskytel.com> wrote in message
> news:mailman.977231769.15324.python-list at python.org...
> 
>> * Alex Martelli <aleaxit at yahoo.com>:
>> 
>>> "Rainer Deyke" <root at rainerdeyke.com> wrote in message
>>> news:KTz%5.48644$x6.22165257 at news2.rdc2.tx.home.com...

<snip derivation of pre-existing documentation> 

> Coincendentally, this corresponds with the definition in section 5.10 of the
> Python Reference Manual:
> 
> "In the context of Boolean operations, and also when expressions are used by
> control flow statements, the following values are interpreted as false:
> None, numeric zero of all types, empty sequences (strings, tuples and
> lists), and empty mappings (dictionaries). All other values are interpreted
> as true."
> 
> -tim

Where do you get one of these time machines?

I-want-one-ly y'rs,

Emile van Sebille
emile at fenx.com



From sbrunning at bigfoot.com  Tue Dec  5 12:09:53 2000
From: sbrunning at bigfoot.com (Simon B.)
Date: Tue, 05 Dec 2000 17:09:53 GMT
Subject: Turning a flat file into a hierarchical structure
References: <90j5km$9sv$1@nnrp1.deja.com>
Message-ID: <90j7gr$bj6$1@nnrp1.deja.com>

In article <90j5km$9sv$1 at nnrp1.deja.com>,
  Simon B. <sbrunning at bigfoot.com> wrote:

Sod's law, isn't it - you spend hours trying to puzzle something out,
then withing minutes of posting, the answer comes to you. The working
code (along with slightly more complex test data) looks like this:

testdata = ((1, 'egg '), (2, 'bacon'), (2, 'sausage'), (3, 'spam'),
(4, 'tomato'), (2, 'beans'), (2, 'spam'), (3, 'spam'), (2, 'spam'))

mylist = []
appendlist = mylist
lastlevel = testdata[0][0]
parentlist = []

for level, text in testdata:
    if level == lastlevel:
        appendlist.append((level, text, []))
    elif level > lastlevel:
        parentlist.append(appendlist)
        appendlist = appendlist[-1][-1]
        appendlist.append((level, text, []))
    if level < lastlevel:
        while level < appendlist[-1][0]:
            appendlist = parentlist.pop(-1)
        appendlist.append((level, text, []))
    lastlevel = level

print mylist

It goes without saying, though - any pointers from the gurus gratefully
accepted!

--
Simon Brunning
What is a wedding? Webster's Dictionary defines a wedding as 'The
process of removing weeds from one's garden.' - Homer Simpson (giving a
lecture on marriage)


Sent via Deja.com http://www.deja.com/
Before you buy.


From jdries at mail.com  Sun Dec 31 13:02:20 2000
From: jdries at mail.com (Jan Dries)
Date: Sun, 31 Dec 2000 19:02:20 +0100
Subject: How can I conveniently print an expression?
References: <20001230231017.A30045@llnl.gov>
Message-ID: <3A4F74AC.25FB45D8@mail.com>


Issac Trotts wrote:
> 
> In Python, one might naively try something like
> 
> def printexpr(expr_string):
>   expr = eval(expr_string)
>   print expr_string+' = '+`expr`
> 
> which would be used to say things like
> 
> printexpr('len(sys.argv)')
> 
> but of course this does not always work because when Python executes
> eval(expr_string), it assumes that the elements of the expression
> specified by expr_string are to be found in the global scope. 
>
> If there were some mechanism that allowed us to access the variables
> of a calling function from within a called function, then the problem
> would be solved.  Does anyone know of such a mechanism?

I once came up with a solution for this. The following should do the
trick for you:

from sys import exc_info

def caller_symbols():
    try:
        raise StandardError
    except StandardError:
        t = exc_info()[2].tb_frame
        return (t.f_back.f_back.f_globals,t.f_back.f_back.f_locals)

def printexpr(expr_string):
    caller_globals,caller_locals = caller_symbols()
    expr = eval(expr_string,caller_globals,caller_locals)
    print expr_string+' = '+`expr`


Regards,
Jan



From jamarijr at my-deja.com  Thu Dec 14 09:57:54 2000
From: jamarijr at my-deja.com (Arinté)
Date: Thu, 14 Dec 2000 14:57:54 GMT
Subject: MS + Python = help???
Message-ID: <91an5e$qfm$1@nnrp1.deja.com>

What is the address of the webpage that had the python help in the form
of one of the new micrsoft html help format things.

tia


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


From tnt-audio at t-online.de  Sun Dec 10 19:48:10 2000
From: tnt-audio at t-online.de (Thomas Thiele)
Date: Mon, 11 Dec 2000 01:48:10 +0100
Subject: type vs. class
Message-ID: <3A34244A.415F4DFB@t-online.de>

Hallo!

simple typetest

>>> type(3)
<type 'int'>

>>> class X:
	pass
	
>>> x = X()
>>> type(x)
<type 'instance'>

How can I realize that I will not get type 'instance' but type 'X'? I
have different classes, more than one, and I want to check out which
class it is.  Are all classes type instance?


From dale at out-think.NOSPAMco.uk  Mon Dec  4 13:59:01 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Mon, 04 Dec 2000 18:59:01 +0000
Subject: Mutual import?
References: <r16n2tkqepahf288gmu6tku5d2sctb72o0@4ax.com> <90ggf9030hi@news2.newsguy.com>
Message-ID: <c9qn2tc2r7gacaalhqajmk2ivviopiejvq@4ax.com>

Thanks.

I came to the same conclusion and reworked it to avoid the mutual dependence.

It turned out to be a simpler job than I had expected.

(unlike Unicode which is turning into a nightmare!)

--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants




From sorular at netscape.net  Wed Dec  6 00:27:01 2000
From: sorular at netscape.net (sorular)
Date: Wed, 06 Dec 2000 05:27:01 GMT
Subject: Using dictionaries
References: <3a2c9264.503856587@news-server.bigpond.net.au> <90i8a001u10@news2.newsguy.com>
Message-ID: <3a2dca59.583722939@news-server.bigpond.net.au>

>> and in the "main.py" these will be all renamed with the local names
>> ie
>> Val1 = dictionary.dict['key1']
>> Val2 = dictionary.dict['key2']
>> ....
>> ValN = dictionary.dict['keyN']
>> etc....
>
>First, I assume you mean configuration.dict, unless for some
>peculiar reason you've chosen to "import configuration as dictionary"
>(but why would you do that?).

Some silly mistake... as you said it should have been like
	Val1 = configuration.dict['key1']
	etc...

......

>As you can see, Python offers a lot of similar but not
>identical approaches for the task you've set yourself.
>It's not part of the party line, but, in Python, too,
>there's more than one way to do it.  How to choose?  A
>golden rule of programming is...:
>
>    *Do the simplest thing that could possibly work*
.......
>
>    from configuration import *

This is what I decided to do.. it does what I want to do pretty well
without any complication...  nor any dictionaries.!

One slight problem :

Another dictionary look up where necessary values are received from a
single-item dictionary (later on I will open this up to a larger
number of items..!) (I saw a similar one in Mr Guido's site I
think!).....

The dictionary would be {'UserXXXX':'123123'}

----Configuration.py (start)----
...
UserName=UserXXXX
UserNumb="123123"
...
----Configuration.py (ends)----

----Main.py (start)----
...
from configuration import *
class Authenticator:
    def __init__(self):
        try:
	 somethingelse()
        except:

self.authenticate=Dict_lookup(UserName=function(UserNumb))

    def lookup_key (self,*args):
        return apply(self.authenticate.lookup_key,args)

class Dict_lookup:
    def __init__(self,**kw):
        self.dict=kw

    def lookup_key (self,sn):
        return self.dict[sn]
...
----Main.py (ends)----

The whole point is to call up the dictionary lookup for the values in
single item dictionary from configuration.py (which is user entered
parameters) when somethingelse() is not available...

However, python gives me error when I run it like this as :
           Exceptions.KeyError : UserName

but still if I change the above line with :
    def __init__(self):
        self.authenticate=Dict_lookup(UserXXXX=function(UserNumb))

then it works fine...!!!

I should be able to the call UserName (which I tried with single,
double and reverse quotes 
	ie
	'UserName'  or  "UserName" or `UserName`

or is there something I am not seing here??

thanks

HK


From gregj at pobox.com  Tue Dec  5 02:36:09 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Tue, 05 Dec 2000 07:36:09 GMT
Subject: Help with Macintosh files on NT
References: <908oj4$hm1$1@nnrp1.deja.com>
Message-ID: <JV0X5.116163$U46.3745292@news1.sttls1.wa.home.com>

<geos at bigfoot.com> wrote in message news:908oj4$hm1$1 at nnrp1.deja.com...
> I'm getting a bit desperate.  I can't seem to find the answer
> anywhere.  In any case, this is the scenario.  We have an NT machine
> with Macintosh Services turned on.  There are Macintosh files saved on
> the machine with special characters which show up as a bold pipe (|)
> when browsing using explorer (see attached file), but ? when viewed in
> dos.  how can i actually identify those files in python?
>
> b=os.listdir('.')
> >>> print b
> ['Cloud Gallery', 'EC 12?94', 'Federal Clip Art', 'File 10?10?2000
> copy', 'ICON?', 'Network Trash
> Folder', 'TheVolumeSettingsFolder', '\306 A and E together', '\357 i
> with 2 dots', '\325 O with tilde', '\371 u with accent', '\373 u with a
> hat', '\260 degree', '\231 TradeMark', "? You're a star"]
> >>> for path in b:
> ... print os.stat(path)

It looks like the | or ? characters are actually forward or backward
slashes. Both of those characters are legal in Mac filenames, but \ is not
legal in Windows, and / is not legal on Unix. It also looks like some of the
filenames contain non-printable characters and characters in the high ASCII
range, which may be displayed as ? or some other character.

I don't know how you can get around this offhand... maybe fiddling with the
environ settings. It could be a shortcoming in whatever Windows system call
os.listdir is using.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From gregj at pobox.com  Tue Dec 12 02:37:07 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Tue, 12 Dec 2000 07:37:07 GMT
Subject: type vs. class
References: <3A34244A.415F4DFB@t-online.de> <uE_Y5.133498$U46.4497870@news1.sttls1.wa.home.com> <91275n02trc@news2.newsguy.com>
Message-ID: <DAkZ5.138543$U46.4631867@news1.sttls1.wa.home.com>

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

> However, to satisfy Thomas's request as he posed it...: when x is an
> instance of some class, you can learn WHICH class it's an instance
> of, through x's __class__ attribute (you can get the NAME of the
> class as a string, through the class-object's own __name__ attribute):
>
> >>> class X:
> ...   pass
> ...
> >>> x=X()
> >>> x.__class__
> <class __main__.X at 007F76CC>
> >>> x.__class__.__name__
> 'X'
> >>>

Thanks... I didn't know about the __class__ attribute. I've never had a
reason to use it.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From qdlaty at wielun.dhs.org  Wed Dec 27 05:10:39 2000
From: qdlaty at wielun.dhs.org (QdlatY)
Date: Wed, 27 Dec 2000 11:10:39 +0100
Subject: Python vs. PHP (& Java?)
References: <mailman.977897470.4734.python-list@python.org>
Message-ID: <3A49C01F.642A7202@wielun.dhs.org>

Hello!

As I remember python can be scripted into html in shtml pages. I never
tried it but this will propably works.


Yours
QdlatY


From dale at out-think.NOSPAMco.uk  Mon Dec  4 09:12:48 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Mon, 04 Dec 2000 14:12:48 +0000
Subject: Beginers question
References: <_3NW5.44729$nh5.3218436@newsread1.prod.itd.earthlink.net>
Message-ID: <gf9n2tg7faupmsn9g4kkkp14e40dv9gpho@4ax.com>

"robert george" <rgeorgesprint at earthlink.net> wrote:

>I started the beginers tutorial. My operating system is Winme.
>The tutorial says I should get a" C:\WINDOWS> " prompt. But I get
>a"C:\WINDOWS\DESKTOP>" prompt.
>Why is that? Will Python work with Winme?
>Thanks for your help
>An extreme novice.
>Rob
>

Withtout knowing exactly what you're following, I'd stick my neck out and suggest that the path (and
hence the prompt) won't make any difference to your results. If it does however, type:

CD ..

(CD followed by two dots) and you'll get the path suggested in the tutorial.

Good luck.
--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants




From gerhard.nospam at bigfoot.de  Tue Dec 19 22:27:08 2000
From: gerhard.nospam at bigfoot.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Wed, 20 Dec 2000 04:27:08 +0100
Subject: python interface to mysql
References: <mailman.977246402.26310.python-list@python.org>
Message-ID: <3A40270C.762FC6DD@bigfoot.de>

Don't know about MySQLmodule, but IIRC it has not been updated long
since, and Andy Dustman's MySQLdb module can be considered its
successor. It works fine with Python 2.0. The URL is
http://dustman.net/andy/python/MySQLdb/

Gerhard H?ring

anand v wrote:
> 
> Hi,
> 
> I have RedHat linux 6.1 and I have mysql version 3.22.32
> and Python 2.0 and I have downloaded MySQLmodule-1.4 and followed the
> instructions in README file and built the MySQL module inside
> the python and I ran pyhton at the command prompt and I tried
> import MySQL and it worked fine
> 
> But when I tried using through a HTML with a python script
> that did not work??
> 
> Does MySQLmodule-1.4 work with Python2.0???
> 
> Could you please help me what's going on??
-- 
Sorry for the fake email, please use the real one below to reply.
contact: g e r h a r d @ b i g f o o t . d e
web:     http://highqualdev.com


From aleaxit at yahoo.com  Fri Dec 22 03:50:56 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 22 Dec 2000 09:50:56 +0100
Subject: Good introduction to functional programming with Python?
References: <mailman.977465166.28872.python-list@python.org>
Message-ID: <91v4m001una@news2.newsguy.com>

"Tim Peters" <tim.one at home.com> wrote in message
news:mailman.977465166.28872.python-list at python.org...
> [Richard P. Muller]
> > Can someone recommend a decent introduction to functional programming
> > in Python?
>
> I doubt it, but if you want to learn FP I strongly recommend checking out
> the freely available Haskell material:
>
>     http://www.haskell.org/

http://sourceforge.net/projects/xoltar-toolkit/ has Python utilities
for better support for some Haskell-like functional programming in
Python, including the key Haskell semantic idea, lazy evaluation.

Haskell is my favourite FP language, and it does have some nice
surface similarities with Python (use of whitespace, the list
comprehension idea), but it differs deeply from Python in two
key issues (which are NOT really central to FP...):

    a. lazy evaluation is always the default semantics
    b. very strict (and elegant) static typing

Python evaluates eagerly by default (like almost all languages!),
and its typing is always dynamic.

If one wants to broaden one's mind, looking at the world from a
very different angle than Python's, then Haskell makes a LOT of
sense.  A 'triangulation' is best done by choosing two 'base
points' that are not TOO close to each other, and nobody would
accuse Haskell and Python of being too close, given [a] & [b].

Incidentally, while haskell.org is a great site, I would strongly
suggest that books are by far the best way to learn Haskell.

Hudak's "Haskell School of Expression" (Cambridge University Press)
may well wow you if you're into multimedia & such; Thompson's more
traditional "Craft of Functional Programming" 2nd ed (Addison
Wesley) is workmanlike, thorough, and clear.  From there you
can move on to Fethi's and Lapalme's "Algorithms: a Functional
Programming Approach" (Addison-Wesley) and Okasaki's "Purely
Functional Data Structures", and finally HAVE a good grasp of
FP the Haskell way (Okasaki's work actually focuses on ML, with
Haskell equivalents in an appendix, so you'll be getting a small
apercu of non-lazy FP thrown in for free:-).

On the other hand, if one's focus is on understanding how best
to apply FP *in a Python setting*, this FP curriculum may prove
too round-about -- the conceptual detours through static typing
and "laziness as a way of life", in particular, are surely
scenic and mind-broadening, but not specifically relevant.  There
are faster and more direct ways...


> > I haven't found any suitable intro, and I'd like to learn more
> > about map(), apply(), lambda, and the like.
>
> Haskell will teach you that.  You can come back and apply what you learned
> in Python then -- although some people will cluck at you <wink>.
>
> > I understand the syntax, but I'd like to learn something of
> > the functional programming context these commands come from.
>
> If you want to learn French, don't go to Manchester.  Functional
programming
> features are not central to Python -- its support for them is spotty,
clumsy
> and unlikely to get fancier.  Learn FP in a context where it's natural!
And
> Haskell is the most Pythonic of all languages that have nothing in common
> with Python.

There must indeed be some Pythonic 'aura' to Haskell, I guess, given
the somewhat surprising correlation between liking Haskell and liking
Python among people who've tried both.  But points [a] and [b] above
are BIG technical differences, and not centrally relevant to FP.

There ARE functional programming languages with dynamic typing AND
eager (strict) evaluation models like Python.  A 'pure' FP with these
characters is up-and-coming Erlang -- but I would not necessarily
suggest it here, because its central semantic concern is with issues
of parallelism and concurrency, which have relatively little to do
with the issue.  A NON-pure language one SHOULD consider for its
didactical strength is Scheme.  Abelson's and Sussman's "Structure
and Interpretation of Computer Programs" will teach one Scheme in
a strongly FP-oriented setting, and it's a great book on its own --
I consider SICP as one of the main conceptual motivation to learn
a little Scheme, just as Kernighan and Pike's "The Practice of
Programming" is motivation enough to learn some C, or Fowler's
"Refactoring" motivates one to pick up 'just enough Java to get by'.

Scheme's syntax will feel pretty alien, what with all the darned
parentheses, but, with Abelson and Sussman at hand, one soon
gets past that and focuses on really deep and fascinating concerns.

http://www.cs.rice.edu/CS/PLT/packages/drscheme/ gives a free
and very good implementation, with all the trimmings and a lot
of equally-free didactical support, good for any Win32 box, Mac,
or Unixoid.  (Not that Haskell's and Erlang's implementations
are any less free, cross-platform, and good-quality, but, IMHO,
Rice's Scheme has characteristics that make it even better).


Alex





From skidrow at nownuri.net  Mon Dec 11 08:34:40 2000
From: skidrow at nownuri.net (Seo ji won)
Date: Mon, 11 Dec 2000 22:34:40 +0900
Subject: How can I use printer port in python?...
References: <90vnol$l5p$1@news1-2.kornet.net> <3A33C429.FE4EF79C@engcorp.com>
Message-ID: <912l22$gv$1@news1-2.kornet.net>

Oh, I'm using Win2k(server..) and using python 2.0,

and want to use parallel port...

I looked at those archives which you suggested, but there's no direct
answer..

-- or  I  don't know if its answer :)

Do I have to use c code to use parallel port on windows?..

If I have to use c, just give me brief sample for it... ;-)

if not, tell me how can I deal with parallel port with python...

Thanks alot, in advance

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3A33C429.FE4EF79C at engcorp.com...
> "??????" wrote:
> >
> > As the title says, I was wondering if there's anyway I can
> >
> > use printer port in python...
>
> You don't mention which platform nor what you want to do with the
> "printer port", nor even whether the "printer port" you refer to is a
> parallel port, serial port, or some other port.
>
> We could assume you mean a parallel port, and that you are on Windows,
> and that you don't actually plan to talk to a printer here, but then
> again we could be wrong.  Help us out!
>
> On top of which, may I suggest you check the archives for the last week
> and read the discussion which just took place with respect to (what I
> assume is) exactly this topic, under the subject "parallel port access?"




From gbreed at cix.compulink.co.uk  Wed Dec 20 07:58:18 2000
From: gbreed at cix.compulink.co.uk (gbreed at cix.compulink.co.uk)
Date: 20 Dec 2000 12:58:18 GMT
Subject: while true: !!!
References: <t4073164g9shaf@corp.supernews.com>
Message-ID: <91qada$1dh$1@plutonium.compulink.co.uk>

In article <t4073164g9shaf at corp.supernews.com>, 
andrew_dot_henshaw_at_earthling_dot_net (Andrew Henshaw) wrote:

> I had never thought of this before, but it is kind of interesting that
>     not 'this is a true string'
> doesn't produce
>     ''
> but instead produces 0.

That also means that "not not" is a "convert to boolean" equivalent.

> Similarly, not [0, 1, 2] could produce [] .
> 
> In fact, I could see where that might be useful.  On the other hand (to 
> be
> consistent), what would
>     not []
> produce? :)

Where would it be useful?

I suppose if not [] still returned 1, at least not not would always return 
a "boolean".  But it would mean you couldn't guarantee the type of not 
performed on a list.

If you want to empty a sequence, but preserve the type, try filter(lambda 
x:0, asequence).

> Also, shouldn't false == false ?
>     '' == []
> returns 0.
>   :)
> 
> Well, at least
>     (not '') == (not [])
> returns 1.

Yes, that's where the existing behaviour comes in useful.  As Python's 
fairly consistent in not expecting truth values to be of any given type, 
it's not as useful as it could be.  But it's still nice to know it's 
there.


                Graham


From jj at void.si  Fri Dec 29 06:52:58 2000
From: jj at void.si (jj)
Date: Fri, 29 Dec 2000 12:52:58 +0100
Subject: msxml3 and IStream interface
Message-ID: <QV_26.228$Io4.17256@news.siol.net>

Hi,
I want to use MSXML3 library for writing xml in file, but I dont know how to
implement IStream interface in acceptable way. In following source is
stripped Stream class from python examples. Too prove that all other code is
working there is working example of saving in COM structured document file.
I would also appreciate comments about my coding, especially how to get
another interface pointer from object. I found only working solution using
king of import typelib with makepy.py utility.

Regars, JJ

# example ###################################################

import win32com.client, os
import pythoncom, msxml_TLB
#msxml_TLB: makepy.py -o msxml_TLB.py "Microsoft XML, v3.0"

class Stream:
 '''from examples, but it is not working well, btw it is sptripped version
juto to prove bug'''
 _public_methods_ = [ 'Read', 'Write']
 _com_interfaces_ = [ pythoncom.IID_IStream ]

 def Read(self, amount): return 0
 def Write(self, data): return 0

class SaxWriter: #wrapper around xmlwriter
 def __init__(self):
  self.elements = []
  self.wr = win32com.client.Dispatch("Msxml2.MXXMLWriter")
  # !!! reason for importing typelibrary, could be done without importing
typelib !!!!
  self.ch =msxml_TLB.IVBSAXContentHandler(self.wr)
  self.atrs = win32com.client.Dispatch("Msxml2.SAXAttributes")

 def startDocument(self):
  self.ch.startDocument()

 def startElement(self,name):
  self.ch.startElement("", "", name, self.atrs)
  self.atrs.clear()
  self.elements.append(name)

 def addAttr(self, atr, value):
  self.atrs.addAttribute("", "", atr, "", value)

 def endElement(self):
  self.ch.endElement("", "", self.elements.pop())

 def characterData(self, s):
  self.ch.characters(s)

 def endDocument(self):
  self.ch.endDocument()

strg=None
def STRG(fname):
 os.unlink(fname)
 global strg
 strg=pythoncom.StgCreateDocfile(fname, 0x12, 0)
 return strg.CreateStream('a', 0x12, 0, 0)

def test():
 s = SaxWriter()
 st = Stream()  # !!!! its not working
 #st =STRG('c:/tmp/strg.txt')  # this is working
 s.wr.output= st
 s.startDocument()
 s.addAttr("cover", "hard")
 s.startElement("book")
 s.characterData("kedf < sdf > dfgl sdfglj sdlfkgj ")
 s.startElement("book1")
 s.characterData("nested element")
 s.endElement()
 s.endElement()
 s.endDocument()
 #print s.wr.output
 global strg
 del strg

test()




From SBrunning at trisystems.co.uk  Fri Dec  8 06:44:40 2000
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Fri, 8 Dec 2000 11:44:40 -0000 
Subject: Newbie : Tk an data input
Message-ID: <31575A892FF6D1118F5800600846864D5B158C@intrepid>

> From:	Patrick [SMTP:patrick.fichou at galactica.it]
> Sent:	Friday, December 08, 2000 11:03 AM
> Is someone kind enough to send me an example of a program making a data
> input into a file (an MSAccess Data Base would be great).
 
I would suggest that you use mxODBC - you can get it from
<http://www.lemburg.com/files/python/>. Then DB is as simple as:

myConn = ODBC.Windows.Connect('myDNS')
myCursor = myConn.Cursor()
myCursor.execute("insert into SPAM values('%s', '%s')" % (egg, sausage))
myConn.commit()

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From aleaxit at yahoo.com  Sat Dec 16 15:47:02 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 16 Dec 2000 21:47:02 +0100
Subject: Singleton vs Proxies DP (was Re: Solution: Direct access to Printer I/O lines)
References: <917k9b$9i9$1@nnrp1.deja.com> <slrn93hrjn.2vp.ssthapa@localhost.localdomain> <91cnp3$g59$1@nnrp1.deja.com> <91d23f0jk7@news1.newsguy.com> <91gfnp$7r4$1@animus.fel.iae.nl>
Message-ID: <91gkh801e49@news2.newsguy.com>

"Carel Fellinger" <cfelling at iae.nl> wrote in message
news:91gfnp$7r4$1 at animus.fel.iae.nl...
> Alex Martelli <aleaxit at yahoo.com> wrote:
> ...
> > This gives me a good chance for one of my favourite rants,
> > '"Singleton" design-pattern considered harmful'!-)  But
> > 'Singleton' is not being ideally applied, so I'll skip that.
>
> And in an other thread you wet my appetite and direct me here!
> So let us hear that rant, mister, and enlight us mere mortals
> who still think that there is legitimate use of this singleton
> pattern.

It's not rare to desire that a certain resource (e.g., a RDBMS
running on another process/machine) be handled through one,
and one only, 'bottleneck'.  Singleton does ensure that, yes.
There is a space for singletons.  But the Singleton _Design
Pattern_ need not be the best way to have singletons.

Within the component that implements the singleton, it
does not matter how and where you place the state, and
the code that knows about that state's representation and
accesses and modifies it directly.  You may choose to
have that state as instance-variables of an object, that
code in its methods, and ensure a single instantiation of
that object, but that's a somewhat roundabout way for
languages which allow per-class (C++) or per-module
(Python) functions and data.  But it's a minor decision, of
implementation more than of design; what matters is the
interface you choose to expose from that component.

And _should_ that interface be one of a single object that
gets instantiated once, so it must be accessed through
some single 'give me a reference to Mr. Singleton please'
function?  I do not see why that could ever be the ideal
interface.  There are all sorts of issues that give slight
problems, and I don't see any compensating benefit when
compared with the following approach.

Why not have client-code ask for objects normally,
use them, then drop them when it's done?  If client
code asks for 'another object' -- why not hand it out?

It doesn't matter if the 'handed out' objects have
different *identities*... as long as they share *state
and behavior*, no?  I forget on what name this simple
pattern was originally published in "C++ Report",
but let me call it 'lightweight proxies' (LWP, versus S
for the singleton pattern).  It's just one of many, many
examples of the benefits that can often be had by
splitting notions of id and state in OO programming.

An example of benefit in Python: say that LWP is the
class in the library module that implements the
lightweight proxies pattern.  It exposes functionality
and state for some underlying in-one-instance-only
object S.  That functionality may not be in the best
shape for client-code, that may want to adapt it --
maybe add or tweak some methods from the rest of
the client code in a 'middleware' stratum.

class DecoratedLWP(LWP):
    def anewmethod(self,x):
        # translate argument and result inches<->mm
        return self.anoldmethod(x*25.4)/25.4
    def anoverride(self,y,z):
        # make sure all calls are logged
        result = LWP.anoverride(self,y)
        z.logit(y,self,result)
        return result

No problem inheriting from LWP -- instantiating
a new DecoratedLWP will instantiate an LWP, but
that's all right, you can instantiate any number of
LWP's you want -- they all internally refer to S,
but who cares.  Inheriting-from-singleton isn't
that easy -- and if two different client modules
want to inherit-from-singleton differently, it's a
big conflict.  So any 'decoration' of singleton must
come through less-convenient encapsulation and
delegation.

A different advantage in C++ is that client code
can instantiate LWP instance with normal operator
'new' (less of an issue in Python, since classes and
functions are both callable in similar ways) and
when done dispose of them with normal 'delete'
(no issue in Python, since references-to-S might
be dropped normally with delete; an issue in C++,
where 'dropping refs to the singleton' would need
to invent one more arbitrary protocol).

A different advantage in COM is that singleton does
NOT work there... not generally.  in-process objects
cannot be system-wide singletons, for example --
not via normal generation methods; their instancing,
exposing and handing out of references would have
to be coordinated from other 'manager' objects. LWP
objects can be in-process or whatever, and handle
everything they may need to coordinate with any
underlying object 'behind the scenes'.


This may be framed as handing out conveniences,
and/or equivalently removing hassles, from the client
code side; basically at minor levels, most of the time,
but these things mount up.  Some of the hassles do
disappear, others may just be moved to server code,
but that's still much better than leaving client code
to deal with them.  It seems to me that Singleton
takes little or no account of the 'library code' vs 'client
code' split.

'Evolvability' of LWP vs S is a particularly strong
point.  If at some point client-code needs to start
receiving references to _more_ than one library-side
object, the passage is much smoother if client code
already expected to receive multiple objects (which
possibly share state) rather than references to a
single one (1 identity/many states is much trickier
than many identity/1 state...:-).


Alex





From aleaxit at yahoo.com  Wed Dec 20 06:22:16 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 20 Dec 2000 12:22:16 +0100
Subject: while (thread.alive()):
References: <mailman.977283904.26696.python-list@python.org>
Message-ID: <91q4pv01851@news2.newsguy.com>

"Jay O'Connor" <joconnor at cybermesa.com> wrote in message
news:mailman.977283904.26696.python-list at python.org...
    [snip]
> Believe it or not, both Ada and Smalltalk have boolean types.

So does Fortran.  It calls them 'LOGICAL'.  So?

> Both languages were designed in the 70s, commercially implemented in the
> 80s; one is dynamically typed, one is statically typed.

And Fortran was first designed and implemented in the '50s (although
I'm not sure if the 1954 version already had LOGICAL).  So?

> Why are we still maintaining these leftover idioms such as using
> integers as booleans.  If we are to carry the state of the art of
> programming beyond it's current level, we need to be programming a t a
> higher level of abstraction than is easily supported while still
> dragging anachronisms such as this around.

The idea of having LOGICAL (or Boolean, whatever) as a separate
type is far older -- as old as the oldest surviving high-level
language.  C, which introduced the simplification of NOT having
a separate boolean type although generally moving towards precise
typing (which its inspiration-language, BCPL, didn't), came much
later.  So how can you possibly claim the _newer_ notion is an
"anachronism"?!  If 'when was it invented' had anything to do
with the issue (and it's, IMHO, quite silly to claim it does),
then the 'anachronism' would have to be the older notion of
having LOGICAL as a separate type.

As for abstraction,
    if anyobject:
        # whatever
is quite high-abstraction enough for me: whatever 'anyobject'
is referring to will define its own interpretation of 'being
false' -- polymorphically and in a perfectly abstract way.


> Those who claim that 'while 1:' is obvious, it's not.

Neither is 'X=X+1' *obvious* _at all_, to somebody who has
never programmed; I remember, on my first encounter with
programming, that it was the thing that puzzled me most --
"no, X does *NOT* equal X+1, what's this silliness?!".  That
doesn't necessarily mean Python should adopt <- to indicate
'assignment' (although, except for the backwards compatibility
issue, and the parsing problems betwen 'x < (-2)' and assignment
of 2 to x, I wouldn't mind it -- ':=', now, THAT would be
a bad choice indeed...:-).  The very specialized meanings of
'=' in some languages (Fortran, PL/I, C, C++, Java, Python,
Basic, &c) become reasonably clear with a *very* modest
amount of study -- if only all programming issues were
as easy to grasp as such surface-syntax-sugar ones!-)


> It's only obvious if you've programmed "C" and other such languages.
> It's makes no sense really, unless you've learned the 'code

And you claim that
    while 'true':
would be enormously clearer to a beginner, without study, than
    while 1:
...?  Ridiculous claim, but, go ahead -- use "while 'true':" and be
happy, Python won't complain in the least.


> My first professional language was Ada83.  From there I went to
> Smalltalk.  When I finally started working with languages such as C,
> C++, etc... I was at first confused ( while (1)...what does that mean)
> and then revolted.
>
> I'm continuously amused that people try to take language design forward
> and still drag obsolete ideas along for the ride.
>
> Python's got a lot of cool stuff, but the lack of a real Boolean object
> is a glaring omission

And once again, you have completely failed to make any case for
it.  Your very personal aesthetical preferences, as shaped by
Ada or Smalltalk, are no 'case' at all, of course; just *show*
us what a Python-plus-boolean-type would enable, as the payback
of the needed additions to the language, that current Python
doesn't.  Vacuous claims of 'obsolete' are made particularly
ridiculous by your apparent (and erroneous) beliefs regarding
dates of introduction in typed-languages of 'logical as a
separate type' vs removal of this type-distinction -- as if
the 'dates of invention' mattered at all, of course.

*Entities are not to be multiplied needlessly".  So, show the
NEED -- the NECESSITY for this additional *entity*; the actual,
practical benefits it would bring.  (If you want *several*
additional entities, be sure to assert and justify each of
them).

Be sure to state CLEARLY (that would be the day...) which
proposals are meant for Python-as-it-stands (no introduction
of backwards incompatibilities, as, for example, additional
reserved-words would imply), and which ones for Py3K-of-
the-future (incompatibilities admitted, it the resulting
benefits are really really big)...


Alex





From aahz at panix.com  Wed Dec 13 11:46:20 2000
From: aahz at panix.com (Aahz Maruch)
Date: 13 Dec 2000 08:46:20 -0800
Subject: Moving from Perl to Python
References: <3a3a9552.76673185@news.tcn.net>
Message-ID: <91894s$lms$1@panix6.panix.com>

In article <3a3a9552.76673185 at news.tcn.net>,
Richard Zilavec <rzilavec at tcn.net> wrote:
>
>man python on my machine doesn't work, I'm assuming something is wrong
>or maybe its a perldoc equivalent?  Where can I find the FAQs?

Go to http://www.python.org/ where you'll find a wealth of
documentation.  Most Python installs contain that version's docs in a
Doc/ directory.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

What if there were no rhetorical questions?  --Aahz


From jon at bezek.dstc.monash.edu.au  Sun Dec 10 23:18:37 2000
From: jon at bezek.dstc.monash.edu.au (Jonathan Giddy)
Date: 11 Dec 2000 15:18:37 +1100
Subject: Backslash escape in regular expressions
Message-ID: <jon.976507791@bezek.dstc.monash.edu.au>

Hi,

According to the re module documentation, backslash either escapes special
characters, or signals a special sequence.  The special sequences are
then listed.

However, as this code shows, there are some special sequences (mainly
the whitespace characters) that are special, but aren't listed.  Is this 
a lapse in the re implementation or the re documentation?  Can I safely
expect re.compile(r'\(hello\)\n') to always match '(hello)\n' (the current
behaviour) and not match '(hello)n' (the documented behaviour?)

Thanks,
	Jon.


import re, traceback

for c in 'abcdefghijklmnopqrstuvwxyz':
    print c, '-',
    try:
	if c in 'AbBdDsSwWZ':
	    print 'documented escape'
	else:
	    r = re.compile('\\' + c)
	    if r.match(c):
		print 'documented ordinary'
	    else:
		escape = eval('"\\%s"' % c)
		if r.match(escape):
		    print 'undocumented escape "\\%s"' % c
		else:
		    print '??'
    except:
	print
	traceback.print_exc()


From eliz at is.elta.co.il  Sun Dec 24 01:55:43 2000
From: eliz at is.elta.co.il (Eli Zaretskii)
Date: Sun, 24 Dec 2000 08:55:43 +0200
Subject: Problem with python mode
References: <3a44bcf1.73260781@nntp.interaccess.com> <3A44BF14.F7721A06@is.elta.co.il> <3a464e89.3641015@nntp.interaccess.com>
Message-ID: <3A459DEF.D182406D@is.elta.co.il>

Thaddeus L. Olczyk wrote:
> 
> It comes after the line:
> (if (featurep 'xpm)
> 
> so I suspect that xpm is not supposed to be present in NTEmacs.
> The question then is why does it test as being present?

It shouldn't.  Please recheck, and if "M-: (featurep 'xpm) RET" prints
anything but "nil", please see where does it come from.  Perhaps your
.emacs loads some package which provides xpm.


From DavidL at nonospamraqia.com  Mon Dec 11 16:59:26 2000
From: DavidL at nonospamraqia.com (David Lees)
Date: Mon, 11 Dec 2000 16:59:26 -0500
Subject: How to start a process?
Message-ID: <3A354E3E.1545D045@nonospamraqia.com>

I would like to start up a number of simultaneos processes from a Python
script.  Can someone explain or refer me to a document that explains
what is going on below.  As a test I am executing a script 'hack.py'. 
Why do I get this OSError when I try using execv, but it runs when I use
'execfile'?

Thanks in advance.

david lees


>>> execfile('hack.py')
foo to you charlie
more hacks
>>> os.execv('e:\python20\hack.py',['foo','bar'])
Traceback (innermost last):
File "<interactive input>", line 1, in ?
OSError: [Errno 8] Exec format error
>>>


From sandj.williams at gte.net  Tue Dec 12 13:16:25 2000
From: sandj.williams at gte.net (Steve Williams)
Date: Tue, 12 Dec 2000 18:16:25 GMT
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <915jcf0c6o@news2.newsguy.com>
Message-ID: <3A366C90.7704333F@gte.net>

Alex Martelli wrote:

> "Steve" <smnordby at yahoo.com> wrote in message
> news:3A36368C.26AADDDE at yahoo.com...
> > I dislike infinite while loops and use:
> >
> > for line in file.readlines()
> >     ...do stuff...
> >
> > No need for a break since it will automagically when it runs out of
> > lines.
>
> Surely best, *IF* you're 100% sure that all the files you
> ever read will always fit comfortably in memory.
>
> If you must provide for reading huge files, this will just
> not work.
>

[snip]

I think we're building a strawman here--we all know readlines() holds a
trap for the unwary.

The 'delightful' example would be

    While 1:
        recordSet = readline(intComfortableNumber)
        if not recordSet:
            break
        for record in recordSet:
            do your thing

The 'convenient' example would be:

    recordSet = readline(intComfortableNumber)
     while recordSet:
          for record in recordSet:
              do your thing
          recordSet = readline(intComfortableNumber)

Perhaps



From kirschh at lionbioscience.com  Wed Dec 13 02:44:08 2000
From: kirschh at lionbioscience.com (Harald Kirsch)
Date: 13 Dec 2000 08:44:08 +0100
Subject: Q: looping with two counter variables through a list
Message-ID: <yv2zoi0ojef.fsf@lionsp093.lion-ag.de>

In Tcl I can write

  set mylist {0 1 2 3 4 5}
  foreach {a b} $mylist {
    puts $a
  }

to print the even numbers of $mylist. Obviously

   mylist = [0, 1, 2, 3, 4, 5]
   for a,b in mylist:
     print a

does not work (ValueError: unpack sequence of wrong size). 

What is Python's idiom for running with more than one variable through
a flat list?

  Harald Kirsch

-- 
----------------+------------------------------------------------------
Harald Kirsch   | kirschh at lionbioscience.com | "How old is the epsilon?"
LION Bioscience | +49 6221 4038 172          |        -- Paul Erd?s
       *** Please do not send me copies of your posts. ***


From fredrik at effbot.org  Tue Dec 19 11:58:43 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 19 Dec 2000 16:58:43 GMT
Subject: tkinter event binding to mouse double click
References: <91n7l6$7ij$1@news.nuri.net> <vpL%5.459$AH6.36619@newsc.telia.net> <91o31f$vi$1@news.nuri.net> <4iM%5.468$AH6.37352@newsc.telia.net>
Message-ID: <7tM%5.471$AH6.37504@newsc.telia.net>

I wrote:
> it may be a problem with Tk on your platform

fwiw, it's a known problem in 8.2.1:

    http://www.deja.com/getdoc.xp?AN=612566673

however, I cannot find it in the current bug database:

    http://sourceforge.net/bugs/index.php?group_id=12997

so maybe it's been fixed in 8.3 or later?  (if not, maybe it's
time to resubmit the bug report...)

</F>




From aleaxit at yahoo.com  Tue Dec 12 18:28:33 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 13 Dec 2000 00:28:33 +0100
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <915jcf0c6o@news2.newsguy.com> <9162p3$7t2$1@animus.fel.iae.nl>
Message-ID: <916cd31b3b@news1.newsguy.com>

"Carel Fellinger" <cfelling at iae.nl> wrote in message
news:9162p3$7t2$1 at animus.fel.iae.nl...
> Alex Martelli <aleaxit at yahoo.com> wrote:
>
> > Firmitas, Utilitas, Venustas -- in *THIS* order, as Vitruvius
>
> No no, Beauty is the one and true guiding principle, Beauty and
Simplicity.

...the TWO and true guiding principles are Beauty and Simplicity and
Solidity -- no, wait: among our guiding principles are such things as...


Alex





From tuttledon at hotmail.com  Mon Dec  4 09:58:13 2000
From: tuttledon at hotmail.com (Don Tuttle)
Date: Mon, 04 Dec 2000 14:58:13 GMT
Subject: Programming with Python
References: <90fq39$jgq$1@nnrp1.deja.com>
Message-ID: <9iOW5.55847$L26.16633720@typhoon.southeast.rr.com>

"Simon B." <sbrunning at bigfoot.com> wrote in message
news:90fq39$jgq$1 at nnrp1.deja.com...
> I came across a Python book this weekend that I haven't heard of
> before - Programming with Python, by Tim Altom
> (see <http://www.prima-tech.com/book.asp?id=3942>). Anyone have this?
> Any good?

IMO, this is the only bad Python book on the market.  Don't waste your
money.

Don




From sbrunning at bigfoot.com  Tue Dec  5 06:38:23 2000
From: sbrunning at bigfoot.com (Simon B.)
Date: Tue, 05 Dec 2000 11:38:23 GMT
Subject: Programming with Python
References: <90fq39$jgq$1@nnrp1.deja.com> <9iOW5.55847$L26.16633720@typhoon.southeast.rr.com>
Message-ID: <90ik3f$rhp$1@nnrp1.deja.com>

In article <9iOW5.55847$L26.16633720 at typhoon.southeast.rr.com>,
  "Don Tuttle" <tuttledon at hotmail.com> wrote:
> IMO, this is the only bad Python book on the market.  Don't waste your
> money.

Pity, 'cos dictEdit.py sound like *exactly* what I need. But if it's no
good...

--
Simon Brunning
As far as the laws of mathematics refer to reality, they are not
certain; and as far as they are certain, they do not refer to reality. -
 Albert Einstein


Sent via Deja.com http://www.deja.com/
Before you buy.


From dsh8290 at rit.edu  Thu Dec 14 08:35:37 2000
From: dsh8290 at rit.edu (D-Man)
Date: Thu, 14 Dec 2000 08:35:37 -0500
Subject: Moving from Perl to Python
In-Reply-To: <3a3a9552.76673185@news.tcn.net>; from rzilavec@tcn.net on Wed, Dec 13, 2000 at 10:40:04 -0500
References: <3a3a9552.76673185@news.tcn.net>
Message-ID: <20001214083537.C21300@westsidecnc.rh.rit.edu>

Just another opinion : I start with python by reading the tutorial on
www.python.org,  then when I need / want  a library function I just browse the
library reference.  It's great way to save money!

BTW, that was as a second year student in college and I never programmed before
my freshman year

:-)
-D

On Wed, 13 Dec 2000 10:40:04 Richard Zilavec wrote:
 | 
 | I just got back from two book stores, nothing on Python, I was
 | shocked!  So I went online and found a pile but I hate ordering
 | online, I like to browse through them first.  Anyways, I really like
 | O'Reilly books, is the Programming Python really advanced or is more
 | along the lines of the Programming Perl where it covers the basics to
 | advanced?
 | 
 | man python on my machine doesn't work, I'm assuming something is wrong
 | or maybe its a perldoc equivalent?  Where can I find the FAQs?
 | 
 | Thanks in advanced for any help or suggestions.
 | 
 | --
 |  Richard Zilavec
 |  rzilavec at tcn.net
 | -- 
 | http://www.python.org/mailman/listinfo/python-list
 | 




From bvdpoel at uniserve.com  Tue Dec 19 22:54:55 2000
From: bvdpoel at uniserve.com (Bob van der Poel)
Date: Tue, 19 Dec 2000 20:54:55 -0700
Subject: length of wav files
Message-ID: <3A402D8E.81978F3B@uniserve.com>

	
I am looking at the docs for the wave library and can't seem to find the
one bit of information I want to extract: the running time of a wav
file. So, before I have to relearn my math, has someone some code which
will deterime the time in seconds of a wav sound file?

Thanks.

-- 
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bvdpoel at uniserve.com
WWW:   http://users.uniserve.com/~bvdpoel



From phrxy at csv.warwick.ac.uk  Sun Dec 10 17:14:48 2000
From: phrxy at csv.warwick.ac.uk (John J. Lee)
Date: Sun, 10 Dec 2000 22:14:48 +0000
Subject: C API mismatch with 2.0
Message-ID: <Pine.SOL.4.30.0012102209040.29674-100000@mimosa.csv.warwick.ac.uk>

I compiled Michael Miller's _tableio.c (the C extension bit of TableIO.py)
on windows using cygwin (with Robert Kern's help), and got the following
warning about the C API.  I used the python20.lib from the same
distribution of python 2.0 that I'm trying to import TableIO into.

Can somebody tell me what's going on?  Where did this API version number
come from for _tableio?

Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
>>> import TableIO
WARNING: Python C API version mismatch for module _tableio:
  This Python has API version 1009, module _tableio has version 1007.
>>>

Thanks for any help.


John




From ngps at madcap.dyndns.org  Mon Dec 11 08:46:43 2000
From: ngps at madcap.dyndns.org (Ng Pheng Siong)
Date: 11 Dec 2000 13:46:43 GMT
Subject: corba, soap, com, etc...
References: <pKsY5.58762$yR4.1496678@news1.rdc1.tx.home.com> <3QzY5.9825$dg6.637046@pouncer.easynews.com>
Message-ID: <912ls3$bn8$1@violet.singnet.com.sg>

According to Jiva DeVoe  <jiva at no-spam.com>:
> Check out fnorb.  http://www.fnorb.org  It's an excellent open
> source CORBA 
> implementation for python.

Just downloaded Fnorb. According to the docu, it supports
BOA. BOA has been superceded by POA. 

See the CORBA FAQ here:

    http://www.javacommerce.com/CorbaFaq/toc-exhaustive.html


OmniOrb3 supports POA:

    www.uk.research.att.com/omniORB/omniORBpy/



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



From fgleeson at my-deja.com  Sat Dec 30 18:42:08 2000
From: fgleeson at my-deja.com (fgleeson at my-deja.com)
Date: Sat, 30 Dec 2000 23:42:08 GMT
Subject: Komodo
References: <B672A66F.72CC%bobhicks@adelphia.net> <3A4D6D0B.20000@ActiveState.com>
Message-ID: <92lrsg$2pa$1@nnrp1.deja.com>


> * For those who can't infer the obvious, I'm part of the Komodo dev team :-]
>
>

I know this is off topic, but since you are on the Komodo dev team I can not
resist. Komodo looks like a really nice product, which I would like to use.
However it seems to be missing some essential features for a perl/python IDE.

For instance typically in perl\python scripts you need to pass in some
parameters to control the scripts behavior. If I am running from the command
line its simple perl myscript.pl opt1 opt2 etc..

I can not find a way to do this. Launching komodow c:\scripts\myscript.pl
opt1 Komodo can not find my script even if I cd to the directory containing
my script.

Am I missing something? Is Komodo only meant to be used for cgi and therefore
does not allow parameter passing? By the way what is the difference between
komodo.exe and komodow.exe?


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


From dieter at handshake.de  Sun Dec  3 16:32:01 2000
From: dieter at handshake.de (Dieter Maurer)
Date: 03 Dec 2000 22:32:01 +0100
Subject: [Bug] Python2.0: Unicode/string.letters inconsistency
Message-ID: <x7u28l6vji.fsf@lindm.dm>

I have a problem with a Unicode/string.letters inconsistency:

  During "xslt" processing, a unicode string is passed to
  "urlpath.urlpath". This fails with the famous

     UnicodeError: ASCII decoding error: ordinal not in range(128)

  The reason:
  
    "string.letters" contains not only the ASCII letters but
    all "ISO-8859-1" letters.

    "string.letters" is used in "urlparse" to build
    "scheme_chars" against with the URL characters are
    compared. "scheme_chars" cannot be coerced in a Unicode
    string, as it contains non ASCII characters -> the exception.

In my view, it is inconsistent to disallow default unicode
encoding/decoding for non-ASCII characters but place
non-ASCII characters in such a central place as "string.letters".


Dieter
   



From noahspurrier at my-deja.com  Wed Dec 27 23:12:43 2000
From: noahspurrier at my-deja.com (noahspurrier at my-deja.com)
Date: Thu, 28 Dec 2000 04:12:43 GMT
Subject: Does Python ftplib support APPEND?
References: <92edhe$g7f$1@nnrp1.deja.com>
Message-ID: <92eejo$h36$1@nnrp1.deja.com>

As usual, I figured this out almost immediately after I asked...

import ftplib
ftp = ftplib.FTP('localhost')
ftp.login ('user','password')
fin = open ('foo.txt', 'r')
ftp.storbinary ('APPE foo2.txt', fin, 1)

I think that works...
Yours,
Noah


In article <92edhe$g7f$1 at nnrp1.deja.com>,
  noahspurrier at my-deja.com wrote:
>
>
> As far as I can tell Python ftplib does not
> support ftp append (from RFC959). Is that true? Is there
> any way around this for me? We need to use append.
>
> We're using Python 2.0
>
> Yours,
> Noah
>
> Sent via Deja.com
> http://www.deja.com/
>


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


From owen at astroNOJNK.washington.edu.invalid  Mon Dec 11 11:45:52 2000
From: owen at astroNOJNK.washington.edu.invalid (Russell E. Owen)
Date: Mon, 11 Dec 2000 08:45:52 -0800
Subject: Tkinter prob in Python 2.0, Macintosh
References: <muvol.remove-this-ABD013.16350709122000@news.knology.net>
Message-ID: <9130bq$cnsu$1@nntp6.u.washington.edu>

In article <muvol.remove-this-ABD013.16350709122000 at news.knology.net>, 
Mike <muvol.remove-this at knology.net> wrote:

>...The problem occurs when I press the Quit key.  The Tk window goes away, 
>but 
>I do not get the >>> prompt back in the interpreter.  Even though I can 
>still select the menus, choosing Quit from the file menu has no effect.  
>Also, Command-Q does not work.  At this point, the About box has changed 
>from "About Python..." to "About Tcl & Tk"
>
>I have to force quit the Python Interpreter to regain control.

Try de-selecting the startup option "Enter interactive mode after 
script". This problem was present in 1.5.2c and based on your 
description it may well still be present in 2.0. (For a one-time test, 
run the interpreter with the option key down. For a longer-term 
solution, run EditPythonPrefs.)

Warning: Tk is not fully functional in MacPython 2.0 -- file events are 
broken. Stick to MacPython 1.5.2 for now if you need file events.

-- Russell


From moshez at zadka.site.co.il  Fri Dec 29 09:54:56 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Fri, 29 Dec 2000 16:54:56 +0200 (IST)
Subject: how do i check if a file exists with python?
In-Reply-To: <fBN26.99045$yR4.2640173@news1.rdc1.tx.home.com>
References: <fBN26.99045$yR4.2640173@news1.rdc1.tx.home.com>, <RlM26.98766$yR4.2639444@news1.rdc1.tx.home.com> <92g5v8$r5k$1@nnrp1.deja.com>
Message-ID: <20001229145456.070CCA84F@darjeeling.zadka.site.co.il>

[Scott Hathaway, on how to find a file]
> Thanks for both responses!

Unfortunately, I think they did you more harm then good by telling you
the answer: I doubt you need it. Now, it may be that I'm wrong, in which
case just go ahead an use os.path.isfile, but if you're doing something
like

def foo():
	if os.path.isfile("foo"):
		return open("foo")
	else:
		return None

Or something to that effect, *don't*. It's evil, and will come back
to bite you (unreadable existing files, files changing between isfile
and open, etc.).

Instead, simple do:

def foo():
	try:
		return open("foo")
	except IOError:
		return None

In other words, don't try to avoid exceptions -- know where you want
to deal with them, and do so.
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From removethis_kranki2 at earthlink.net  Fri Dec 22 11:56:52 2000
From: removethis_kranki2 at earthlink.net (Bob W.)
Date: Fri, 22 Dec 2000 16:56:52 GMT
Subject: Newbie Question
References: <05m64tkqev0aavea80ttjse1sj2gd447hg@4ax.com> <hqs64tc5jjndav3p6nvmu67t7hqaf5go00@4ax.com> <91vuj5$5n7is$1@ID-11957.news.dfncis.de> <42074t8a8e8604sgl77egadao721tk62ik@4ax.com> <92005r$5e67v$1@ID-11957.news.dfncis.de> <9200gh$5o0sr$1@ID-11957.news.dfncis.de>
Message-ID: <ha174t4g5r5bfp766j13bo1i34nbq0jvq8@4ax.com>

Well, I got it to work.  I added an Arg parameter just after self in
the definition for __CheckFile.

    def __CheckFile( self, Arg, szDirName, szFileNames ):

That seems to let it work.  Apparently, the first self reference is
automatically added in the call which I had not expected.  Thanks a
lot for the help.  Now I just have to figure out why this works this
way.  

Crashed PythonWin too!  Guess that I will get better with time and
experience!  Maybe! :))

On Fri, 22 Dec 2000 08:46:42 -0800, "Emile van Sebille"
<emile at fenx.com> wrote:

>BTW, it also works with
>
>    def run(self):
>        os.path.walk(self.root, Class.func, self)



From wware at world.std.com  Thu Dec 28 19:00:12 2000
From: wware at world.std.com (Will Ware)
Date: Fri, 29 Dec 2000 00:00:12 GMT
Subject: Erlang (was Re: Good introduction to functional programming with Python?)
References: <mailman.977984412.30211.python-list@python.org> <92f0il01a5u@news1.newsguy.com> <86wvckqw4h.fsf@localhost.my.domain>
Message-ID: <G6AyoC.CK1@world.std.com>

Brad Knotwell (knotwell at ix.netcom.com) wrote:
> I'm now wondering if Stackless Python's uthreads are cooperatively or
> pre-emptively multi-tasked.

Unless you're digging into their inner workings, microthreads will appear
to you as preemptively multitasking.
-- 
import string,time,os;print string.join((lambda x:x[:10]+x[8:])(map(
lambda x:string.center("*"*(lambda x:((x<24) ### Seasons Greetings, Will Ware
*(x-3))+3)(x),24),range(1,28, 2))),"\n") ################ wware at world.std.com


From tgagne at ix.netcom.com  Wed Dec 20 22:58:52 2000
From: tgagne at ix.netcom.com (Thomas Gagne)
Date: Wed, 20 Dec 2000 22:58:52 -0500
Subject: nevermind
References: <3A40EEBF.3262040F@ix.netcom.com> <ulmtahl1h.fsf@ctwd0143.fitlinxx.com> <3A417C3B.E226C419@ix.netcom.com>
Message-ID: <3A417FFC.56DC8C40@ix.netcom.com>

The str() built-in function did a wonderful job.  I hope you didn't waste too
much time on this.

--
.tom





From fgeiger at datec.at  Tue Dec 26 10:01:07 2000
From: fgeiger at datec.at (Franz GEIGER)
Date: Tue, 26 Dec 2000 16:01:07 +0100
Subject: A Q concerning map (a comparison to Maple V)
References: <mailman.977595008.22017.python-list@python.org>
Message-ID: <92c41e$n1p$2@newsreaderg1.core.theplanet.net>

> Note that the listcomps in 2.0 make scalar broadcast much easier to spell;
> e.g.
>
>     [replace(x, "c", "C") for x in ("abc", "cde")]
>     [x**2 + 5 for x in L]
>

Indeed, Peter, this is a Python way to do it. Yet I'm not very used to use
list comprehensions, but I'm working on that. Thank you!

Best regards
Franz GEIGER


"Tim Peters" <tim.one at home.com> wrote in message
news:mailman.977595008.22017.python-list at python.org...
> [Franz GEIGER]
> > > map(replace, ["abc", "cde"], "c", "C");
> > > instead of having to create a lambda function.
> > > Wouldn't it make sense to have that in Python too? Do there come
> > > more such opportunities into your mind?
>
> [Peter Schneider-Kamp]
> > If I get you right you intend this to mean:
> >
> >   map(replace, ["abc", "cde"], ["c"]*2, ["C"]*2)
> >
> > So basically your proposal is to reuse map parameters
> > which are no sequences in every iteration over the
> > parameters who are,
>
> It's hard to tell for sure, but that was my guess too.  In array
languages,
> this kind of thing is ubiquitous and is often called "scalar broadcast"
> (where a scalar is any atomic value (as opposed to an array), and is
> "broadcast" to every position of the arrays in which it's combined via
some
> operation).
>
> There was a long debate about this a few years ago on c.l.py, where I
> championed scalar broadcast in map specifically.  It collapsed under its
own
> hideous weight when-- as such things always do <wink> --that originally
> modest goal got hijacked by people seeking to extend the semantics to
every
> corner of the language.  That won't happen.
>
> > ...
> > But how would you decide if "abc" is a sequence parameter
> > or not?
>
> That is a problem!  Kinda.  Mostly people want to broadcast numbers (as in
> Franz's original example), and there's no problem there.  A cheap
workaround
> for sequences that are desired to be treated as scalars would have been to
> introduce a new builtin scalar() function, that simply hid the
> "sequenceness" of its argument from map.
>
> Note that the listcomps in 2.0 make scalar broadcast much easier to spell;
> e.g.
>
>     [replace(x, "c", "C") for x in ("abc", "cde")]
>     [x**2 + 5 for x in L]
>
> So the better solution is not to extend map, but to forget it <0.9 wink>.
>
> a-life-strategy-of-universal-applicability-ly y'rs  - tim
>
>






From mal at bewoner.dma.be  Sun Dec 24 16:16:34 2000
From: mal at bewoner.dma.be (Lieven Marchand)
Date: 24 Dec 2000 22:16:34 +0100
Subject: One Python 2.1 idea
References: <mailman.977684587.22689.python-list@python.org>
Message-ID: <m3hf3tcyfh.fsf@localhost.localdomain>

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

> [Lieven Marchand]
> > I think Common Lisp shows that getting very good speed in Python is
> > quite feasable. The original CMUCL implementers weren't such a large
> > group. Their highly optimising compiler (which incidentally is also
> > called Python) has on occasion beaten FORTRAN at numerics. I don't
> > know why some people in the Python community think compiling Python is
> > such a problem.
> 
> The reasons vary, in large part depending on whether or not they've ever
> worked on a compiler <wink>.  Whatever, the fact is that there isn't a
> Python compiler, nor a credible plan for such a beast to come into
> existence.  If the CMUCL compiler group contained at least one person, it
> was infinitely larger than the group of people working on a Python compiler
> (Vyper notwithstanding).
> 

Perhaps because the community doesn't feel the need for a compiler
very strongly? I'd say the same effort that got you JPython could get
you a compiler.

> > Practically all the problems have been tackled and
> > solved 20 years ago in the Lisp community.
> 
> Excellent!  Then we would greatly appreciate it if you took one of their
> compilers, plugged in a Python parser, and released the result.  Sounds like
> it shouldn't take you more than a week or two, given that all the problems
> were solved so long ago.

I didn't say it wouldn't be a lot of work. I was reacting to some
people that feel that because Python is more dynamic than e.g. C it's
impossible to compile.

-- 
Lieven Marchand <mal at bewoner.dma.be>
Lambda calculus - Call us a mad club


From MarkH at ActiveState.com  Thu Dec 28 07:34:51 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Thu, 28 Dec 2000 12:34:51 GMT
Subject: PythonWin Question
References: <3A443EF1.125426ED@home.com>
Message-ID: <3A4B3244.6080502@ActiveState.com>

John Q. Public wrote:

> While trying to add a journaling mechanism to python, I came across an
> odd behavior from PythonWin when I tried to override sys.stdout.  For
> both the console and IDLE versions of the interpreter, I can override
> sys.stdout, but not for PythonWin.
> 

This is a "feature".  Quite often I found people running scripts that 
themself set sys.stdout.  This would then mess-up Pythonwin after the 
script had completed - so Pythonwin resets stdout after running a script.

Mark.



From fredrik at effbot.org  Tue Dec 12 12:53:23 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 12 Dec 2000 17:53:23 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <915lrg$lp9$1@nnrp1.deja.com>
Message-ID: <nCtZ5.3569$Kd1.415539@newsb.telia.net>

Simon wrote:
> Pardon my choice of diction, but I am against inventing
> new terminiologies (modified-in-place??).
>
> The correct terminology is mutable (for objects which can be modified in
> place), and immutable (for objects which cannot).

Well, I doubt anyone who doesn't understand the meaning
of "modifying something in place" understands exactly what
"mutable" means...

(especially since the latter is as likely to mean "may change
itself" as "may be changed" ;-)

But maybe I should use "in situ" to make it sound a bit more
formal...

</F>




From robin at alldunn.com  Fri Dec 15 12:15:09 2000
From: robin at alldunn.com (Robin Dunn)
Date: Fri, 15 Dec 2000 09:15:09 -0800
Subject: launching a python program from GNU make (Win2K)
References: <918lp0$6q4$1@nnrp1.deja.com> <slrn93jn66.p9r.hat@se-46.wpa.wtb.tue.nl>
Message-ID: <Cks_5.170$SKV4.45482166@news.randori.com>

"Albert Hofkamp" <hat at se-46.wpa.wtb.tue.nl> wrote in message
news:slrn93jn66.p9r.hat at se-46.wpa.wtb.tue.nl...
> On Wed, 13 Dec 2000 20:22:02 GMT, jschmitt at vmlabs.com
<jschmitt at vmlabs.com> wrote:
> >From my experience, GNU make has the most features that I want to use.
> >However, I find it so hard to use under Windows.  For example, I wrote
>
> No kidding :-)
>
> >a small python program that I can launch from the command line and
> >execute without problems.  I can't seem to launch it from GNU make.  I
>

I have a version of GNU make that I use once in a while on Win32 (all
varieties) that does just fine.  It's a bit older (3.74.4) but I've had no
problems with it.  I don't remember where I got it, but it wasn't part of
the cygwin toolkit but some separate collection of unix-like command line
tools.

I'll send that copy of gmake.exe to anybody that's interested.

>
> >something to do with GNU make going out of it's way to launch sh.exe if
> >it's on your hard drive, rather than launching command.exe to shell out
>
> Nope, it is simply not possible to launch a command-line program from
> another command-line program.


This is not true.  I do it every day.


--
Robin Dunn
Software Craftsman
robin at AllDunn.com
http://wxPython.org     Java give you jitters?
http://wxPROs.com        Relax with wxPython!





From jflores at codeit.com  Mon Dec 18 14:54:26 2000
From: jflores at codeit.com (Julio Flores Schwarzbeck)
Date: Mon, 18 Dec 2000 11:54:26 -0800
Subject: float to string conversion
In-Reply-To: <91lmt3$4b8$1@nnrp1.deja.com>
Message-ID: <5.0.2.1.0.20001218115206.00a257e0@mail.codeit.com>

Take a look at the following python snippet (key in from the python prompt)..

x=5.25
y="Bad Check"
print "Error for ammount %f, Error Message: %s" % (x, y)
print "Error for ammount %s, Error Message: %s" % (x, y)

Hope it helps.

Julio


At 07:00 PM 12/18/00 +0000, deragon at 8d.com wrote:
>Greetings.
>
>
>   I am embarrassed to ask this since it is trivial, but
>   I cannot figure out how to convert a float to a string (obviously,
>   I am new to python).  Somehow, I could not find in the 2.0
>   documentation the syntax to perform the operation.
>
>   The following in my function:
>
>     return MSISDN + " check FAILED:  $" + balance + errormsg
>
>   where balance is a floating number and errormsg is a string
>   generates the following error:
>
>     TypeError: cannot add type "float" to string
>
>
>Sincerely,
>Hans Deragon
>
>
>Sent via Deja.com
>http://www.deja.com/
>--
>http://www.python.org/mailman/listinfo/python-list




From karacim at my-deja.com  Tue Dec 12 01:56:43 2000
From: karacim at my-deja.com (karacim at my-deja.com)
Date: Tue, 12 Dec 2000 06:56:43 GMT
Subject: PgSQL and double quotes...
Message-ID: <914i7a$q0a$1@nnrp1.deja.com>

Hi,
PgSQL doesn't seem to like the data that is
sometimes in binary format and in somecases
includes double/single quotes (ie "" or '' etc..)
which causesng PgSQL to invoke parsing error...!

I wrote a simple class before the execute sintax
as :
  string.join(string.split(text, "'"), "''")
but that still doesnt get rid of this "parsing
error" from PgSQL... anyone has any ideas?

thx

Aus


Sent via Deja.com http://www.deja.com/
Before you buy.


From rapto at arrakis.es  Mon Dec 18 13:01:15 2000
From: rapto at arrakis.es (rapto at arrakis.es)
Date: Mon, 18 Dec 2000 18:01:15 GMT
Subject: More docs in Spanish/Más madera
Message-ID: <91ljd7$17v$1@nnrp1.deja.com>

New files in
Ficheros nuevos en
http://pySpanishDoc/sourceforge.net/


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


From joseamartin at venmex.com  Thu Dec 14 07:33:12 2000
From: joseamartin at venmex.com (Jose Martin)
Date: Thu, 14 Dec 2000 08:33:12 -0400
Subject: Changing Procesor Flags and word state From Python
Message-ID: <t3hg1fpdpvkg7d@corp.supernews.com>

Is there any way to work with CPU registers and the flags directly from
python?

...?





From eblakes at drwtrading.com  Thu Dec  7 18:50:00 2000
From: eblakes at drwtrading.com (Edward Blakes)
Date: Thu, 7 Dec 2000 18:50:00 -0500
Subject: Implementing IPersistStorage using PythonCOM
Message-ID: <t308m58pkecra3@corp.supernews.com>

Hi all,

I am having a problem creating a COM server that implements IPersistStorage.
Below is the code for the server. I am using ActivePython build 202.

I have removed the bodies of the function members to keep this posting as
small as possible. Also, they do not seem to be the root of the problem.
After the code I have some additional information about the problem.

Here is the code.

CATID_PerspectiveAttributeTypeServer = pythoncom.MakeIID("{a guid}")
CLSID_PythonSample = pythoncom.MakeIID("{a guid}")

class DummyAttr:
    SampleProperty = ""

    _public_attrs_ = ["SampleProperty"]
    _public_methods_ = ["DoSomething", "GetClassID", "IsDirty", "InitNew",
"Load", "Save", "SaveCompleted", "HandsOffStorage"]
    _reg_clsid_ = CLSID_PythonSample
    _reg_progid_ = "PerspectiveAttribute.PythonSample"
    _reg_catids_ = [CATID_PerspectiveAttributeTypeServer]
    _com_interfaces_ = [pythoncom.IID_IPersist,
pythoncom.IID_IPersistStorage]

    def __init__(self):
        ...
    def DoSomething(self):
        ...
    def GetClassID(self):
        ...
    def IsDirty(self):
        ...
    def InitNew(self):
        ...
    def Load(self, pStg):
        ...
    def Save(self, pStg, fSameAsLoad):
        ...
    def SaveCompleted(self, pStg):
        ...
    def HandOffStorage(self):
        ...

class LockBytes:
... (An almost exact copy of the class in testPersist.py for testing only)

def Test():
    import win32com.client
    import win32com.server.util

    # Create the server
    origObj = DummyAttr()
    aServer = win32com.server.util.wrap(origObj,
pythoncom.IID_IPersistStorage)
...
    # Test persistence
    aLockBytes = win32com.server.util.wrap(LockBytes(),
pythoncom.IID_ILockBytes)
    pStg = pythoncom.StgCreateDocfileOnILockBytes(aLockBytes,
storagecon.STGM_DIRECT|storagecon.STGM_CREATE|storagecon.STGM_READWRITE|stor
agecon.STGM_SHARE_EXCLUSIVE, 0)

    origObj.Save(pStg,1)   <--- This is works.
    aServer.Save(pStg, 1)  <--- This fails with the traceback
    pStg.Commit(storagecon.STGC_DEFAULT)
...

def Register():
    from win32com.server.register import UseCommandLine
    return UseCommandLine(DummyAttr)

if __name__ == '__main__':
    import sys
    if "/test" in sys.argv:
        print "Testing..."
        Test()
    else:
        Register()

End code.

When running the code in "test" mode I get the following traceback message.
Note that the line numbers won't match because I was using the original
source.

Traceback (most recent call last):
  File "PythonAttributeSample.py", line 176, in ?
      Test()
     File "PythonAttributeSample.py", line 158, in Test
      aServer.Save(pStg, 1)
   pywintypes.com_error: (-2147352573, 'Member not found.', (0, None,
'Member not found', None, 0, -2147352573), None)

When stepping through the code in the PythonWin debugger, stepping into
aServer.Save() goes to a method that looks like something that implements
IDispatch::Invoke(). Not what I expect, even though aServer is a
PyIPersistStorage. A dir(aServer) (and, coincidentally a dir(origObj)) shows
a method named Save.

When I create the server using another client (C++), I get error 0x80020003
(which matches the traceback) when using IPersistStorage::Save() or
IPersistStorage::Load(). The server is instantiated without incident and the
QueryInterface for IID_IPersistStorage works perfectly. Category
registration works well, too. (The component category registration support
is sweet! This is not something VB can do on its own.)

Admittedly, I am a newbie to Python, though not a newbie to COM. This is
probably a more ambitious project than a newbie should tackle but I have a
specific need to build a automation server that can persist to IStorage but
is easily and quickly modifiable. Python fits the bill. I'd appreciate any
help in figuring this one out. I just don't see what I am doing wrong.

Thanks in advance for any help.

Ed Blakes
eblakes at drwtrading.com




From sbrunning at bigfoot.com  Mon Dec  4 05:05:53 2000
From: sbrunning at bigfoot.com (Simon B.)
Date: Mon, 04 Dec 2000 10:05:53 GMT
Subject: Python mailing list/usenet bridge.
References: <8vt9t4$r0f$1@nnrp1.deja.com> <900365$fp6$1@nnrp1.deja.com> <905t5s$6ef$1@nnrp1.deja.com>
Message-ID: <200012041005.EAA23178@x56.deja.com>

*Still* nothing. This is getting a bit irritating - deja is horrid.
*Please*, anyone know who I can nag to get this sorted?

--
Simon Brunning
I was thrown out of college for cheating on the metaphysics exam; I
looked into the soul of the boy sitting next to me. - Woody Allen



From barrett at stsci.edu  Tue Dec  5 10:01:05 2000
From: barrett at stsci.edu (Paul Barrett)
Date: Tue, 05 Dec 2000 10:01:05 -0500
Subject: A better python interface for buffer objects
References: <3A2C1188.D384E47C@stsci.edu> <90h8320n4r@news2.newsguy.com>
Message-ID: <3A2D0331.C551FEDF@stsci.edu>

Alex Martelli wrote:

> > confusing?)  Since I can't use a string object as a writeable buffer, is
> > there some other object that I can use as a writeable buffer, not
>
> Maybe an array.array object?  Haven't tested... it should, I guess.
>
> Alex

I've tried this and it also appears to be a read-only buffer!

I'm not aware of any other options at this time.

I suppose a PEP may be in order here.

--
Dr. Paul Barrett       Space Telescope Science Institute
Phone: 410-338-4475    ESS/Science Software Group
FAX:   410-338-4767    Baltimore, MD 21218





From news at dorb.com  Sun Dec 31 11:43:59 2000
From: news at dorb.com (Darrell)
Date: Sun, 31 Dec 2000 16:43:59 GMT
Subject: HTML Parser
References: <mailman.978224102.28797.python-list@python.org> <n1z36.208621$U46.6407059@news1.sttls1.wa.home.com> <y5z36.208644$U46.6407855@news1.sttls1.wa.home.com>
Message-ID: <jnJ36.5$Rb.28443@typhoon.nyroc.rr.com>

When parsing large documents I found this to be much faster.

>>> import re, time
>>> s="xxx<dog a='a'>yyyy"
>>> s1=s*1000000
>>> t1=time.time();res=re.findall("<[^>]*?>", s1);time.time()-t1
45.437000036239624
>>> t1=time.time();res=re.findall("<.*?>", s1);time.time()-t1
82.343999981880188
>>>

Although to my surprise the patched version of _sre.pyd I use inverts these
results.

>>> t1=time.time();res=re.findall("<[^>]*?>", s1);time.time()-t1
75.577999949455261
>>> t1=time.time();res=re.findall("<.*?>", s1);time.time()-t1
49.296999931335449

The patch looks for ".*?" and optimizes out the recursion.

Both versions run much faster using sre than pre.
Here's pre
>>> t1=time.time();res=re.findall("<[^>]*?>", s1);time.time()-t1
83.828999996185303
>>> t1=time.time();res=re.findall("<.*?>", s1);time.time()-t1
123.43700003623962
>>>

--Darrell

"Greg Jorgensen" <gregj at pobox.com> wrote:
> "Greg Jorgensen" <gregj at pobox.com> wrote:
>
> > rx = re.compile('(<.*?>)', re.MULTILINE)
>
> Oops -- that should be:
>
>     rx = re.compile('(<.*?>)', re.DOTALL)
>
> That makes the . match the newlines. You need that because HTML tags can
> span lines.
>





From dalke at acm.org  Fri Dec 15 00:30:49 2000
From: dalke at acm.org (Andrew Dalke)
Date: Thu, 14 Dec 2000 22:30:49 -0700
Subject: How to use mxTextTools
References: <2bc4OuYTMVp87e49nlcz01wv0ybq@4ax.com>
Message-ID: <91ca70$6cp$1@slb7.atl.mindspring.net>

Paul Moore wrote:
>I'm looking at mxTextTools to see if it would be suitable for some
>types of text parsing work I am interested in (nothing concrete yet,
>so I can't give specifics...)

I am using it as the basis of a parsing project for biopython.org
called Martel (http://www.biopython.org/~dalke/Martel/ ).  It
turned out to be very useful, but its learning curve was quite
high.  I think it took me about 4 days of playing around with
it to feel okay using it, and another few weeks of using it
to feel proficient.

>What I dont't see (yet), and I can't find any good examples for, is
>what to do with the resulting taglist. There seem to be no functions
>for working with taglists, and the lists themselves seem like
>relatively complex data structures, so is it right that I should be
>manipulating them "by hand"?

There aren't any that I've come across.  One of the downside of
Martel is that it produces pretty unoptimized tagtables.  I know
of a few ways to get better performance out of them but they
requiring modifications which are too cumbersome to implement
directly by hand.  So if you are interested in working on such
things, get ahold of me :)

>More information, or better still, some complete examples, would be
>very helpful.

You can look at Martel.  It's a parser generator which uses a
(possibly very large) regular expression as the format description.
It creates a corresponding tagtable which is used by the parser
to parse those files.

So if you know regular expresssions you'll be able to see how
to make some of those constructs to mxTextTools.  However,
mxTextTools is more powerful than regular expressions so it
doesn't work the other way.

                    Andrew
                    dalke at acm.org





From fredrik at effbot.org  Tue Dec 19 19:44:26 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 20 Dec 2000 00:44:26 GMT
Subject: xml-rpc module question
References: <dmC%5.75166$yR4.2151642@news1.rdc1.tx.home.com> <j7L%5.879$VF3.1528@news.more.net> <%3T%5.76381$yR4.2212181@news1.rdc1.tx.home.com> <V9T%5.651$Qb7.44185@newsb.telia.net>
Message-ID: <KhT%5.652$Qb7.44208@newsb.telia.net>

I wrote:
>     def call(self, method, params):
>         if method == 'CubeIt':
>             return self.CubeIt(params)

correction: params is a tuple.  tuples cannot be cubed.

better make that:

             return self.CubeIt(params[0])

or, better:

             return apply(self.CubeIt, params)

(anybody know a good book on debugging with print
statements? ;-)

</F>




From bob at lenaerts.org  Tue Dec 26 11:09:15 2000
From: bob at lenaerts.org (bob at lenaerts.org)
Date: Tue, 26 Dec 2000 16:09:15 GMT
Subject: did you ever see something like this ?  8177
Message-ID: <Lo326.89120$Ww.1931373@afrodite.telenet-ops.be>

take a look at http://www.webline.be and see the rates for hosting your site. Its almost for free. And you must check out its speed.
WAAAAAUUUW. great.
take a look
ytxokcfhxcwlculxljckdfysqrpnxurjrevlbjxmgipwuhwbxpcdwf



From akuchlin at mems-exchange.org  Mon Dec  4 16:22:16 2000
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 04 Dec 2000 16:22:16 -0500
Subject: int multiplication overflow and the future of python
References: <ghaelzoc9lp.fsf@synopsys.com>
Message-ID: <3dlmtv286v.fsf@platinum.cnri.reston.va.us>

Ram Bhamidipaty <ramb at synopsys.com> writes:
> there is a fix....but I have not been able to find it. Is there
> a patch that I should apply to upgrade my 1.6 source to the
> latest version?

This is because of the broken glibc that Red Hat shipped with 7.0; see
the brief Python 2.0 FAQ at
http://www.python.org/cgi-bin/moinmoin/FrequentlyAskedQuestions .  The
fix is to get the updated glibc RPM from Red Hat's errata page and
install it.

> This leads to a larger question: Which version is being maintained?
> Is active developement taking place for both 1.6 and 2.0? I've been

2.0.  1.6 is basically the state of the Python CVS tree as of when the
developers left CNRI in May 2000, plus a minimal set of bug and
compatibility fixes.  2.0 is the CVS tree with some extra months of
development, so there are more bugfixes and more new features.
There's no real reason to stick with 1.6; I'm not aware of any
instabilities that 2.0 introduced.  (If anyone knows of such, *please*
report the problem!)

--amk



From DavidL at nospamy_raqia.com  Thu Dec 14 15:55:38 2000
From: DavidL at nospamy_raqia.com (David Lees)
Date: Thu, 14 Dec 2000 15:55:38 -0500
Subject: Use of httplib?
References: <3A37E456.772811C9@nospammy_raqia.com>
Message-ID: <3A3933CA.FF688279@nospamy_raqia.com>

Thanks to the kind help of Steve Holden, I have a much better solution
than the httlib one below.  I used the urllib, with the code:

f = urllib.urlopen("http://192.168.0.3:50006",body)
m = f.info()
NumBytes = string.atoi(m.getheaders('content-length')[0])
BytesRead = 0

while (BytesRead < NumBytes):
    data = f.readline()
    BytesRead = BytesRead+len(data)
    print data

f.close()

This works fine and seems easier than the http below or socket level
that I previously tried.

david



David Lees wrote:
> 
> I am running the client code below to try and send and receive text from
> a server.  My problem is that I do not know how to recieve back an
> arbitrary number of lines.  In a test case, when I know the exact number
> of lines as shown below (e.g. 11) I can use readline.  However, in
> general I do not know how many lines there are and the program hangs on
> the connection when  I accidentaly read to the end.  The comment out
> pieces of code also either hang or do not give back all the text.
> 
> I am copying from the Library Reference section 11.4, but it does not
> seem to document the readline, read or readlines methods for http.
> 
> Thanks in advance for any help/pointers.
> 
> david lees
> 
> h = httplib.HTTP("192.168.0.3:50006")
> h.putrequest('POST','/rpchandler HTTP/1.0')
> h.putheader('Content-Type', 'text/xml')
> h.putheader("Content-length", "%d" % len(bod))
> h.endheaders()
> h.send(bod)
> reply, msg, hdrs = h.getreply()
> print reply
> print msg
> print hdrs
> f = h.getfile()
> 
> #data=f.read()
> #print data
> 
> #data=f.readlines()
> #print data
> 
> for i in range(11):
>     data=f.readline()
>     print data
> 
> 
> #while 1:
> #   data=f.readline()
> #    if not data:
> #       break
> #    print data
> 
> f.close()
> h.close()


From fredrik at effbot.org  Thu Dec 21 13:06:28 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Thu, 21 Dec 2000 18:06:28 GMT
Subject: A Suggestion for Python Colon Syntax
References: <91tc8h$2pp$1@hecate.umd.edu> <W2r06.1076$AH6.86249@newsc.telia.net> <91tfg1$282$1@hecate.umd.edu>
Message-ID: <EEr06.842$Qb7.73987@newsb.telia.net>

"William Djaja Tjokroaminata" wrote:
> If a feature is borrowed from C or Java or Perl, I think it is good.  But
> a feature that seems to add nothing from an obscure language to me
> is questionable.

oops.  methinks you should read up on Python's history...

start here:

    http://www.python.org/doc/essays/foreword.html

    "I decided to write an interpreter for the new scripting
    language I had been thinking about lately: a descendant
    of ABC that would appeal to Unix/C hackers. I chose
    Python as a working title for the project /.../

(incidentally, that text also talks a lot about readability...)

</F>




From tim.one at home.com  Sun Dec 31 05:13:27 2000
From: tim.one at home.com (Tim Peters)
Date: Sun, 31 Dec 2000 05:13:27 -0500
Subject: Python vs. PHP (& Java?)
In-Reply-To: <92ig0g$jgl$1@nnrp1.deja.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEPHIFAA.tim.one@home.com>

[rturpin at my-deja.com]
> ...
> Looking at something where the stakes are much less, the
> benefit for each individual more clear, and the new skill
> is easily learned, cultural inertia STILL wins out over
> better notation. No one claims that the QWERTY keyboard
> is technically superior to the Dvorzak. I suspect, though,
> that the QWERTY keyboard will remain the major character
> input device for the next two or three decades.

Perhaps if the Chinese never take up typing <wink>.  WRT English speakers,
Liebowitz and Margolis published a now-famous paper ("The Fable of the
Keys") that kicked off a wave of serious skepticism about the oft-claimed
technical superiority of the Dvorak layout:

    http://wwwpub.utdallas.edu/~liebowit/keys1.html

They had an economic axe to grind (i.e., their main interest wasn't
typewriter design), but all the same it's an interesting read.

>(Before thinking that voice recognition will knock it out soon,
> see how long your throat lasts speaking ten pages of text.  Twice,
> with edits.)

That's not hard for an experienced speech-recog user (but is probably harder
for an inexperienced one than for an inexperienced typist -- contra Star
Trek, these are both learned skills, and people are very resistant to the
notion that they need to *learn* how to speak to a computer).  Note that the
most steadfastly loyal market for speech recog is among people who need to
produce truly vast quantities of text every day, like medical and legal
transcriptionists.  Ten pages isn't a strain if you're using a normal,
quiet-room conversational voice (picture reading a book chapter to your
kid!).  Eight hours per day of repeating *anything* may lead to medical
problems, though.

with-the-exception-of-programming-python-ly y'rs  - tim




From aleaxit at yahoo.com  Tue Dec 26 19:16:12 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 27 Dec 2000 01:16:12 +0100
Subject: One Python 2.1 idea
References: <LNBBLJKPBEHFEDALKOLCCEIFIFAA.tim.one@home.com> <mailman.977868490.18316.python-list@python.org>
Message-ID: <92bcfj03078@news1.newsguy.com>

"Darrell Gallion" <darrell at dorb.com> wrote in message
news:mailman.977868490.18316.python-list at python.org...
    [snip]
> You lost me here Tim.
>
> >>> class A:
> ...     def f(self):
> ...             pass
> ...
> >>> a=A()
> >>> a.f.im_self is a
> 1

In this case, a.f (the bound-method) is transiently
constructed from A.f (the unbound-method), and
goes away if no reference to it is kept; and indeed
after this sequence:

>>> sys.getrefcount(a)
2

so variable 'a' is the only reference extant to this
instance object (remember getrefcount always
returns one more!-).

> >>> a.f=a.f
> >>> a.f.im_self is a
> 1
> >>>
>
> Don't see what's changed after a.f=a.f

Now, you've bound a reference to the bound-method,
so, it's kept alive -- and holds a reference to a.  Indeed,
now:

>>> sys.getrefcount(a)
3

This would be true wherever you kept a reference to
bound-method a.f -- but if that reference was in some
slot not connected with a, no cycles would exist.  As
the slot you're using (a.f) is in a's instance dictionary,
then the references form a circular 'loop':

a -> its instance dictionary -> boundmethod a.f -> ... -> a


Alex






From dalke at acm.org  Fri Dec 15 10:51:26 2000
From: dalke at acm.org (Andrew Dalke)
Date: Fri, 15 Dec 2000 08:51:26 -0700
Subject: __repr__ policy problem: should module be prefixed to output?
References: <3A3A3A6D.2C3B1A91@sbc.su.se>
Message-ID: <91dekh$7gq$1@nntp9.atl.mindspring.net>

Per Kraulis:
>I have received some constructive
>criticism, one issue being save/restore functionality. My plan is to
>define a __repr__ for each type in the module.
>
>However, I have realized that there is a policy problem.
>Should __repr__ add the module name as prefix to the output
>string, or not? Or is there some other solution?

You should take a look at the pickle (and cPickle) modules.
See http://www.python.org/doc/current/lib/module-pickle.html

                    Andrew Dalke
                    dalke at acm.org





From fredrik at effbot.org  Tue Dec 12 12:42:13 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 12 Dec 2000 17:42:13 GMT
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com>
Message-ID: <VrtZ5.3567$Kd1.415273@newsb.telia.net>

"Steve" wrote:
> I dislike infinite while loops and use:
>
> for line in file.readlines()
>     ...do stuff...

what makes you so sure that's not an infinite loop?

(well, in the current implementation, it isn't: if your file
has more than 2147483647 lines, your code won't read
all of them... ;-)

</F>




From johngrayson at home.com  Tue Dec 19 14:12:42 2000
From: johngrayson at home.com (John Grayson)
Date: Tue, 19 Dec 2000 19:12:42 GMT
Subject: tkinter event binding to mouse double click
References: <91n7l6$7ij$1@news.nuri.net> <91njgc$l4q$1@nnrp1.deja.com> <91nmqj$ls4$1@news.nuri.net>
Message-ID: <91obuv$be6$1@nnrp1.deja.com>

In article <91nmqj$ls4$1 at news.nuri.net>,
  "June Kim" <junaftnoon at nospamplzyahoo.com> wrote:

>
> Well, it doesn't work, either. Probably it will work on linux but not
on
> win32.
> Thanks anyway.
>
> Best Regards,
>
> June
>
>

Hmmm...

I tested it on NT 4.0, so you've got some other problem, Not Win32.


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


From dsh8290 at rit.edu  Tue Dec 19 13:42:53 2000
From: dsh8290 at rit.edu (D-Man)
Date: Tue, 19 Dec 2000 13:42:53 -0500
Subject: Python interpreter in python?
In-Reply-To: <91ndgd$g8d$1@bunyip.cc.uq.edu.au>; from teyc@bigfoot.com on Tue, Dec 19, 2000 at 08:31:03PM +1000
References: <91ndgd$g8d$1@bunyip.cc.uq.edu.au>
Message-ID: <20001219134253.B15468@westsidecnc.rh.rit.edu>

The problem with such a thing is that to run your interpreter you
would need an existing interpreter (written in C/Java or other
language).  The only languages that can be implemented in themselves
are compiled languages like C and C++.

What experimenting are you planning on doing?  Python does have an
'eval' function that will evaluate the given python code.

-D

On Tue, Dec 19, 2000 at 08:31:03PM +1000, Chui Tey wrote:
> Is there a python implementation of the python VM? Basically for
> experimenting?
> 
> Chui
> 



From cjensen at bioeng.ucsd.edu  Tue Dec 12 12:44:05 2000
From: cjensen at bioeng.ucsd.edu (Curtis Jensen)
Date: Tue, 12 Dec 2000 09:44:05 -0800
Subject: Python-like compiled language
References: <913jr3$dmv$1@sylvester.vcn.bc.ca> <3A35835B.AAA92005@bioeng.ucsd.edu> <3A364CC2.263C0958@gssec.bt.co.uk>
Message-ID: <3A3663E5.26E7115D@bioeng.ucsd.edu>

Alan Gauld wrote:
> 
> Curtis Jensen wrote:
> > get LaPack lite for free.  Odepack is free and it's from llnl.  Nag is
> > NOT free, it is very expensive.
> 
> Are you sure about that? I thought I red recently that NAG was
> now open source? Their gift for the millenium or something...
> 
> I may be wrong, I don't use NAG just noticed a magazine newsbyte.
> 
> Alan g.
> 

I'm sure it's not open source.  To be even more sure, I did some
searches.  I found another Nag.  There is the "Network Administrator's
Guide," which is free and can be downloaded from RedHat.  Perhaps that
is what you saw.  There may be other packages called Nag, but the Nag
from the Numerical Algorithms Group is not open source.  It'd be nice if
it was.  We just paid a heap of money for a site license.

-- 
Curtis Jensen
cjensen at bioeng.ucsd.edu
http://www-bioeng.ucsd.edu/~cjensen/
FAX (425) 740-1451


From fredrik at effbot.org  Wed Dec 20 17:47:05 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 20 Dec 2000 22:47:05 GMT
Subject: Python in embedded systems?
References: <_w706.11237$TU6.1069571@ptah.visi.com>
Message-ID: <JFa06.749$Qb7.61608@newsb.telia.net>

Grant Edwards wrote:
> Has anybody heard of a Python interpreter being used in an
> embedded system to provide user scripting capabilities?

    http://www.abo.fi/~iporres/python/

    "Deeply Embedded Python"

    "With this patch, it is possible to build a python
    interpreter WITHOUT support for float, complex
    and long numbers, file objects, parser, compiler
    and depedencies with any operating system."

</F>




From k_j_r_a_y at ix.netcom.com  Wed Dec 13 21:54:00 2000
From: k_j_r_a_y at ix.netcom.com (Keith Ray)
Date: Thu, 14 Dec 2000 02:54:00 GMT
Subject: Programming Habits in Python
References: <cspU5.82459$U46.2617014@news1.sttls1.wa.home.com> <8vvb6p$ct2$1@news.wrc.xerox.com> <XiMU5.87123$U46.2803855@news1.sttls1.wa.home.com> <900cpl02ss1@news2.newsguy.com> <Ue3V5.91231$U46.2956112@news1.sttls1.wa.home.com> <slrn92tv0u.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com> <910m5d$t6v$1@slb3.atl.mindspring.net>
Message-ID: <k_j_r_a_y-958C59.19032913122000@news>

In article <910m5d$t6v$1 at slb3.atl.mindspring.net>, "Andrew Dalke" 
<dalke at acm.org> wrote:
[...]
> As yet another comparison, I've been doing a lot of work with
> mxTextTools, which uses a state table to do text processing.
> It's very fast so I rewrote some of my Python parsing code
> using it.  While there were only about 12 states, it took
> a few hours to write and debug while the Python implementation
> took about 15 minutes.
> 
> I ended up coding the example in C++ instead of Python because
> C++ has gotos and because C++'s STL includes a priority queue.
> The whole code ended up being about 5 times longer than the
> MIX example, which isn't bad since MIX puts a lot of the
> code on one line, and he didn't declare a priority queue.
[...]

Could you have put the code for each of those 12 states into separate 
classes, and put instances of those classes into a dictionary where the 
key is the state-id?

-- 

       <http://pw2.netcom.com/~kjray/>


From curtissol at my-deja.com  Sat Dec 30 20:48:09 2000
From: curtissol at my-deja.com (curtissol at my-deja.com)
Date: Sun, 31 Dec 2000 01:48:09 GMT
Subject: os module - reliable?
Message-ID: <92m38n$88f$1@nnrp1.deja.com>

I am writing exercise 1 pg 212 of "Learning Python". It asks for a
program that prints out the names of all files in a specific directory
and all its subdirectories. The code I have seems relatively straight
forward:

import os, sys

import os, sys

def ShowDirContents(dir):
    contents = os.listdir(dir)

    for file in contents:
        if os.path.isdir(file):
            ShowDirContents(file)
        else:
            print file



if __name__ == "__main__":
    ShowDirContents(os.getcwd()+"\Doc")

However, the test
   if os.path.isdir(file):

seems to only recognize some of subdirectories of "...Python20\Doc".
Those that are recognized it handles just fine. Those it does not
recognize, it simply treats as files. Also, as you may know, the "Doc"
directory contains a subdirectory called 'doc'. When the code sees this,
it goes into an infinite loop only outputing some of files in Doc before
it enters 'doc' again. This behavior goes away if I change the name of
the subdirectory 'doc'.

How reliable are the os and os.path modules? I am running Python 2.0
with NT 4.5 on a PC. Windows NT Explorer shows a number of directories
(folders) that the 'if' test shown above misses.

Thanks


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


From rhamorim at my-deja.com  Mon Dec  4 07:36:48 2000
From: rhamorim at my-deja.com (rhamorim at my-deja.com)
Date: Mon, 04 Dec 2000 12:36:48 GMT
Subject: Pmw.CounterDialog
Message-ID: <90g351$prk$1@nnrp1.deja.com>

Hi, to correct it is simple. You'll have to edit PmwCounter.py. Here it
is:

- Find the function called _changeNumber() (line 273)
- On line 279 there's the return statement for this function. Change it
to return str(value) (and not return str(value)[:-1])

That ocurred because str(long) on Python 2.0 doesn't return a number
with L on the end like Python 1.5.2 did.

Tell me if you have any further problems.

Roberto Amorim


Sent via Deja.com http://www.deja.com/
Before you buy.


From shindich at my-deja.com  Fri Dec  8 20:36:28 2000
From: shindich at my-deja.com (shindich at my-deja.com)
Date: Sat, 09 Dec 2000 01:36:28 GMT
Subject: ANN.: Weak Reference Extension Module has been converted to "C"; Beta 2 is out
References: <8usorp$3i3$1@nnrp1.deja.com> <mtaeb1f4q3.fsf@astron.berkeley.edu> <8uvjmj$dt8$1@nnrp1.deja.com> <mtlmuk9hyc.fsf@astron.berkeley.edu> <B63C5A9E.172D7%graham73@telocity.com> <3A176E0F.39BF350B@felinity.com> <KaNR5.5550$QH2.654863@newsb.telia.net> <8v9dl9$si6$1@nnrp1.deja.com> <90i3e3$f8t$1@nnrp1.deja.com> <p6qu28ek7zg.fsf@informatik.hu-berlin.de>
Message-ID: <90s2as$ibo$1@nnrp1.deja.com>

In article <p6qu28ek7zg.fsf at informatik.hu-berlin.de>,
  Martin von Loewis <loewis at informatik.hu-berlin.de> wrote:
> shindich at my-deja.com writes:
>
> > I am happy to announce that Weak Reference Extension project has
been
> > fully converted to "C". Weak Reference extension is now completely
> > CPython compatible.
>
> That's indeed a big improvement. Congratulations!
>
> Martin
>
Do I sense sarcasm in your voice? ;)


Sent via Deja.com http://www.deja.com/
Before you buy.


From MarkH at ActiveState.com  Mon Dec  4 23:59:55 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Tue, 05 Dec 2000 04:59:55 GMT
Subject: General default exception handler
References: <lj0n2to9vs8bmv3oo49mqgoc6h7esh8t37@4ax.com>
Message-ID: <3A2C74FE.6080300@ActiveState.com>

Dale Strickland-Clark wrote:

> We have a substantial set of COM objects in Python that run a Web site.
> 
> Currently, if an exception is raised, the user gets an unfriendly message in their browser and we
> may never know about it.
> 
> I want to log all unhandled exceptions but I'd rather not have to put a try/except block around
> every entry point - because there are dozens of them.
> 
> Is there a general way to catch all unhandled exceptions so I can log them and pass back something
> nice and soothing to the user?
> 
> I'm planning to use the traceback module to record the exception info.

The only built in way is to use a "dispatcher".  The existing "debug 
dispatcher" is what "debug mode" uses to log the traceback to the 
win32traceutil module.  You can write your own dispatcher to handle 
errors any way you like.

You can set the default dispatcher by setting a _reg_dispatcher_ (or 
something like that) on the class.

Mark.



From jrfsousaREMOVE00592424 at esoterica.pt  Thu Dec 21 20:54:11 2000
From: jrfsousaREMOVE00592424 at esoterica.pt (José Rui Faustino de Sousa)
Date: Fri, 22 Dec 2000 01:54:11 -0000
Subject: Refreshall from Python 1.3
Message-ID: <977505346.331315@osiris.esoterica.pt>

Hi!

I am learning Python using:

Internet Programming with Python

and

Programming Python

The problem is that both of these books target Python 1.3...

Where can I find some "refreshall" course to put me up to date with
Python 2.0?

Preferably online... Allready too much $$$ on books... ;-)

Best regards
Jos? Rui

--
========================================================================
The trouble with people is not that they don"t know but that they know
so much that ain"t so. - "Josh Billings" (Henry Wheeler Shaw).
========================================================================
Iam://Jos? Rui Faustino de Sousa http://homepage.esoterica.pt/~jrfsousa/
mailto://jrfsousaREMOVE00592424 at esoterica.pt phone://+351-239444940
address://rua Carlos A. Pinto de Abreu n? 30C, 1? 3040 Coimbra Portugal
========================================================================
Real Programmers do sig blocks in Fortran 95.
========================================================================





From db3l at fitlinxx.com  Fri Dec  1 20:47:14 2000
From: db3l at fitlinxx.com (David Bolen)
Date: 01 Dec 2000 20:47:14 -0500
Subject: win32net - How to connect to a remote network share from a service
References: <3A26D207.84B8F0B9@home.com> <uwvdl6khn.fsf@ctwd0143.fitlinxx.com> <3A282FDA.9E97DC67@home.com>
Message-ID: <uelzrlhlp.fsf@ctwd0143.fitlinxx.com>

Mark Tompkins <mdtompkins at home.com> writes:

> BTW, have you any suggestions as to specific Windows API documentation where
> I might look, to gain a better understanding of the issues involved.

In general, any of the stuff wrapped by the win32all modules is best
served by the native Win32 documentation (as augmented with some
Python specifics from the win32all documentation).

The best place for this sort of documentation is the platform SDK
(alone or as part of MSDN), which if you don't have a local copy is
also available on the web at http://msdn.microsoft.com, or more
specifically for the library documentation at:

    http://msdn.microsoft.com/library/default.asp

This particular issue (impact of running services as LocalSystem on
their network access) took me a day or two to locate the first time I
ran into it with my own service (non-Python, not to mention first real
Win32 application).  Once I found the info it was pretty clear but I
had a heck of a time figuring out what to search for :-)

Here's a pointer to an MSJ article that provides some background:

    http://msdn.microsoft.com/library/periodic/period98/service2.htm

and you might also look at the knowledge base articles Q122702,
Q124184 and Q132679 (note that you'll have to select the knowledgebase
from the search page to be able to look up these entries).

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From carsten.dot.geckeler at gmx.dot.de  Fri Dec  1 07:08:05 2000
From: carsten.dot.geckeler at gmx.dot.de (Carsten Geckeler)
Date: Fri, 1 Dec 2000 13:08:05 +0100 (MET)
Subject: re bug?
In-Reply-To: <90781q+pvk3@eGroups.com>
Message-ID: <Pine.LNX.4.10.10012011307500.479-100000@erde.kosmos.all>

On Fri, 1 Dec 2000 rubygeek at yahoo.com wrote:

> active-python2.0, win2k
> 
> > re.sub("(\d+) (\d+)","\1\2","abc 34 23")
> "abc \001\002"
> 
> > re.sub("(\d+) (\d+)","\\1\\2","abc 34 23")
> "abc 3423"
> 
> can anyone explain this?

I'll try.  This is because of python's way how to escape characters.  In
normal strings (you use them), the backslash is to escape special
characters "\n" will give you a newline (code 10?).  In your case you are
using "\d" and "\1"/"\2".  For "\d" python has no special meaning, so it
takes a backslash and the character `d'.  But "\1" is used as the octal
coding of a character, in this case octal, decimal 1.  That's why you get
the first result.  In the second you give python a "\\" which means a
single backslash character `\'.  Actually you are quite lucky, that "\d"
has no special meaning.

To make it short: For regular expressions where you have to use often
backslashes it's best to use so-called raw strings r"...":

re.sub(r"(\d+) (\d+)", r"\1\2", "abc 34 23")
"abc 3423"

Raw stings have no backslash escaping (except for the quotes, but the
backslash will stay in the string).  For more information on this look in
the Python Reference Manual "1.4.1 String literals".

Cheers, Carsten
-- 
Carsten Geckeler:  carsten dot geckeler at gmx dot de
To get proper email-address replace `dot' and `at' by the corresponding symbols.





From joseamartin at venmex.com  Thu Dec  7 10:47:02 2000
From: joseamartin at venmex.com (Jose Martin)
Date: Thu, 7 Dec 2000 11:47:02 -0400
Subject: DCOracle Need Help
References: <t2v9fu2pht2d69@corp.supernews.com> <20001207164453.A7670@yetix.sz-sb.de>
Message-ID: <001501c06064$f1982510$9101a8c0@venmex.com>

Hi, tanks.
i am trying to make the binares for buffer and oci 
it is hard !.

How can i use DCOracle on Windows ?
there is an easy way ?


  ----- Original Message ----- 
  From: Andreas Jung 
  To: Jose Martin 
  Cc: python-list at python.org 
  Sent: Thursday, December 07, 2000 11:44 AM
  Subject: Re: DCOracle Need Help


  On Thu, Dec 07, 2000 at 10:48:32AM -0400, Jose Martin wrote:
  > I am trying to use DCOracle on Python 2.0 under Windows or Linux
  > but when i try to make my scritp :
  > 
  > The first line is
  > from DCOracle import *

  This is not the recommended way to use DCOracle because DCOracle
  is a Python package. Instead use the following:

  import DCOracle

  conn = DCOracle.Connect('my_code/is_bad') 
  ....

  This initializes and imports all neccessary modules you need
  for DCOracle. Hope it works !

  Cheers,
  Andreas
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20001207/b7a27954/attachment.html>

From gregholmes at my-deja.com  Tue Dec 19 08:17:50 2000
From: gregholmes at my-deja.com (gregholmes at my-deja.com)
Date: Tue, 19 Dec 2000 13:17:50 GMT
Subject: tkinter event binding to mouse double click
References: <91n7l6$7ij$1@news.nuri.net> <91njgc$l4q$1@nnrp1.deja.com>
Message-ID: <91nn5r$nq2$1@nnrp1.deja.com>

I think you just explained some mysterious headaches I have had!
Arghh ...

In article <91njgc$l4q$1 at nnrp1.deja.com>,
  John Grayson <johngrayson at home.com> wrote:
> > >
>
> The problem is that you have to click on the *frame* not the (empty)
> space in the frame.
>
> Try this:
>
> frame = Frame(root, bg='red', width=100, height=100)


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


From bh at intevation.de  Mon Dec 11 09:26:31 2000
From: bh at intevation.de (Bernhard Herzog)
Date: 11 Dec 2000 15:26:31 +0100
Subject: exec'ing an import
References: <3A315C1A.90CB8D98@seebelow.org>
Message-ID: <6qk897f2zs.fsf@abnoba.intevation.de>

Grant Griffin <not.this at seebelow.org> writes:

> Finally, I discovered that that when I executed the script "script.py"
> using just the script name (like "script") I got the ImportError, but
> when I executed it by calling the Python interpreter (ike "python
> script.py") it worked fine.

Could it be that you have several Python interpreters installed and
invoking "script" and "python script.py" use different ones (with
resulting different sys.paths)?

  Bernhard

-- 
Intevation GmbH                                 http://intevation.de/
MapIt!                                               http://mapit.de/
Sketch                                 http://sketch.sourceforge.net/


From ben.hutchings at roundpoint.com  Fri Dec 22 08:49:57 2000
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 22 Dec 2000 13:49:57 +0000
Subject: while (thread.alive()):
References: <mailman.977283904.26696.python-list@python.org> <91tgvm$3q8$2@taliesin.netcom.net.uk>
Message-ID: <usnngy396.fsf@roundpoint.com>

"Garry Knight" <garryknight at bigfoot.com> writes:

> In article <mailman.977283904.26696.python-list at python.org> "Jay
> O'Connor" <joconnor at cybermesa.com> wrote:
> 
> > I'm continuously amused that people try to take language design
> > forward and still drag obsolete ideas along for the ride.
> 
> Like having a print command that doesn't, by default, send output
> to the printer...  :o)

I think that bug was introduced by the shift to glass terminals.
Would you like to use a Teletype instead?

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From gmcm at hypernet.com  Sat Dec  2 21:30:42 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: 03 Dec 2000 02:30:42 GMT
Subject: problems with installer 20_3i
References: <_o%V5.47439$yR4.1041961@news1.rdc1.tx.home.com>
Message-ID: <8FFED7BE9gmcmhypernetcom@199.171.54.155>

[posted and mailed]

Scott Hathaway wrote:

>After running:
>
>python standalone.py bw.py
>
>It seems to go ok, but when trying to run the .exe, I get:
>
>C:\Python\install\dist_bw>bw
>Traceback (most recent call last):
>  File "<string>", line 3, in ?
>  File "c:\python\lib\imputil.py", line 91, in _import_hook
>    raise ImportError, 'No module named ' + fqname
>ImportError: No module named time
>
>What did I do wrong?

Nothing. Barry Scott just pointed out to me that the Python 2.0 imputil 
does not by default install any method of importing builtins (even though 
it installs, by default, a file system importer).

Untested: at the top level of archive_rt (after the imports), put:

sys.path.append(imputil.BuiltinImporter())

archive_rt will be imported by the C code long before your script runs, so 
builtins should be available.

- Gordon



From news at dorb.com  Thu Dec 14 22:53:01 2000
From: news at dorb.com (Darrell)
Date: Fri, 15 Dec 2000 03:53:01 GMT
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <tcmay-FCA22F.14073713122000@news.got.net> <918v7a02scn@news1.newsguy.com> <tcmay-04E80C.16181313122000@news.got.net> <91a1ed02a4j@news2.newsguy.com> <3A389B1A.4F33FFD4@cfmu.eurocontrol.be> <91ae8s02llf@news2.newsguy.com> <3A397778.2E6D84D3@cosc.canterbury.ac.nz>
Message-ID: <xAg_5.13536$7g.1916815@typhoon.nyroc.rr.com>

Don't understand why allowing only a single bed within
a process would be handy?

Wouldn't dropping a bed instance into a scene be good enough?

must-be-missing-somethingly y'rs  - Darrell


"Greg Ewing" wrote:
> Just to throw another language into the mix,
> the TADS (Text Adventure Development System)
> lets you declare objects with multiple
> classes, e.g.
>
>    bed: Furniture, Surface, Scenery
>       ...
>
> declares a singleton object called 'bed'
> which belongs to the classes Furniture,
> Surface and Scenery.
>
> This is really just syntactic sugar, since
> you could just as well declare a new class
> inheriting from all those and then instantiate
> it. But it's extremely handy in a language
> designed for applications where the vast
> majority of objects tend to be singletons.
>
> To mention something vaguely Python-relevant:
> if Python had a convenient syntax for declaring
> singletons, it could be a pretty good language for
> Interactive Fiction...
>





From trentm at ActiveState.com  Tue Dec 19 20:29:33 2000
From: trentm at ActiveState.com (Trent Mick)
Date: Tue, 19 Dec 2000 17:29:33 -0800
Subject: How to get the os.system data into a variable
In-Reply-To: <91o5ih$2kv$1@news.wrc.xerox.com>; from cabrera@wrc.xerox.com on Tue, Dec 19, 2000 at 05:23:29PM +0000
References: <91o5ih$2kv$1@news.wrc.xerox.com>
Message-ID: <20001219172933.B29727@ActiveState.com>

On Tue, Dec 19, 2000 at 05:23:29PM +0000, jicman wrote:
> I know that
> 
> t = os.system('ls -l')

o = os.popen('ls -l')
output = o.read()
o.close()

"output" now contains the output. Also look at os.popen2, os.popen3, etc for
when you want handles on stderr and stdin inaddtion to the usual stdout.

Cheers,
Trent


-- 
Trent Mick
TrentM at ActiveState.com



From pbjorn at uswest.net  Mon Dec 25 14:12:41 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Mon, 25 Dec 2000 12:12:41 -0700
Subject: Some Python 2.1 ideas
References: <925tl301ub0@news1.newsguy.com> <mailman.977590932.19335.python-list@python.org> <Cw616.23054$qN6.3308205@typhoon.we.rr.com> <3A452881.7686E717@uswest.net> <%vb16.187233$U46.5801214@news1.sttls1.wa.home.com> <3A464D5D.B541C04E@uswest.net> <mailman.977688488.25438.python-list@python.org> <20001225150250.6FA06A84F@darjeeling.zadka.site.co.il> <mailman.977760789.10224.python-list@python.org> <roy-EA8C3B.11361325122000@news.panix.com>
Message-ID: <3A479C29.2631F27B@uswest.net>

Roy Smith wrote:

> "Bob Alexander" <bobalex at home.com> wrote:
> > Finally, I'd be disappointed if the name "chomp" actually survived! I
> > just used it in my initial post as a placeholder until a better name
> > came along, since some folks would likely recall it from Perl.
>
> We've already got the string methods strip, lstrip, and rstrip.  Why not
> just add an optional second argument which is the class of characters to
> strip.  Then chomp just becomes string.rstrip (input_string, '\n').

Even better, add two arguments similar to string split (character class,
number of occurrences).

   "foo\n\n".rstrip('\n',1) -> "foo\n"

I think this is a good idea by itself and should be implemented for all of
strip, lstrip, and rstrip, however it doesn't solve the problem we want to
solve with chomp (removing a line terminator, whatever the line terminator
happens to be).

-- bjorn




From anewcombe at earthlink.net  Sat Dec  9 16:11:53 2000
From: anewcombe at earthlink.net (Amos Newcombe)
Date: Sat, 9 Dec 2000 16:11:53 -0500
Subject: dumbdbm bug fixed
Message-ID: <a05001901b6584e55adbf@[38.26.127.234]>

In dumbdbm.py (v 2.0) it says

>- seems to contain a bug when updating...

The bug I found was that the length of an item in the index (.dir) 
file doesn't get updated when the record in the .dat file does.

The fix is simple. Just call self._commit() at the end of 
_Database.__setitem__(), so it looks like this.

	def __setitem__(self, key, val):
		if not type(key) == type('') == type(val):
			raise TypeError, "keys and values must be strings"
		if not self._index.has_key(key):
			(pos, siz) = self._addval(val)
			self._addkey(key, (pos, siz))
		else:
			pos, siz = self._index[key]
			oldblocks = (siz + _BLOCKSIZE - 1) / _BLOCKSIZE
			newblocks = (len(val) + _BLOCKSIZE - 1) / _BLOCKSIZE
			if newblocks <= oldblocks:
				pos, siz = self._setval(pos, val)
				self._index[key] = pos, siz
			else:
				pos, siz = self._addval(val)
				self._index[key] = pos, siz
			self._commit()

Note that in the other branch of the main if...else statement, 
self._addkey() takes care of updating the index.

Amos Newcombe
-- 
The difference between theory and practice is that, in theory, there 
is no difference between theory and practice. 



From luthi at vaw.baug.ethz.ch  Mon Dec  4 06:50:51 2000
From: luthi at vaw.baug.ethz.ch (luthi at vaw.baug.ethz.ch)
Date: 04 Dec 2000 12:50:51 +0100
Subject: Python+Emacs+GUD ?
Message-ID: <yfmn1eciew4.fsf@collet.i-did-not-set--mail-host-address--so-shoot-me>

According to the Python archives this question has been posted 6 years
ago... still without an answer?

Q: Has someone accomplished Python integration within Emacs GDB-Mode (GUD,
   Grand Unified Debugger)?

With this, Emacs Python-mode mode would be a very handy integrated development
environment.

Thanx

Martin 


From andrer at stud.cs.uit.no  Tue Dec 12 05:03:06 2000
From: andrer at stud.cs.uit.no (=?iso-8859-1?Q?Andr=E9?= Larsen Risnes)
Date: Tue, 12 Dec 2000 11:03:06 +0100
Subject: Multiple Threads & IRC
References: <3A3415E0.1080607@bellsouth.net> <3A34E448.8B42C1BC@stud.cs.uit.no> <ku8zpm91ms.fsf@lasipalatsi.fi>
Message-ID: <3A35F7DA.8F3A5F09@stud.cs.uit.no>

Erno Kuusela wrote:
> 
> 
> well, it is from 1993. newer version is in rfc2810-rfc2813.
> 
>   -- erno

Oi. Thank you. 

alr.


From fredrik at effbot.org  Sat Dec  9 04:51:50 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sat, 09 Dec 2000 09:51:50 GMT
Subject: Dijkstra's Shortest Path algorithm
References: <Pine.GSO.4.21.0012090042370.15110-100000@y.glue.umd.edu>
Message-ID: <WgnY5.2711$O%.428005@newsc.telia.net>

Roy Katz wrote:
> For a recent assignment in my computer science class, we coded up a
> graph which implements the Shortest Path algorithm (it's a model of
> the Internet, with 'web pages' as nodes, and the latency between them
> as edges).  It took a good two or three days for me in C++, and
> I was wondering if Python (or Numerical Python) has anything like this
> (a Graph module??).  If there isn't, perhaps I should post my class at
> SourceForge and try to make an extension out of it :)

start here:

    http://www.python.org/doc/essays/graphs.html

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From BARRON.F.SNYDER at msg.ameritech.com  Mon Dec  4 12:15:29 2000
From: BARRON.F.SNYDER at msg.ameritech.com (SNYDER, BARRON F. (AIT))
Date: Mon, 4 Dec 2000 12:15:29 -0500 
Subject: Newsgroup Bridge (was - no subject)
Message-ID: <35BD410BA148D411A7ED00508BCFFBDA65FEFD@msgil65170u05.nbk2305.il.ameritech.com>

I'm fairly new here but the way I believe it works is that there is a
"bridge" between this list and the python newsgroup comp.lang.python.
Messages posted here show up there and messages posted on the newsgroup also
show up here. For some reason, the "bridge" is broken and has yet to be
fixed. Most of the traffic comes from the newsgroup and so that is why you
don't see many messages right now.

Because I'm behind a firewall, I can't read newsgroups (except with
deja.com, but that is too painfully slow) and so will be patiently waiting
for the "bridge" to be repaired. <sigh>

Barron


-----Original Message-----
From: Eric Stechmann [mailto:estechmann at abivest.com]
Sent: Monday, December 04, 2000 10:57 AM
To: python-list at python.org
Subject: (no subject)


Sorry to waste bandwidth but...

   I'm receiving little or nothing on this newsgroup.  Can anyone
   fill me in on the situation, please?

Many thanks

Eric
--------------------------------------------------------------------------
Eric Stechmann                  Direct:  +1 (651) 234-1217
Software Critter                   Fax:  +1 (651) 490-1484
American Biosystems, Inc.        E-mail:  estechmann at abivest.com
1020 West County Road F            URL:  www.abivest.com
St.Paul  MN  55126
--------------------------------------------------------------------------
The program said "Requires Windows 9* or better" so I bought a Macintosh.



From cfelling at iae.nl  Sat Dec 16 14:34:27 2000
From: cfelling at iae.nl (Carel Fellinger)
Date: 16 Dec 2000 20:34:27 +0100
Subject: Can't get Tkinter to run from command line
References: <MEM_5.1372$L3.323917@typhoon2.ba-dsg.net>
Message-ID: <91gg43$7vi$1@animus.fel.iae.nl>

Steven Sartorius <ssartor at bellatlantic.net> wrote:

you allmost got it, it's just that main is spelled differently:)

> if __name__ == 'main': gui().mainloop()

try:

  if __name__ == '__main__': gui().mainloop()

-- 
groetjes, carel


From jschmitt at vmlabs.com  Wed Dec  6 21:16:37 2000
From: jschmitt at vmlabs.com (jschmitt at vmlabs.com)
Date: Thu, 07 Dec 2000 02:16:37 GMT
Subject: UnboundLocalError: Local variable 'Tee' referenced before assignment
Message-ID: <90mru0$cv4$1@nnrp1.deja.com>

I'm getting an error message that I don't understand.  I'm using this
handy class that someone kindly on comp.lang.python:

class Tee:
    """
    Think of the unix command-line utility
    """
    def __init__( self, *fileobjects ):
        self.fileobjects = fileobjects

    def write( self, string ):
        for fileobject in self.fileobjects:
            fileobject.write( string )

and I'm using it kind of like this:
reportfile = open( os.path.join( workingdir, "report.txt" ), "w" )
sys.stdout = Tee( reportfile, sys.stdout )    # THIS LINE IS THE PROBLEM
# do some print's here...
del Tee
reportfile.close()

but I get this error message:

UnboundLocalError: Local variable 'Tee' referenced before assignment

Does anyone have a spare clue?  It works the way I expect when I use it
interactively in the python shell.

John


Sent via Deja.com http://www.deja.com/
Before you buy.


From geoff at elj.com  Sat Dec 30 21:00:38 2000
From: geoff at elj.com (geoff at elj.com)
Date: Sun, 31 Dec 2000 02:00:38 GMT
Subject: elj-win32-dev: project summary (31 Dec 2000)
Message-ID: <92m404$8pb$1@nnrp1.deja.com>

   -- posted to comp.lang.python since it was the Python win32
   -- book that inspired this project in SmallEiffel.

Just an end of year update for the elj-win32-dev project:

   http://www.egroups.com/group/elj-win32-dev/

Some useful starting points:

   http://elj-win32.sourceforge.net/
   http://efsa.sourceforge.net/cgi-bin/view/Main/EljWin32All

The current status and some background was recently summarised by Uwe
Sander - usander at online.de - on our mailing list:

   http://www.egroups.com/message/elj-win32-dev/401     or
   http://www.elj.com/elj.cgi?article=20001231a  (has some links + gif)

--8<--
Okay, a short history, our current state and what we plan to do:

 * Geoff started the project last summer with the intention to give
   [Small]eiffel programmers full access to the win32 api. He called
   out for assistence and Birgit and me joined the team.

 * A few months later, we started a a new subproject, called ccl
   (component and class library), as an abstract layer over the pure api
   calls.

The idea is, to isolate the api calls in a way that they can be
exchanged to gain independence from a special os but have them (the
api calls) still available if a programmer needs them for special
purposes. Why we are going this way? There are two reasons for it:

 * We don't want to maintain two libraries, a c based one and an
   eiffel based one on top of the c code. Yes, we have some c code
   in it, but the functions are mostly simple 1 to 1 mappers and
   for this fact nearly bug free right from the start.

 * We found, that in real world application development it is very
   useful to have the possibillty to invoke api calls if there is
   nothing in the library which fulfills your needs. ccl is not
   finished, but it supports a form class, all windows standard
   controls and some of the common controls of the windows api now.

The next subproject, started a few weeks ago, is kk, which is in a very
early state. This is the two way design tool on the top of the ccl.

What we are doing now:

 * Birgit is improving the win32 api calls in a way that we need much
   lesser c code.

 * Geoff is coding lots of examples for win32 api programming with
   eiffel, collecting tons of information resources for further
   development and providing some help with ccl and kk. He is a
   very busy man ;-)

 * Sven and me are currently rebuilding the streaming mechanism in the
   heart of the ccl, so that xml becomes the standard stream format for
   exporting and importing component properties. This will mostly be
   used from the visual designer of kk.

What we plan:

 * Finishing the windows controls encapsulation of ccl,

 * building an abstract database access layer for the ccl,

 * developing of data sensitive controls and finally, after further
   development of kk,

 * porting the whole thing to linux.

If we don't find more helping hands, I guess we need 6 - 9 months
until we can start the port to linux.

I hope I have answered your questions.
-->8--

Since this message Uwe has added an interface to the Postgresql database
which will be added to a CVS project in the near future.

As you can see we have come a long way, but there is still a long way to
go.  To subscribe to the elj-win32-dev list:

   http://www.egroups.com/group/elj-win32-dev/

just send a message to elj-win32-dev-subscribe at egroups.com

Happy New Year to all ..

Regards .. Geoff

-- geoff at elj.com
-- http://www.elj.com/elj-daily.cgi


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


From bww00 at amdahl.com  Wed Dec  6 10:25:59 2000
From: bww00 at amdahl.com (Bryan Webb)
Date: 06 Dec 2000 15:25:59 GMT
Subject: Help/example of bit testing
Message-ID: <90llq7$8fc@dispatch.concentric.net>

Hi,
    I am reading a file (binary)of 1024 char recs (bitmaps) . I can read the
file , bit im having trouble finding examples of testing bits. I need to
look at each bit of the above record.
Looking for examples and ideas.

Thanks
Bryan Webb




From dhumphrey at acm.org  Wed Dec 20 16:15:34 2000
From: dhumphrey at acm.org (David L. Humphrey)
Date: Wed, 20 Dec 2000 15:15:34 -0600
Subject: Best documentation for writing wrappers for C++ classes
Message-ID: <9EB39E4E3AF64120.C259930B9E4BEFF5.1C1A4B99EAA59011@lp.airnews.net>

I have several C++ classes that I would like to call from Python code.  I
have reviewed the "Extending and Embedding the Python Interpreter" as well
as selected portions of "Programming Python" but worry a little about the
currency of these sources as well as the general tone of extending with C
code rather than C++ classes.  Is there a document that is recognized as the
standard reference for this information?

Thanks in advance,
David L. Humphrey
Manager, Software Development
Bell Geospace, Inc




From aleaxit at yahoo.com  Mon Dec 18 07:06:27 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 18 Dec 2000 13:06:27 +0100
Subject: Database solution for Mac?
References: <91kjtq$ns3$1@news1-2.kornet.net>
Message-ID: <91kul3029b5@news1.newsguy.com>

"Ahn, Jae-wook" <mail at codex.pe.kr> wrote in message
news:91kjtq$ns3$1 at news1-2.kornet.net...
> Hello,
>
> I have an iBook. I usually do works on Linux or Windows, but sometimes I
> need some database support for my small Mac machine.
>
> I visited some pages, and found out a lot of database support for Python
> including ODBC, but could not see anything that supports Macintosh. Please
> recommend any solution which support database for Mac. Networked or local.

I think Gadfly should build and run happily on a Mac; it's an
SQL relational DBMS that is implemented in Python (with optional
C in you need to speed up certain things), see:

http://www.chordate.com/kwParsing/gadfly.html


Alex





From grante at visi.com  Tue Dec 12 07:53:04 2000
From: grante at visi.com (Grant Edwards)
Date: 12 Dec 2000 06:53:04 -0600
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Dec 12)
Message-ID: <98A74E99845A9CC4.F233FB7345AE8DED.EF9894A1BC63F754@lp.airnews.net>

    Alex Martelli discusses mutual dependencies between modules.
       http://deja.com/=dnc/getdoc.xp?AN=701015113
    
    Alex also summaries the Python answer to the lack of "super"
    question.
       http://deja.com/=dnc/getdoc.xp?AN=701004958
    
    Alex Shindich converts Weak Reference Extension Module to "C"
       http://deja.com/=dnc/getdoc.xp?AN=701250836
    
    Alex Martelli answers the question (regarding sorting database
    records) "What is a Schwartzian transform?"
       http://deja.com/=dnc/getdoc.xp?AN=701836861
    
    Jeff Epler introduces PyHTML -- an extension of the Python
    grammar, plus a modified compiler which permits the use of
    HTML-style tags as Python blocks, making it easy to write HTML
    document generators with the block-indentation style of Python.
       http://deja.com/=dnc/getdoc.xp?AN=702170599
    
    Laura Hill posts OOPSLA 2001 Call for Participation
       http://deja.com/=dnc/getdoc.xp?AN=702782743


========================================================================

Everything you want is probably one or two clicks away in these pages:

      Python.org's Python Language Website is the traditional
      center of Pythonia
          http://www.python.org
      although PythonLabs.com bills itself as "The Python Source",
      which is becoming increasingly true in 2000
          http://www.pythonlabs.com/

      PythonWare complements the digest you're reading with the
      daily python url
           http://www.pythonware.com/daily  

      comp.lang.python.announce announces new Python software.  Be
      sure to scan this newly-revitalized newsgroup at least weekly.
          http://deja.com/group/comp.lang.python.announce

      Andrew Kuchling writes marvelous summaries twice a month of
      the action on the python-dev mailing list, where the future
      of Python is truly determined
	  http://www.amk.ca/python/dev

      The Vaults of Parnassus ambitiously collects Python resources
          http://www.vex.net/~x/parnassus/   

      Much of Python's real work takes place on Special-Interest Group
      mailing lists
          http://www.python.org/sigs/

      Python Consortium emerges as an independent nexus of activity
          http://www.python.org/consortium

      Cetus does much of the same
          http://www.cetus-links.de/oo_python.html

      Python FAQTS
          http://python.faqts.com/

      Python To-Do List anticipates some of Python's future direction
          http://www.python.org/cgi-bin/todo.py

      Python Journal is at work on its second issue
          http://www.pythonjournal.com

      Links2Go is a new semi-automated link collection; it's impressive
      what AI can generate
          http://www.links2go.com/search?search=python

      Archive probing trick of the trade:
          http://www.dejanews.com/dnquery.xp?QRY=&DBS=2&ST=PS&defaultOp=AND&LNG=ALL&format=threaded&showsort=date&maxhits=100&groups=comp.lang.python


Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://purl.org/thecliff/python/url.html
or
  http://www.dejanews.com/dnquery.xp?QRY=~g%20comp.lang.python%20Python-URL%21


Suggestions/corrections for next week's posting are always welcome.
[http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
e-mail us ideas directly.]

To receive a new issue of this posting in e-mail each Monday morning,
ask <claird at neosoft.com> to subscribe.  Mention "Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.
-- 

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


From slhath at home.com  Tue Dec 19 20:02:32 2000
From: slhath at home.com (Scott Hathaway)
Date: Wed, 20 Dec 2000 01:02:32 GMT
Subject: xml-rpc module question
References: <dmC%5.75166$yR4.2151642@news1.rdc1.tx.home.com> <91ooea$9qb$1@bunyip.cc.uq.edu.au> <DeT%5.76385$yR4.2212474@news1.rdc1.tx.home.com> <nlT%5.653$Qb7.44417@newsb.telia.net>
Message-ID: <IyT%5.76392$yR4.2213736@news1.rdc1.tx.home.com>

Thanks!  It's alive.

:)

Scott

"Fredrik Lundh" <fredrik at effbot.org> wrote in message
news:nlT%5.653$Qb7.44417 at newsb.telia.net...
> Scott Hathaway wrote:
> > Thanks for the response.  When I try your suggested code, here is the
error
> > I get:
> >
> > C:\Projects\core>python core_client.py
> > Traceback (most recent call last):
> >   File "core_client.py", line 5, in ?
> >     print testing.CubeIt(3)
> >   File "xmlrpclib.py", line 547, in __call__
> >     return self.__send(self.__name, args)
> >   File "xmlrpclib.py", line 627, in __request
> >     response = self.__transport.request(
> > AttributeError: 'int' object has no attribute 'request'
>
> you're passing in a bogus second argument to the xmlrpclib
> Server constructor.  don't do that.
>
> </F>
>
>




From dsh8290 at rit.edu  Thu Dec 21 08:50:32 2000
From: dsh8290 at rit.edu (D-Man)
Date: Thu, 21 Dec 2000 08:50:32 -0500
Subject: Editing files
In-Reply-To: <Bdd06.5591$K6.674078@juliett.dax.net>; from lars@stea.nu on Thu, Dec 21, 2000 at 01:41:53AM +0000
References: <Bdd06.5591$K6.674078@juliett.dax.net>
Message-ID: <20001221085032.A1673@westsidecnc.rh.rit.edu>

You probably want to use seek() and tell().  tell will return the
current location in a file and seek will go to a particular location
(it's argument is the same format as the return value of tell).

I have never had a use for them myself so I don't know if there are
any issues such as overrunning the bounds of what you are replacing.
(ie if the new text is longer than the old)

-D

On Thu, Dec 21, 2000 at 01:41:53AM +0000, dv wrote:
> I'm trying to automate some configuration tasks with python. Is it possible
> to do a "search and replace" or specify where (which line) in a file python
> will write?
> 
> Thanks in advance
> 
> Lars
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list



From sjoerd at oratrix.nl  Mon Dec 11 09:09:28 2000
From: sjoerd at oratrix.nl (Sjoerd Mullender)
Date: Mon, 11 Dec 2000 15:09:28 +0100
Subject: Validating xml files
In-Reply-To: Your message of Mon, 11 Dec 2000 13:48:38 +0100.
             <3A34CD26.DDE8F341@sicon.net> 
References: <3A34CD26.DDE8F341@sicon.net> 
Message-ID: <20001211140928.AF2A631047C@bireme.oratrix.nl>

I have a DTD validator that is still a work-in-progress, but it
behaves pretty well already.  It's slow though.  I can mail it to you
(and anybody else who wants it).

On Mon, Dec 11 2000 Sergio Boix Moriano wrote:

>     Hi all,
> 
>         I?m using the PyXML-0.5.5.1 library to work with xml files. We
> generate the xnl files from a database and we need to validate this
> files cause later we have to do an insert in another dataabse with the
> xml contents, and the file must only have fields and tables that exits
> in the database.
> 
>        I?ve tried to use the DTDParser() class, but lokking at the code
> i?ve seen that the functions in the class doesn?t do nothing. The code
> of this classes in the saxlib module is:
> class DTDHandler:
>     def notationDecl(self, name, publicId, systemId):
>         "Handle a notation declaration event."
>         pass
> 
>     def unparsedEntityDecl(self, name, publicId, systemId, ndata):
>         "Handle an unparsed entity declaration event."
>         pass
> 
> 
>     Does anyone know anyway to read an xml file validating it with a
> DTD, in a way that if the xml is bad formed it gives me an error.
> 
>     Thanks in advance,
> 
> 
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 

-- Sjoerd Mullender <sjoerd.mullender at oratrix.com>



From peter at engcorp.com  Tue Dec 12 23:41:18 2000
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 12 Dec 2000 23:41:18 -0500
Subject: Serial IO for Python 2.0???
References: <9166jh$5hh$1@nnrp1.deja.com>
Message-ID: <3A36FDEE.D19B67B1@engcorp.com>

harrc at my-deja.com wrote:
> 
> Does anyone know of a Win32 serial port I/O solution for Python 2.0?
> Both win32comm and the sio module are compiled for Python 1.5.

Python 2.0 (with Mark Hammond's win32all that is) now has the equivalent
support to win32comm built into the win32file module.  (Thanks for that
pointer before, Mark! :)

Slight differences but it only took about four lines of code to migrate
my old support (not yet released, sorry) from win32comm in 1.5.2 to
win32file in 2.0.


From emile at fenx.com  Wed Dec 13 10:42:11 2000
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 13 Dec 2000 07:42:11 -0800
Subject: Q: looping with two counter variables through a list
References: <yv2zoi0ojef.fsf@lionsp093.lion-ag.de>
Message-ID: <3A3798D3.1020806@fenx.com>

How about:

for i in range(0,len(mylist),2):
     a,b = mylist[i], mylist[i+1]

with approprite trapping for short lists.

Emile van Sebille
emile at fenx.com

Harald Kirsch wrote:

> In Tcl I can write
> 
>   set mylist {0 1 2 3 4 5}
>   foreach {a b} $mylist {
>     puts $a
>   }
> 
> to print the even numbers of $mylist. Obviously
> 
>    mylist = [0, 1, 2, 3, 4, 5]
>    for a,b in mylist:
>      print a
> 
> does not work (ValueError: unpack sequence of wrong size). 
> 
> What is Python's idiom for running with more than one variable through
> a flat list?
> 
>   Harald Kirsch



From uche at ogbuji.net  Thu Dec 21 11:01:06 2000
From: uche at ogbuji.net (Uche Ogbuji)
Date: Thu, 21 Dec 2000 16:01:06 GMT
Subject: missing HtmlBuilder
References: <91lsh8$9p7$1@nnrp1.deja.com> <91npvu018ie@news2.newsguy.com>
Message-ID: <3A42293F.7DC5AF5F@ogbuji.net>

Alex Martelli wrote:
> 
> <ennisr at my-deja.com> wrote in message news:91lsh8$9p7$1 at nnrp1.deja.com...
> > I am a relative newbie to Python, using Python 2.0 under Windows.
> > 4DOM from Fourthought, Inc. has also been installed.
> >
> > I have run into a problem trying to run the DOM examples from David
> > Mertz article "Charming Python: The dynamics of DOM"
> > (http://gnosis.cx/publish/programming/charming_python_2.txt).
> >
> > Modules referenced in the following import statements seem to be
> > missing.
> >
> >     from xml.dom import core
> >     from xml.dom.html_builder import HtmlBuilder
> >     from xml.dom.utils import FileReader
> >
> > Do I have the wrong software loaded?  Where can I find the missing
> > modules?
> 
> I think you need the PyXml stuff, e.g. from
>     http://sourceforge.net/projects/pyxml
> probably in some older version.
> 
> 4DOM may have equivalent or better functionality, but
> I suspect it will not match what you read in that
> online book, which appears to target older PyXml stuff.

Actually, PyXML now incoporates 4DOM as its DOM library.  The old code
mentioned above is no longer maintained by anyone.

Here is code given in the article

import sys
from xml.dom import core
from xml.dom.html_builder import HtmlBuilder

# Construct an HtmlBuilder object and feed the data to it
b = HtmlBuilder()
b.feed(sys.stdin.read())

# Get the newly-constructed document object
doc = b.document

# Output it as XML
print doc.toxml()


And here is how it would look in PyXML or 4DOM

import sys
from xml.dom import ext
from xml.dom.ext.reader import HtmlLib

#Create a reusable reader object
reader = HtmlLib.Reader()

#Construct a DOM from stdin
doc = reader.fromStream(sys.stdin)

#Output the HTML as XHTML
ext.XHtmlPrettyPrint(doc)


-- 
Uche Ogbuji
Personal:   uche at ogbuji.net		http://uche.ogbuji.net
Work:       uche.ogbuji at fourthought.com	http://Fourthought.com


From scj at mcom.mcom.fr  Wed Dec  6 04:07:57 2000
From: scj at mcom.mcom.fr (Alfred)
Date: Wed, 06 Dec 2000 09:07:57 GMT
Subject: Sort
Message-ID: <NlnX5.1444$yv1.2700363@nnrp5.proxad.net>

Hello world,

I'm trying to sort a List of List, on the third col, but I don't find how to
do this.

Please help me. ;)

Greg




From aleaxit at yahoo.com  Thu Dec 28 05:10:31 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 28 Dec 2000 11:10:31 +0100
Subject: Python vs PHP (& Java?)
References: <mailman.977897649.4877.python-list@python.org> <slrn94l86m.a3h.tim@degree.ath.cx>
Message-ID: <92f3kp01cqe@news1.newsguy.com>

"Tim Hammerquist" <tim at degree.ath.cx> wrote in message
news:slrn94l86m.a3h.tim at degree.ath.cx...
> Ben Ocean <zope at TheWebsons.com> wrote:
> > I'm an intermediate-level programmer that's new to Python (via Zope) and
I
> > like it a lot. However, it appears to me that, while Python is a good
> > replacement for Perl, it cannot replace much of the functionality of
PHP,
> > which can be scripted into HTML (whereas Python, it appears, must be
> > called).
>
> Ok, everyone else made very good points, but this little comment slipped
> out (not that I'm surprised, honestly)  =)   There is no "replacement
> for Perl!"  And Python would do a poor job trying to be Perl (and vice
> versa), nor does anyone here seem to _want_ Python to be/replace Perl.

Not sure what you mean here.  Over my first 2-3 months of Python,
I rewrote _every_ Perl script on my machine in Python, and the gains
in clarity, reliability, and maintainability, were outstanding.

The 'good-applicability niches' of Perl and Python are vastly
overlapped (and not so small as to really deserved being called
'niches', but that's another issue).  I can't think of any program
I would rather write in Perl than in Python, though I guess I
might *have* to, if forced (by somebody pointing a loaded gun at
my head with its safety off, for example).  So, for me, Python HAS
indeed served as "a good replacement for Perl" -- everything I used
to do in Perl, I now do in Python instead; what other definition
of "replacement" are you using instead?


Alex





From wade at lightlink.com  Thu Dec 14 11:49:33 2000
From: wade at lightlink.com (wade at lightlink.com)
Date: Thu, 14 Dec 2000 16:49:33 GMT
Subject: codecs UTF-8 StreamReader ignores errors arg
References: <90imhq$t5m$1@nnrp1.deja.com> <slrn92tsno.jm.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com>
Message-ID: <91atmp$mt$1@nnrp1.deja.com>

In article <slrn92tsno.jm.amk at 207-172-111-
60.s60.tnt1.ann.va.dialup.rcn.com>,
  akuchlin at mems-exchange.org wrote:
> An aside: Wade, did you look at Lib/codecs.py to find the cause?  I
> interpret your message as meaning you did, since your suggested cause
> is quite correct.  There's no need to be shy about saying that this is
> a bug!
>
> --amk
>
It's just so embarrassing to say you've found a bug, then find out you
made a silly mistake!




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


From moshez at zadka.site.co.il  Sun Dec 31 17:49:25 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Mon,  1 Jan 2001 00:49:25 +0200 (IST)
Subject: why no "do : until"?
In-Reply-To: <3A4E95FD.D6C737A7@engcorp.com>
References: <3A4E95FD.D6C737A7@engcorp.com>, <92jj3j$ff7$1@nnrp1.deja.com> <slrn94shbe.150.grey@teleute.rpglink.com> <X8s36.2731$Qb7.305339@newsb.telia.net> <slrn94smds.150.grey@teleute.rpglink.com> <FOt36.17$vd1.2224@e420r-atl2.usenetserver.com> <slrn94t1vj.150.grey@teleute.rpglink.com>
Message-ID: <20001231224925.399F9A84F@darjeeling.zadka.site.co.il>

On Sat, 30 Dec 2000 21:12:13 -0500, Peter Hansen <peter at engcorp.com> wrote:

> The advice that runs
> contrary to your 10+ years seems very reasonable to my 20+ years of
> programming experience, and apparently many others with a variety of
> experiences agree.  

oooooh, oooooh, my dick is longer.
I've been programming with it for, like, 100+ yrs. I haven't really
got an opinion about /F's advice -- I just wanted to show off.

contributing-to-dead-threads-ly y'rs, Z. 
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From loewis at informatik.hu-berlin.de  Fri Dec  8 15:57:12 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 08 Dec 2000 21:57:12 +0100
Subject: Pickles Cross-platform?
References: <3A313570.A71966A@qvlinc.com>
Message-ID: <p6qofymiqc7.fsf@informatik.hu-berlin.de>

Coy Krill <ckrill at qvlinc.com> writes:

> I seem to have to run 'dos2unix' past any pickles that I want to read
> that originated on our Windows machines. Is this typical?  For some
> reason I was under the impression that the unpickler handled the LF vs
> CRLF contortions for you but it doesn't seem to be doing that.

There shouldn't be any CR in a pickled file; just make sure you always
open files in binary ('b') mode.

Regards,
Martin


From robin at jessikat.fsnet.co.uk  Sat Dec  9 10:37:40 2000
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sat, 9 Dec 2000 15:37:40 +0000
Subject: sgmllib nit
Message-ID: <eyErVLAEHlM6EwA4@jessikat.fsnet.co.uk>

The current version of sgmllib doesn't allow the detection of attributes
without values. Instead the attribute name is used. Later the pair
(string.lower(attrname),attrvalue) is appended to the list of
attributes. The current code is
            
            attrname, rest, attrvalue = match.group(1, 2, 3)
            if not rest:
                attrvalue = attrname
            elif attrvalue[:1] == '\'' == attrvalue[-1:] or \
                 attrvalue[:1] == '"' == attrvalue[-1:]:
                attrvalue = attrvalue[1:-1]
            attrs.append((string.lower(attrname), attrvalue))


if this were changed to do the string lower first then one could use the
fact that the name, value pair both have the same id to detect the fact
that no = value was entered.

            attrname, rest, attrvalue = match.group(1, 2, 3)
            attrname = string.lower(attrname)
            if not rest:
                attrvalue = attrname
            elif attrvalue[:1] == '\'' == attrvalue[-1:] or \
                 attrvalue[:1] == '"' == attrvalue[-1:]:
                attrvalue = attrvalue[1:-1]
            attrs.append((attrname, attrvalue))
-- 
Robin Becker


From grey at despair.rpglink.com  Fri Dec 29 14:23:34 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Fri, 29 Dec 2000 19:23:34 -0000
Subject: Great New Book
References: <t4o1vr8hktk531@corp.supernews.com> <92ij47$n3p$1@bob.news.rcn.net> <slrn94pnrk.irm.grey@teleute.rpglink.com> <92inl3$a5o$1@panix2.panix.com>
Message-ID: <slrn94pp5m.irm.grey@teleute.rpglink.com>

On 29 Dec 2000 11:11:31 -0800, Aahz Maruch <aahz at panix.com> wrote:
>In article <slrn94pnrk.irm.grey at teleute.rpglink.com>,
>Steve Lamb <morpheus at here.not.there> wrote:
>>    Not to mention 2'5" isn't all that thick.  I have fictional paperbacks
>>that are that thick.  :)

>Really?  You have paperbacks more than two feet thick????  (No fair
>counting all the Robert Jordans as a single book.)

    Touche'.  2.5".  :)


-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From peter at engcorp.com  Tue Dec 12 23:30:14 2000
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 12 Dec 2000 23:30:14 -0500
Subject: threads and dictionaries
References: <9139fq$ovj$1@nnrp1.deja.com>
Message-ID: <3A36FB56.EC26D2C5@engcorp.com>

bartig at pinpoint.com wrote:
> 
> I'm writing a program that passes dictionaries into a subroutine that's
> threaded.  Are dictionaries thread safe?  If so how would I go about
> making changes to the dictionaries inside of a thread so that other
> threads can take advantage this?

Following on Aahz's answer: it's not so much a question of whether
*dictionaries* are safe, as it is a question of whether the operation(s)
you intended to perform on a dictionary is (are) safe.  Multiple
operations are of course not safe without proper protection.  In Python
many simple operations on basic data types are safe.  Anything specific
in mind?


From thomas at xs4all.net  Sun Dec 31 11:59:40 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Sun, 31 Dec 2000 17:59:40 +0100
Subject: 2001 Enchancement Wishlist
In-Reply-To: <92j83r13gh@news1.newsguy.com>; from aleaxit@yahoo.com on Sat, Dec 30, 2000 at 12:49:28AM +0100
References: <3A4CEACB.220998A0@javanet.com> <Ro736.2511$Qb7.273102@newsb.telia.net> <92j83r13gh@news1.newsguy.com>
Message-ID: <20001231175940.A5435@xs4all.nl>

On Sat, Dec 30, 2000 at 12:49:28AM +0100, Alex Martelli wrote:
> "Fredrik Lundh" <fredrik at effbot.org> wrote in message
> news:Ro736.2511$Qb7.273102 at newsb.telia.net...
>     [snip]
> > (fwiw, singleton classes are pretty unpythonic too.  python
> 
> Amen, Hallelujah.
> 
> > already has a framework for creating exactly one instance of
> > something: the module namespace).
> 
> Pity that isn't as flexible as an object-instance -- can't define
> special methods to determine how it will print, how it will
> let usercode loop on its items, etc.

Of course it can.

<spammodule.py>
class _singleton:
	<whatever you want>

SingleTon = _singleton()

All that is required is for everyone to use spammodule.SingleTon.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From tim.one at home.com  Sun Dec 17 18:21:46 2000
From: tim.one at home.com (Tim Peters)
Date: Sun, 17 Dec 2000 18:21:46 -0500
Subject: Migration test
Message-ID: <LNBBLJKPBEHFEDALKOLCOEFEIEAA.tim.one@home.com>

Ignore me!

business-as-usual-ly y'rs  - tim




From emile at fenx.com  Tue Dec  5 19:00:19 2000
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 05 Dec 2000 16:00:19 -0800
Subject: an apply question
References: <90jsj0$24o$1@front6.grolier.fr>
Message-ID: <3A2D8193.3070401@fenx.com>

Try it without the quotes around the functions.

eg: 'Key1' = func1

Emile van Sebille
emile at fenx.com



Alan Marchand wrote:

> In Python 2.0:
> 
> What am I doing wrong?
> 
> #!/bin/env python
> #############
> MyMap = {
> 'Key1' : 'func1',
> 'Key2' : 'func2',
> 'Key3' : 'func3',
> 'Key4' : 'func4'
> }
> 
> def func1 ( ):
>     print "Here's func1"
> 
> items = MyMap.items( )
> for x in range( len(items) ):
>     key, func = items[x]
>     if key == "Key3":
>         apply( func, ( ) )
> -------------------------
> c:\devpy\python test.py
> 
> Traceback (most recent call last):
>   File "test.py", line 17, in ?
>     apply( func, ( ) )
> TypeError: Call of non-function (type string)
> 
> Thanks.
> Alan.



From mwh21 at cam.ac.uk  Fri Dec 15 06:06:46 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 15 Dec 2000 11:06:46 +0000
Subject: Getting a code obj's function?
References: <91co73$fpm$1@news.rz.uni-karlsruhe.de>
Message-ID: <m3k892ose1.fsf@atrus.jesus.cam.ac.uk>

Bjoern Giesler <un4e at rzstud1.rz.uni-karlsruhe.de> writes:

> Hi,
> 
> another inane question...
> 
> ...is there a standard way to get the function object that a code object is
> associated with? 

Well, by mucking with the new module, you can get two functions with
the same code object.

Also it's possible to execute a code object without an associated
function - like the top-level of a module for example.

> If there is no standard way, could somebody please comment on the
> following: I've modified my Python interpreter so that a code object
> has one more member, co_func.  This gets set to None initially, and
> PyFunc_New() sets it to the new function object that is created.
> What are the potential problems with this approach, if any?  I see
> all kinds of memory leaks / double frees coming, but haven't really
> enough understanding of Python to know what might happen.  My
> assumption was that a code object is assigned to a maximum of
> exactly one function. Is this correct?

I think you can probably come up with something that will work most of
the time, but not all the time.  Whether that's good enough for you is
up to you...

Cheers,
M.

-- 
7. It is easier to write an incorrect program than understand a
   correct one.
  -- Alan Perlis, http://www.cs.yale.edu/homes/perlis-alan/quotes.html


From joconnor at cybermesa.com  Tue Dec 19 22:39:21 2000
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Wed, 20 Dec 2000 03:39:21 +0000
Subject: while (thread.alive()):
Message-ID: <3A4029E9.1EE53250@cybermesa.com>

>The fact that it's taken about three attempts by
>experienced Python people to come up with a full
>and accurate characterisation of what constitute
>'true' and 'false' in Python suggests that things 
>could have been a lot simpler if there had been a 
>dedicated boolean type!

Believe it or not, both Ada and Smalltalk have boolean types.

Both languages were designed in the 70s, commercially implemented in the
80s; one is dynamically typed, one is statically typed.

Why are we still maintaining these leftover idioms such as using
integers as booleans.  If we are to carry the state of the art of
programming beyond it's current level, we need to be programming a t a
higher level of abstraction than is easily supported while still
dragging anachronisms such as this around.

Those who claim that 'while 1:' is obvious, it's not.

It's only obvious if you've programmed "C" and other such languages. 
It's makes no sense really, unless you've learned the 'code

My first professional language was Ada83.  From there I went to
Smalltalk.  When I finally started working with languages such as C,
C++, etc... I was at first confused ( while (1)...what does that mean)
and then revolted.

I'm continuously amused that people try to take language design forward
and still drag obsolete ideas along for the ride.

Python's got a lot of cool stuff, but the lack of a real Boolean object
is a glaring omission

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

"God himself plays the bass strings first when He tunes the soul"



From moshez at zadka.site.co.il  Sun Dec 31 09:56:59 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Sun, 31 Dec 2000 16:56:59 +0200 (IST)
Subject: Python vs PHP (& Java?)
In-Reply-To: <20001230054219.A21277@glacier.fnational.com>
References: <20001230054219.A21277@glacier.fnational.com>, <92f3kp01cqe@news1.newsguy.com>, <mailman.977897649.4877.python-list@python.org> <slrn94l86m.a3h.tim@degree.ath.cx> <92f3kp01cqe@news1.newsguy.com> <20001228162203.7163.qmail@stimpy.scso.com>
Message-ID: <20001231145659.3E7B2A84F@darjeeling.zadka.site.co.il>

[Moshe Zadka]
> (BTW: Are there plans for quixote/mod_snake integration? That'd
> be ultra-cool: one of the benefits of HTML::Mason is that it
> integrates really well with mod_perl)

[Neil Schemenauer]
> What would be the advantage of that over Quixote/FCGI?

For one, the mod_fastcgi license is sucky. I'm not sure how FCGI works,
but staying within the Apache process gives all you all kinds of things
for free -- keeping up specified number of servers, shutting down
a server after a specified number of requests, etc.

-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From aleaxit at yahoo.com  Thu Dec 28 09:12:57 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 28 Dec 2000 15:12:57 +0100
Subject: Interesting behaviour of the assignment
References: <92f4gk$qk6$2@news.nuri.net>
Message-ID: <92fhsr01u88@news1.newsguy.com>

"June Kim" <junaftnoon at nospamplzyahoo.com> wrote in message
news:92f4gk$qk6$2 at news.nuri.net...
> >>> a=100
> >>> b=100
> >>> a is b
> 0
> >>> a=99
> >>> b=99
> >>> a is b
> 1
>
> Is this an intended/expected result? What is the lying structure beneath
> this?

The current Python interpreter uses a special-trick (just
simple caching, actually) for the representation of small
integers, up to a certain threshold.  This just shows the
threshold is now 100 (excluded).

A similar case:

>>> y="aufwiedersehen"
>>> y is "aufwiedersehen"
1
>>> z="arrivederci Roma, good-bye, au revoir!"
>>> z is "arrivederci Roma, good-bye, au revoir!"
0
>>>

Here, the special-trick is "intern'ing", not that it makes
much difference to the application-coder except that intern
IS exposed and available for (doubtful!) application code
direct use...:

>>> intern
<built-in function intern>
>>> z=intern("arrivederci Roma, good-bye, au revoir!")
>>> z is "arrivederci Roma, good-bye, au revoir!"
0
>>> z is intern("arrivederci Roma, good-bye, au revoir!")
1
>>> z
'arrivederci Roma, good-bye, au revoir!'
>>>


I think the correct level of awareness of these internals
issues is the following: 'is' is ``not fully reliable''
(or rather, not fully applicable from an application POV)
when the objects involved may be small integers or string
constants, in the following specific sense: a!=b ensures
not "a is b", but, if a==b, then whether "a is b" or not
depends on internals/implementation issues that you don't
really want to know about.

Thus, limit your use of 'is' to cases in which you know
"object-identity" is indeed meaningful: None, lists, tuples,
dictionaries, object instances, type-objects, etc.
(Actually, in the case of None, the semantics of is and
== are identical, but 'is' can be a tiny little bit faster).

In practice I find myself using == almost by reflex, e.g.
    if type(x)==type(0):
when I KNOW I could safely and reliably write, instead:
    if type(x) is type(0):
and perhaps gain a 1/1000th of a second if this is at
the heart of a heavily-executed loop... '==' is just a
by-far-too-more-frequent case, except perhaps in the
specific idion "x is None"!-)


Alex





From davide at mercurio.localdomain  Fri Dec 29 10:48:07 2000
From: davide at mercurio.localdomain (davide at mercurio.localdomain)
Date: Fri, 29 Dec 2000 16:48:07 +0100
Subject: split hangs
In-Reply-To: <mailman.977998605.7293.python-list@python.org>
References: <mailman.977998605.7293.python-list@python.org>
Message-ID: <14C1lH-44b-00@mercurio.localdomain>

In comp.lang.python, you wrote:

>The tail end of the output is clear:  adding another character to the string
>doubles the time it takes to figure out that the regexp doesn't match.  This
>isn't a bug, it's the way Python (and Perl, and Emacs, and ...) regexps
>work.  A full explanation is quite involved; you can find one in Jonathan
>Friedl's excellent book "Mastering Regular Expressions" (O'Reilly).

I tried the equivalent regexp in emacs
  <\([^<>"]+\("[^"]*"\)?\)+>
with bigger and bigger files and it does not seem to give an exponential
behaviour.

In everything I read regexp matching is given as a linear time problem in
n + m (where n is the size of the testing string and m the size of the
regular expression).

I am a python newbye and I could not find the reason for this exp 
behaviour. The regular expressions seems pretty standard to me, 
and so there should be no need to go beyond regular languages and DFA. 

Would you explain why this algorithm is needed in python?
Do really the other implementation (grep, perl, emacs, ecc...)
use the same algorithm?

Thanks,

				Davide



From trotts at llnl.gov  Sat Dec 23 00:53:34 2000
From: trotts at llnl.gov (Issac Trotts)
Date: Fri, 22 Dec 2000 21:53:34 -0800
Subject: coding convention conversion
In-Reply-To: <oSV06.176$g03.67749@sapphire.mtt.net>; from bruce_dodson@bigfoot.com on Sat, Dec 23, 2000 at 04:29:08AM +0000
References: <20001222122015.A7987@llnl.gov> <mailman.977522226.5104.python-list@python.org> <oSV06.176$g03.67749@sapphire.mtt.net>
Message-ID: <20001222215334.B10077@llnl.gov>

OK, that should do the trick for python.  Any ideas about how to 
do it for C++ (using python, so that this isn't off-topic ;).

Issac


On Sat, Dec 23, 2000 at 04:29:08AM +0000, Bruce Dodson wrote:
> Better yet, import the file as a module, and scan through its public symbols
> (via dir()) to find out which names need to be changed, and then use that
> information to build a sed script which can be applied globally.  You can
> easily take advantage of other meta-information such as the type, e.g. limit
> it to just functions if you want.  This approach is good because the parser
> will recognize more variations than a regexp.
> 
> 
> Something like this is what I had in mind:
> 
> from types import ModuleType, ClassType
> 
> def print_symbols(symbols, prefix=''):
>     for sym in dir(symbols):
>         if sym[:2] == '__': continue
> 
>         v = getattr(symbols, sym)
>         if type(v) is ModuleType: continue
> 
>         print prefix + sym
> 
>         if type(v) is ClassType:
>             print_symbols(v, prefix + sym + '.')
> 
> 
> See also the pyclbr module for a way to do something similar without
> executing the module.
> 
> Bruce
> 
> 
> 
> "D-Man" <dsh8290 at rit.edu> wrote in message
> news:mailman.977522226.5104.python-list at python.org...
> >
> > This shouldn't be too hard with a python script and some regexes with
> > some more string processing in between (or maybe multiple pass).
> >
> > Everything you want to change is bounded by "def" on one side and "("
> > on the other.  Then replace the capitals with _ followed by the lower
> > case version.
> >
> > A quick (partial) solution:
> >
> > re = re.compile( "def ([a-z]+)([A-Z])([a-z]*)\(" )
> >
> > for line in source.readlines() :
> > match = re.match( line )
> > if ( match ) :
> > res = "def " + re.group( 0 ) + "_" + \
> > string.lower( re.group( 1 ) + re.group( 2 ) + "("
> >
> >
> > This isn't complete and won't have the arguments in the result (nor
> > preserve indentation before the "def").
> >
> > I think the re.sub() function will do more of what you want it to.
> >
> >
> > HTH,
> > -D
> >
> >
> > On Fri, Dec 22, 2000 at 12:20:15PM -0800, Issac Trotts wrote:
> > > Does anyone know of a utility to change between the coding conventions
> > > of
> > >
> > > def fooBarBaz(): pass
> > >
> > > and
> > >
> > > def foo_bar_baz(): pass
> > >
> > > It is a pain to do this by hand.
> > >
> > > Thanks,
> > > Issac
> >
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list



From Randy.L.Kemp at motorola.com  Fri Dec  8 12:52:43 2000
From: Randy.L.Kemp at motorola.com (Kemp Randy-W18971)
Date: Fri, 8 Dec 2000 11:52:43 -0600 
Subject: Jpython not work in Win2000
Message-ID: <69BC4C6AE83ED311BC9400805FA7B14AF84CDE@il93exp01.css.mot.com>

I have tried the same thing under Windows 2000 and duplicated his error.  Any suggestions? When I did a jar tf jpython.jar, I noticed some classes starting with Perl.  What gives? 

-----Original Message-----
From: Hongzhen Zhang [mailto:hzzhang at hotmail.com]
Sent: Friday, December 08, 2000 10:57 AM
To: python-list at python.org
Subject: Jpython not work in Win2000


Hi,
I just installed Jpython on Windows 2000. After successful
installation, under dos prompt, I type:
jpython
it complains that 'jpython' is not recognized as an internal
or external command, operable program or batch file.
The jpython is a executable jar file, according to the file
property. How can that happen? But it works in Windows NT.
Any suggestions and recommendations are highly welcome.

zhang




From greg at cosc.canterbury.ac.nz  Tue Dec 12 22:59:29 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Wed, 13 Dec 2000 16:59:29 +1300
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <915jcf0c6o@news2.newsguy.com> <9162p3$7t2$1@animus.fel.iae.nl>
Message-ID: <3A36F421.2E9BD970@cosc.canterbury.ac.nz>

Carel Fellinger wrote:
> I ... would have prefered (I think) something like:
> 
> loop: suite
> break <condition>: suite
> ...

Last time I thought about this, I decided that
I would like something along the lines of

   while:
     line = f.readline()
   gives line != '':
     process(line)

Alas, it would require a new keyword, making it
forever excluded from the idyllic Garden of Python
(before Py3k, anyway).

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


From tim.hochberg at ieee.org  Tue Dec 19 10:54:08 2000
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Tue, 19 Dec 2000 15:54:08 GMT
Subject: while true: !!!
References: <mailman.977180341.10206.python-list@python.org> <slrn93t7bv.le.grey@teleute.rpglink.com> <KTz%5.48644$x6.22165257@news2.rdc2.tx.home.com> <91nc7h0s0p@news2.newsguy.com> <mailman.977231769.15324.python-list@python.org>
Message-ID: <AwL%5.80470$15.17054718@news1.rdc1.az.home.com>

"David Porter" <jcm at bigskytel.com> wrote in message
news:mailman.977231769.15324.python-list at python.org...
> * Alex Martelli <aleaxit at yahoo.com>:
> > "Rainer Deyke" <root at rainerdeyke.com> wrote in message
> > news:KTz%5.48644$x6.22165257 at news2.rdc2.tx.home.com...
> > >
> > > 0 is false, None is false, [] is false, () is false, {} is false, and
x is
> > > false where x is a class instance and x.__nonzero__() returns false.
> > > Everything else is true.  I think.
> >
> > Another couple of 'false's: a class instance which doesn't define
> > __nonzero__ but does define __length__ when the latter returns 0;
> > and, C-implemented user objects, similarly to class instances.
> >
> > I don't _think_ there are any more...
>
> An empty string is also false.

also 0.0, 0L, 0+0j =>  anything the equals zero.

This can be simplified somewhat by replacing [], (), {}, and "a class
instance which doesn't define __nonzero__ but does define length when the
later returns 0" with "a sequence of length zero".


If I attempt to summarize I get the following for false items:

* None
* Anything that equals zero (Including a class instance with __nonzero__
returning false)
* An empty sequence (Including a class instance with __len__ returning 0)

Coincendentally, this corresponds with the definition in section 5.10 of the
Python Reference Manual:

"In the context of Boolean operations, and also when expressions are used by
control flow statements, the following values are interpreted as false:
None, numeric zero of all types, empty sequences (strings, tuples and
lists), and empty mappings (dictionaries). All other values are interpreted
as true."

-tim





From rpgarchive at my-deja.com  Wed Dec 27 01:11:57 2000
From: rpgarchive at my-deja.com (rpgarchive at my-deja.com)
Date: Wed, 27 Dec 2000 06:11:57 GMT
Subject: zipfile.py
References: <3A37AD3E.7233E7B7@ex.ac.uk> <3A37B917.6AC49FEA@interet.com>
Message-ID: <92c17b$m3a$1@nnrp1.deja.com>

I get this same problem, but only when using zip files that have been
created on Linux, using 'zip'.

In article <3A37B917.6AC49FEA at interet.com>,
  "James C. Ahlstrom" <jim at interet.com> wrote:
> "M.J.Fisher" wrote:
> >
> > Hi,
> >
> > I have been having some trouble with using zipfile.py, I need to
unzip
> > .zip files some created in windows and
> > some in unix.
> >
> > zlib.error: Error -3 while decompressing: incomplete dynamic bit
lengths
> > tree
> >
>
> It looks like there is a version incompatibility in your
> version of zlib on Windows or Unix.  Zipfile depends on zlib
> to compress/uncompress files.  You can test this by specifying
> a compression method of ZIP_STORED for all files so they are not
> compressed.  Or you could check zlib versions.
>
> Jim Ahlstrom
>


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


From wolfson at midway.uchicago.edu  Wed Dec 13 15:29:34 2000
From: wolfson at midway.uchicago.edu (Ben Wolfson)
Date: Wed, 13 Dec 2000 20:29:34 GMT
Subject: A question.
References: <9152a0$acb$1@panorama.wcss.wroc.pl> <916cd20b3b@news1.newsguy.com> <jMAZ5.170$x3.2004@uchinews> <917gta020ig@news2.newsguy.com>
Message-ID: <O_QZ5.27$v3.256@uchinews>

In article <917gta020ig at news2.newsguy.com>,
Alex Martelli <aleaxit at yahoo.com> wrote:
>"Ben Wolfson" <wolfson at midway.uchicago.edu> wrote in message
>news:jMAZ5.170$x3.2004 at uchinews...
>> In article <916cd20b3b at news1.newsguy.com>,
>> Alex Martelli <aleaxit at yahoo.com> wrote:
>>
>> > [I don't wink, I don't adverbialize, I iconize
>> >only very occasionally, the polysyllabic eruditeness of my hallmark
>>                                           ^^^^^^^^^^^
>> >circumlocutions pushes my Gunning Fog AND my Flesch-Kincaid up
>> >in the high 20's -- what more could I do to shew hostility...?]
>>
>> *Surely* you mean "erudition"!
>
>Abso-bloody-lutely not, o Ben.  Why use a relatively common
>9-letter word when a perfectly correct 11-letter synonym of
>it, that is used much more rarely, offers itself?  Hey,
>'eruditeness' gets 68 Google hits in all (including those
>in Merriam-Webster's Online Collegiate Thesaurus, American
>Heritage, etc, which just serve to proclaim it as a fully
>correct and legitimate synonym for 'erudition'), while your
>preferred 'erudition' get 46.600 -- talk about hoi polloi!

However, I'm afraid the August Majesty that is the OED deigns give it only a
passing mention--must be because any Joe Schmoe off the street would
naturally formulate "eruditeness" from "erudite".  The prevalence of
"erudition" on the web only goes to show that either
    1.  English speakers are uncommonly well educated in the vagaries of
their tongue
    2.  "Eruditeness" is such a mark of an unimaginative, declasse mind that
only the kind of unreconstructed hooligan who seldom has access to a
computer would use it.

>Glad I landed a nicely-sized fish with this one.  I was
>rather betting on 'shew', though -- doesn't it just _look_
>like a typo?-)

Nah,  I think "shew" is more dignified-looking than "show", even if I am
tempted to pronounce it "shyoo".  That, and I didn't really notice it.

-- 
BTR
It infuriates me that Dracula may have needed my soul more than I do.
 -- Stanley Donwood, "Dracula"



From aleaxit at yahoo.com  Wed Dec 13 03:46:07 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 13 Dec 2000 09:46:07 +0100
Subject: looping with two counter variables through a list
References: <yv2zoi0ojef.fsf@lionsp093.lion-ag.de>
Message-ID: <917d1201qfj@news2.newsguy.com>

"Harald Kirsch" <kirschh at lionbioscience.com> wrote in message
news:yv2zoi0ojef.fsf at lionsp093.lion-ag.de...
>
> In Tcl I can write
>
>   set mylist {0 1 2 3 4 5}
>   foreach {a b} $mylist {
>     puts $a
>   }
>
> to print the even numbers of $mylist. Obviously
>
>    mylist = [0, 1, 2, 3, 4, 5]
>    for a,b in mylist:
>      print a
>
> does not work (ValueError: unpack sequence of wrong size).

Yep: the 'for a,b in mylist' implies that mylist is a
sequence of two-item subsequences ('tuple unpacking').


> What is Python's idiom for running with more than one variable through
> a flat list?

The underlying one is looping on indices as opposed to values:

    for i in range(0, len(mylist)-1, 2):
        a = mylist[i]
        b = mylist[i+1]
        # whatever else you need


The meta-idiom one generally uses in Python for all sorts of
"impedance-matching" tasks (when the language, or a library,
offers patterns/idioms A, B, C, and you'd like to use X, Y, Z
instead in your client-code) is wrapping it up in some kind
of intermediate, auxiliary object.

Here, for example:

class LoopByN:
    def __init__(self, seq, N):
        self.seq = seq
        self.N = N
    def __getitem__(self, i):
        last = self.N * (i+1)
        if last>len(self.seq):
            raise IndexError
        return self.seq[self.N*i:last]

and then:

    for a, b in LoopByN(mylist, 2):
        # whatever

This assumes you want to 'silently slice off' the last
len(sequence) % N items of the sequence, those that do
not fit in the 'loop-by-N' idea.  If you want, e.g., to
conceptually "pad" the sequence to a multiple of N items,
or other slight semantic variation, that's pretty easy,
too (and similarly, if you want to support sequences,
that are not built-in ones, that do not support slicing
as this proposed LoopByN class is using).


Alex






From jp at ulgo.koti.com.pl  Wed Dec 27 11:58:48 2000
From: jp at ulgo.koti.com.pl (Jacek =?iso-8859-2?Q?Pop=B3awski?=)
Date: 27 Dec 2000 16:58:48 GMT
Subject: tkinter and scroll
Message-ID: <slrn94k7u8.14h.jp@localhost.localdomain>

I want to create window with:
- text (diffrent fonts)
- images
- buttons
and scroll it (becouse it can be larger than screen)
Is it possible to do it in Tkinter?

from tutorial: (scrollbar)

"This widget is used to implement scrolled listboxes, canvases, and
 text fields."
 
what about buttons and other widgets? what about all-in-one ?

If it's impossible with Tkinter - is it possible with pygtk?

-- 
Safe in dreams
Away from where they are
Let me be nowhere
Just another star                            "Feed my head" - Ronnie James Dio


From katz at Glue.umd.edu  Wed Dec 13 13:10:03 2000
From: katz at Glue.umd.edu (Roy Katz)
Date: Wed, 13 Dec 2000 13:10:03 -0500
Subject: Dijkstra's Shortest Path algorithm
In-Reply-To: <3A37055C.D831CACF@yale.edu>
References: <Pine.GSO.4.21.0012090042370.15110-100000@y.glue.umd.edu> <dvrd3t40u9l7atgdro59m5ttg7dsfv7c4i@4ax.com> <Pine.GSO.4.21.0012122249480.19682-100000@y.glue.umd.edu> <3A37055C.D831CACF@yale.edu>
Message-ID: <Pine.GSO.4.21.0012131307060.761-100000@y.glue.umd.edu>

On Wed, 13 Dec 2000, Paul Magwene wrote:
> If you're interested in graphs and programming, how about a clique
> finding algorithm for undirected graphs?  



Alright, I checked out some background material on cliques, and I'll take
a look at Andrew Dalke's code (I looked at his site).  If I understand
correctly, all that is needed is to generate a clique-finding extension in
C? 

This raises another question for me:  Suppose my exension (1) relies on
another extension (2); is there a more efficient way for (1) to access
(2) without going through the Python layer?

Roey




From mak at mikroplan.com.pl  Tue Dec  5 19:32:48 2000
From: mak at mikroplan.com.pl (Grzegorz Makarewicz)
Date: Wed, 6 Dec 2000 01:32:48 +0100
Subject: How do packages and .pth files work under Windows (Py 2.0)
In-Reply-To: <90jk6r$rga$2@news.vanderbilt.edu>
Message-ID: <NDBBIKNLJKPLOLAJJJAPKEAJFDAA.mak@mikroplan.com.pl>

Create somename.pth in python root, eg.:
c:\blach
d:\blach

execute python and you will find c:\blach and d:\blach appended to sys.path

For more information look into lib/site.py:
def addsitedir(sitedir):
    ...
    for name in names:
        if name[-4:] == ".pth":
            addpackage(sitedir, name)

and code from global scope executed after addpackage:

prefixes = [sys.prefix]
if sys.exec_prefix != sys.prefix:
    prefixes.append(sys.exec_prefix)
for prefix in prefixes:
    if prefix:
       ...
       addsitedir(...)

mak


> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Jonathan Gilligan
> Sent: Tuesday, December 05, 2000 9:46 PM
> To: python-list at python.org
> Subject: Re: How do packages and .pth files work under Windows (Py 2.0)
> 
> 
> Thanks, but this doesn't tell me about when python reads .pth files,
> which is the crux of my problem. Between this and writing a
> __init__.py file which properly initializes the variable __all__, I
> managed to work around my immediate problem.
> 
> I have managed to resolve another part of my problem---it seems that
> Python reads the default value of the registry key
> "HKLM\Software\Python\CurrentVersion\2.0\PythonPath" only if it cannot
> resolve PYTHONHOME. If it can resolve PYTHONHOME, then it ignores the
> default value in "HKLM\Software\Python\CurrentVersion\2.0\PythonPath"
> and substitutes some default directories instead for the PYTHONPATH.
> It does add any values in keys below PythonPath, but not the value at
> the top of the tree.
> 
> This is undocumented, but can be found in the C code in getpathp.c.
> 
> I'd still like to know how to use .pth files properly, though, if
> anyone can help.
> 
> Jonathan
> 
> "Stuart" <sgm103 at cs.york.ac.uk> wrote in message
> news:z26X5.15943$R77.1248181 at nnrp4.clara.net...
> > > It would help me enormously if there were a way to get Python to
> print
> > > out a list of where it looks when it tries to resolve an "import".
> Is
> > > this possible?
> >
> > I'm not certain, but i think this does the trick:
> >
> > import sys
> > print sys.path
> 
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list



From phd at phd.pp.ru  Thu Dec 21 07:33:54 2000
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 21 Dec 2000 15:33:54 +0300 (MSK)
Subject: 3D GUI Project
Message-ID: <Pine.LNX.4.30.0012211531310.26142-100000@fep132.fep.ru>

http://slashdot.org/articles/00/12/20/2310213.shtml

"A guy that calls himself "matrixnan" introduced this project on NANs
Blender homepage. It's gonna be a GPLd 3D GUI for Linux using Blender as
construction kit. Blender is a professional freeware 3D
Animation/Modelling/Applicationkonstruction kit that features Python as
Plugin language (Plugins are a big deal in the 3D buisness). Coding of the
Project uses/will use Python, C and C++."

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




From dsh8290 at rit.edu  Tue Dec 12 19:23:02 2000
From: dsh8290 at rit.edu (D-Man)
Date: Tue, 12 Dec 2000 19:23:02 -0500
Subject: Plan9 port of python
In-Reply-To: <87ofyhood3.fsf@litterbox.meowing.net>; from meowbot@meowing.net on Tue, Dec 12, 2000 at 06:44:40 -0500
References: <914hfv$pi1$1@nnrp1.deja.com> <87ofyhood3.fsf@litterbox.meowing.net>
Message-ID: <20001212192302.G16853@westsidecnc.rh.rit.edu>

I had heard of Plan9 briefly a couple of years ago.  I'm just wondering who
has/uses Plan9.

-D


On Tue, 12 Dec 2000 06:44:40 Angora Meowbot wrote:
 | dsavitsk <dsavitsk at my-deja.com> wrote:
 | 
 | > does a port of python exist for plan9
 | 
 | Yes, but it still needs work.
 | <URL:ftp://csociety-ftp.ecn.purdue.edu/pub/plan9/python/>
 | 
 | > or inferno?
 | 
 | I don't think so.
 | -- 
 | http://www.python.org/mailman/listinfo/python-list
 | 




From bernhard at abnoba.intevation.de  Thu Dec 28 04:24:44 2000
From: bernhard at abnoba.intevation.de (Bernhard Reiter)
Date: 28 Dec 2000 09:24:44 GMT
Subject: PEP 222 draft
References: <slrn947bj1.1rs.amk@207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com>
Message-ID: <92f0ss$re4$1@newsserver.rrzn.uni-hannover.de>

In article <slrn947bj1.1rs.amk at 207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com>,
	amk at mira.erols.com (A.M. Kuchling) writes:

>     * fcgi.py : A new module adding support for the FastCGI protocol.
>       Robin Dunn's code needs to be ported to Windows, though.

I'll second this.

fcgi.py only uses Unix Domain sockets for now.
It cannot deal with tcp ports or Windows named pipes.
Both would be nice to have, but at least the last one does not seem
to be trivial.

Bernhard



-- 
Professional Service around Free Software                (intevation.net)  
The FreeGIS Project                                         (freegis.org)
Association for a Free Informational Infrastructure            (ffii.org)
FSF Europe                                            (www.fsfeurope.org)


From aleaxit at yahoo.com  Fri Dec 29 12:04:21 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 29 Dec 2000 18:04:21 +0100
Subject: Great New Book
References: <t4o1vr8hktk531@corp.supernews.com> <92ico9$jde$1@panix2.panix.com>
Message-ID: <92ig6k02apk@news2.newsguy.com>

"Aahz Maruch" <aahz at panix.com> wrote in message
news:92ico9$jde$1 at panix2.panix.com...
> In article <t4o1vr8hktk531 at corp.supernews.com>,
> Jason Cunliffe <jasonic at nomadicsltd.com> wrote:
> >
> >The only shame is not the author's fault - it is the still crazy high
price
> >$45 and thick paper which bulks the 760pages to 2.5" thick.
>
> Hmmm....  I don't know how thick the pages actually are, but I sure
> wouldn't want a book to have tissue-paper pages.  It's a tricky balance.

63.5 mm for 760 pages works out at over 83 microns per page.

Rough comparison with a few high-quality books around here (O'Reilly,
Cambridge University Press, Addison-Wesley) appears to show a range
of, roughly, 30 to 60 microns per page.  And it's anything BUT
'tissue-paper' (or fine India paper, for that matter).  Over 80u/pg
is _really_ thick paper -- not quite cardboard, but...:-)


Alex





From chicago_george at my-deja.com  Wed Dec 13 13:16:17 2000
From: chicago_george at my-deja.com (chicago_george at my-deja.com)
Date: Wed, 13 Dec 2000 18:16:17 GMT
Subject: Import Error: no module named Java
Message-ID: <918ed8$vfq$1@nnrp1.deja.com>

I'm using JPython 1.1 on Java ver 1.1
Does anyone have any idea why I get the error msg

ImportError: no module named java

I get that when I try doing something like;
>>> from java import awt

WHY??!!  Anyone.. please??

Thanks!
George


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


From claird at starbase.neosoft.com  Wed Dec 27 09:01:43 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 27 Dec 2000 08:01:43 -0600
Subject: Python vs PHP (& Java?)
References: <mailman.977897649.4877.python-list@python.org> <92cdn00hck@news2.newsguy.com> <87k88mxcf4.fsf@eugene.prosa.it>
Message-ID: <D722E432FF377E01.657BAF2ECAA117A3.909A66D671189EE3@lp.airnews.net>

In article <87k88mxcf4.fsf at eugene.prosa.it>,
>"Alex Martelli" <aleaxit at yahoo.com> writes:
David N. Welton <davidw at prosa.it> wrote:
>
>> So, _what_ "much functionality" of PHP couldn't be "replaced" by
>> Python solutions?
>
>Personally, I use Tcl for this sort of thing, but I think that Python
>is pretty good too, for a very important reason:
>
>PHP is used on the web and only on the web.  Python, Tcl, Java, and
>company are languages that you can take and reuse elsewhere.
			.
			.
			.
Incidentally, the PHPers certainly are aware of
this situation, and it's likely that PHP will
become available through command-line and/or
MySQL stored-procedure applications, sometime
in 2001.  I expect to write more on this, pro-
bably in January.
-- 

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


From arturo_perez at wanadoo.es  Wed Dec  6 04:15:04 2000
From: arturo_perez at wanadoo.es (Arturo Pérez Mulas)
Date: Wed, 06 Dec 2000 09:15:04 GMT
Subject: [DAO Database in win32] What is more efficient
References: <0g8X5.152$7U6.3922@m2newsread.uni2.es> <PXlX5.117656$U46.3885672@news1.sttls1.wa.home.com>
Message-ID: <ssnX5.41$rZ6.2186@m2newsread.uni2.es>

> If you aren't using ADO you should be. DAO was replaced by ADO some time
> ago.

By the way, is mxODBC any better?

Thanks




From dsh8290 at rit.edu  Mon Dec 18 22:50:50 2000
From: dsh8290 at rit.edu (D-Man)
Date: Mon, 18 Dec 2000 22:50:50 -0500
Subject: An "exit" function for C modules?
In-Reply-To: <Aaz%5.2310$1%2.101471@sjc-read.news.verio.net>; from lk@veriomail.com on Tue, Dec 19, 2000 at 01:51:28AM +0000
References: <Aaz%5.2310$1%2.101471@sjc-read.news.verio.net>
Message-ID: <20001218225050.A13303@westsidecnc.rh.rit.edu>

On Tue, Dec 19, 2000 at 01:51:28AM +0000, Lawrence Kesteloot wrote:
> Should I use "atexit()"?  I'm concerned that I don't know
> when that's called -- it could be after some cleanup has
> been done by Python that makes decref dangerous.
> 

The functions registered with atexit() are called when the program
(main) exits.  That would be when the script is done and the
interpreter shuts down (or when the interpreter dies becuase an
exception wasn't caught).

I believe the functions registered with atexit() are called in the
order they are registered.  This means that *if* (I don't know) the
interpreter registers some functions before you register yours it may
not work as you want it to.

> It seems as though there should be an "exit<modulename>"
> function like the "init<modulename>" call at an appropriate
> time that allows cleanup of globals.
> 

Sound like a good idea to me.

> Lawrence
> 

-D



From amk at mira.erols.com  Thu Dec 28 18:29:28 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 28 Dec 2000 23:29:28 GMT
Subject: Python vs PHP (& Java?)
References: <92f3kp01cqe@news1.newsguy.com>, <mailman.977897649.4877.python-list@python.org> <slrn94l86m.a3h.tim@degree.ath.cx> <mailman.978021012.21874.python-list@python.org>
Message-ID: <slrn94nk02.jk.amk@207-172-57-128.s128.tnt2.ann.va.dialup.rcn.com>

On 28 Dec 2000 16:22:03 -0000, Moshe Zadka <moshez at zadka.site.co.il> wrote:
>(BTW: Are there plans for quixote/mod_snake integration? That'd
>be ultra-cool: one of the benefits of HTML::Mason is that it
>integrates really well with mod_perl)

Not at this point in time; we use FastCGI, which has been an adequate
performance boost so far.

--amk



From sandj.williams at gte.net  Fri Dec 22 12:32:01 2000
From: sandj.williams at gte.net (Steve Williams)
Date: Fri, 22 Dec 2000 17:32:01 GMT
Subject: ANNOUNCE: soundex.py
References: <mailman.977338109.32467.python-list@python.org>
		<3A414897.D82C953C@gte.net> <mailman.977504503.25273.python-list@python.org>
Message-ID: <3A439139.87682C18@gte.net>

Skip Montanaro wrote:

> >>>>> "Steve" == Steve Williams <sandj.williams at gte.net> writes:
>
>     Steve> I've found you can good results with a variable length soundex
>     Steve> string--the more information you give the routine (first names,
>     Steve> middle names, prefixes and suffixes) the better/smaller the
>     Steve> result set.
>
> Your wish is my command.  The module at
>
>     http://www.musi-cal.com/~skip/python/soundex.py
>
> now computes a variable-length soundex string if you set soundex.NDIGITS to
> 0:
>

[schnipp]

Thank you very much.




From moshez at zadka.site.co.il  Thu Dec 28 10:54:41 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: 28 Dec 2000 15:54:41 -0000
Subject: Interesting behaviour of the assignment
In-Reply-To: <92f4gk$qk6$2@news.nuri.net>
References: <92f4gk$qk6$2@news.nuri.net>
Message-ID: <20001228155441.6997.qmail@stimpy.scso.com>

On Thu, 28 Dec 2000, "June Kim" <junaftnoon at nospamplzyahoo.com> wrote:

> >>> a=100
> >>> b=100
> >>> a is b
> 0
> >>> a=99
> >>> b=99
> >>> a is b
> 1
> 
> Is this an intended/expected result? What is the lying structure beneath
> this?

The Python implementation makes no promises of the identity of 
immutable data types which are ==.
In other words, if "a==b", and both a and b are of the same immutable
data type (e.g., strings, integers or tuples) then it might or might
not be the case that "a is b". 

This allows the implementation to do all sorts of nifty optimizations.



From cfelling at iae.nl  Tue Dec 12 17:50:17 2000
From: cfelling at iae.nl (Carel Fellinger)
Date: 12 Dec 2000 23:50:17 +0100
Subject: Label-Value
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <r0pzoi11lhz.fsf@black.cs.washington.edu> <5jvZ5.3588$Kd1.418770@newsb.telia.net> <r0py9xl1k2n.fsf@black.cs.washington.edu>
Message-ID: <916a39$9it$1@animus.fel.iae.nl>

Daniel Wood <daniel at cs.washington.edu> wrote:

> Hmmm, I'm open to the suggestion that I need to rethink my plan at a
> larger scale, but isn't a mutable int likely to sometimes be useful?
> (Just like a mutable list is sometimes useful.)  And if so, is [ 2 ]
> the currently favored idiom?  (And if so, isn't there something
> better?)

What about a class? The only thing that is impossible is to set a new
value to its instances with the "instance = value" syntax. But then
neither does "a=1; a=2" assign a new value to "a", it simply makes "a
refer to a different object as Frederiq has explained recently on the
list.  So the proper way to bind new values to an instance is to
generate a new instance like "a = MutableInt(3)", but then "a.set(3)"
might be acceptable in many cases. The fun of using a class is that
given it has the needed special methods defined you can easily use
this mutable-int-instance in expressions like in "a * 3".

-- 
groetjes, carel


From fgeiger at datec.at  Sat Dec 30 10:02:20 2000
From: fgeiger at datec.at (Franz GEIGER)
Date: Sat, 30 Dec 2000 16:02:20 +0100
Subject: Python for prototyping of the user interface
References: <20001228.17165633@mis.configured.host>
Message-ID: <92ktj8$vhk$1@newsreaderg1.core.theplanet.net>

If you like wxWindows, try wxDesigner or Boa Constructor.

Regards
Franz

"Volkmar Busch" <vbusch at gmx.de> wrote in message
news:20001228.17165633 at mis.configured.host...
Hi,

I took my first steps in using Python, and I really like the language. I
read a lot about using Python for prototyping. As a basis for discussion
with our users we use GUI prototypes. Building complex user interfaces in
Python to me seems quite complicated, because

- I am missing a Gui builder
- I am missing controls like notebooks and tables

So I'd like to asked the experienced Python users how they solve this
problem.

Thanks in advance,
Volkmar




From ivnowa at hvision.nl  Fri Dec 15 05:58:42 2000
From: ivnowa at hvision.nl (Hans Nowak)
Date: Fri, 15 Dec 2000 10:58:42 GMT
Subject: Scopes and exceptions
References: <91chgc$7m8$1@merlin.rembrandtstr.bocholt.de>
Message-ID: <3A39F7CA.F8A@hvision.nl>

Magnus Heino wrote:
> 
> Could someone please explain this to me:
> 
> [magnus at daysleeper magnus]$ cat e.py
> 
> a = 'test'
> b = 'test'
> 
> print 'a      = %s' % id(a)
> print 'b      = %s' % id(b)
> 
> def func():
>         print "'test' = %s" % id('test')
> 
> func()
> 
> try:
>         test, '%s' % (a,)
> except NameError, e:
>         print 'e      = %s' % id(e)
>         print 'a      = %s' % id(a)
>         print 'b      = %s' % id(b)
>         print "'test' = %s" % id('test')
> 
> [magnus at daysleeper magnus]$ python e.py
> a      = 136227368
> b      = 136227368
> 'test' = 136227368
> e      = 135997204
> a      = 136227368
> b      = 136227368
> 'test' = 136227368
> [magnus at daysleeper magnus]$

I don't understand what is so special about this? Python stores a string
'test' in memory... a points to this string, then b, so id(a), id(b) and
id('test') all yield the same result. This is also the case after the
exception. Variable e has a different id, because it's a different
variable; not a string, but an instance of exceptions.NameError. 

HTH,

--Hans Nowak


From sh at ttsoftware.co.uk  Fri Dec 15 09:26:05 2000
From: sh at ttsoftware.co.uk (Steve Horne)
Date: Fri, 15 Dec 2000 14:26:05 +0000
Subject: Python advocacy
References: <91bln3$lmr$1@nnrp1.deja.com> <3A39AC3E.FA75C4FF@engcorp.com>
Message-ID: <nl5k3t03on9ehf9d50nov3coi71bu8tgo4@4ax.com>

On Fri, 15 Dec 2000 00:29:34 -0500, Peter Hansen <peter at engcorp.com>
wrote:

>jschmitt at vmlabs.com wrote:
>> 
>> I just read this good article about Perl advocacy.  I think it applies
>> to Python equally.  I found the link mentioned from /. this morning.
>> 
>> http://www.perl.com/pub/2000/12/advocacy.html?wwwrrr_20001213.txt [snip]
>
>I didn't bother reading the whole thing because I noticed early on (and
>confirmed throughout) that the author is oblivious to the huge
>investment people put into developing systems in a particular language
>(not to mention the investment in *learning* that language) and the fact
>that as long as the language stays popular, improves, and continues to
>serve their needs well, they get a payoff from that investment.

Pardon. OK - it wasn't an issue he discussed - but that doesn't make
him oblivious. From what he was describing, he has made a much bigger
investment in these things than the average programmer. To quote from
the article...

: One of the things I found most dismaying about the Perl 6 RFC process
: was the parochialism of many of the submissions. The submitter would
: be trying to solve some problem, and would come up with a crappy
: solution. And meanwhile, there would be a perfectly serviceable
: solution in the language next door, just waiting to be reused.
: But it seemed as though a lot of the people making proposals only
: knew Perl, and not any other languages.

Hardly the words of someone with no investment in Perl, or for that
matter in the development of programming languages, are they?

I'd be interested in knowing how you can 'confirm throughout' without
reading the whole thing. It sounds to me that you have made exactly
the kind of presumption-before-the-facts that the author was
complaining about.

If you read the whole article, for instance, you would know that the
author was not complaining about advocacy in general, merely about the
way it tends to be done. In fact, here's another quote...

: I don't really hate advocacy. I just hate the way we do it most of
: the time.

You, on the other hand, seem to be advocating advocacy in exactly the
tribal sense that the author describes and, seeing an opinion other
than your ideal, seem to wrongly assume that his article is nothing
more than a shallow justification for the opinion that 'advocacy is
bad'.

My opinion: I largely agree with this article. On comp.lang.c, for
instance, I am sick to death of people being bullied for merely
mentioning C++, or using words like 'assert' in the general
programming sense rather than as they are described in the ANSI C
standard. And I am also honest enough to say that I've sometimes
fallen into the tribal advocacy trap myself - particularly with
respect to Perl, which I strongly dislike. But overall my dislike of
Perl is based on a number of bad features of Perl, not on the fact
that I like python.

>If the language "dies", that investment is largely lost.
>
>The theory is that by advocating a language one can increase the
>likelihood of its survival and improvement.  If true, then advocating a
>language directly serves ones own needs.

OK - assume you have a co-worker who does a similar job to you. Do
you...

a.  Sabotage everything he does in the hope that he will be sacked,
    leaving you safe.
b.  Co-operate with him in the hopes of getting more of the work done,
    and of possibly learning some new tricks that you can use to
    improve your own value in the future.

I have an investment in a number of languages and, like the author, I
don't see the successes of some as failures of the others. I will
quite happily advocate Pascal, Modula 2, Ada, C, C++, Java, BASIC,
Icon, Lisp, Haskell, Miranda, Prolog, Forth and various others. I will
not advocate COBOL, but I will advocate some features of COBOL. I have
made a large investment in all of these languages, and will continue
to make investments in most of them and in any others that I happen to
find useful.

Note the language Icon. If you check up, you will find that it is in
fact all but dead. Is my investment a waste? - No. The concepts that
Icon pioneered are so useful that I have developed a C++ library to
support similar concepts, and someone recently announced a generator
library for Python in this newsgroup.

>There is no reason to focus on the fact that people appear to take this
>sort of thing personally.  That's probably just human nature, and likely
>not even a Bad Thing.  The point is there are good and wholly rationale
>reasons for advocacy even in its current style, and you could say that
>anyone who does *not* advocate a particular language is contributing to
>its demise (the "if you're not part of the solution, you're part of the
>problem" theory of life :-).

Yes, advocacy is important. It is a large part of how new languages
become successful. But exclusive there-is-only-one-answer advocacy is
not the same thing. It does not lead to progress - only to stagnation.
It is the reason a whole range of ideas are dismissed as unnecessary
and irrelevant, instead of being sought out as new tools to add to the
toolbox - tools which may be better than any other tool for certain
jobs.

Just as for Perl, Python would not be in its current form except that
people recognise that it has failings, and that there are better
methods in other languages. An obvious example - the list
comprehensions would not be useful except that filter, map and lambda
have weaknesses. List comprehensions are copied directly from
functional languages such as Haskell, which in turn copied them from
mathematical set notation. This does not mean that older versions of
Python were universally bad, or that Haskell is universally good. It
simply means that Python 2 got a bit better by copying a good feature
from another language (or in this case, paradigm).

In fact, Guido van Rossum himself was an ABC advocate who recognised
that ABC had failings - if he hadn't been, Python would never have
existed.

>I think I'll just ignore whatever other, possibly valid points
>Mark-Jason Dominus may have made and carry on advocating as I always
>have.  He hasn't convinced me.

How could he convince you - there are none so blind as those who will
not see. You had already decided what you thought he meant without
even reading the article. You made a judgement before bothering to
gather all the relevant facts, based on a shallow viewpoint. In short,
you are displaying a kind of prejudice.

But at least you admit that what you are ignoring could be valid. We
should all be grateful for small mercies.

-- 
Steve Horne
Home : steve at lurking.demon.co.uk
Work : sh at ttsoftware.co.uk


From grey at despair.rpglink.com  Fri Dec 29 15:51:13 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Fri, 29 Dec 2000 20:51:13 -0000
Subject: Great New Book
References: <t4o1vr8hktk531@corp.supernews.com> <92ij47$n3p$1@bob.news.rcn.net> <slrn94pnrk.irm.grey@teleute.rpglink.com> <92inl3$a5o$1@panix2.panix.com> <slrn94pp5m.irm.grey@teleute.rpglink.com> <roy-8A57F2.15444129122000@news.panix.com>
Message-ID: <slrn94pua1.irm.grey@teleute.rpglink.com>

On Fri, 29 Dec 2000 15:44:41 -0500, Roy Smith <roy at panix.com> wrote:
>But I thought in python " and ' meant the same thing? :-)

    Bwahahahahaha!!!!  Good one.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From dalke at acm.org  Fri Dec 29 16:20:56 2000
From: dalke at acm.org (Andrew Dalke)
Date: Fri, 29 Dec 2000 14:20:56 -0700
Subject: 2001 Enchancement Wishlist
References: <3A4CEACB.220998A0@javanet.com>
Message-ID: <92iv76$flh$1@slb2.atl.mindspring.net>

Raymond Hettinger wrote:
>1.  List.append(elem) and List.pop() are great.  Let's add two more
>     methods, List.prepend(elem) and List.leftpop().  The goal is to
>     make queues,deques, and reverse stacks easier, faster, and clearer
>     to implement.

Adding the method does not make things faster.  There would need
to be an implementation change so that popping from the left
doesn't take O(N) time.  This has come up before.  Guido even
mentioned some names for "prepend" and "leftpop" which came from ..
some other language (forgot which).

The thing is, there are reasons for different data structures
like queues over lists.  Trying to encapsulate all of them in
one data structure is hard - and trying to do it efficiently is
impossible.  Instead, why not develop a library of data
structures for Python, with queue, trees of various sorts,
stacks and (to make some happy :) lists which resize in N*log(N)
time instead of N**2 time as N >> 1000.

BTW, prepend is easily done with an insert at 0.

>I recommend that we extend the same courtesy to
>   the % string formatting operator.  For example:
>
>             print "Spam with %s, eggs, and spam" % aUserDictionary

Time machine beat you to it.  I wrote a docstring formatting
package with templates based on % interpolation using a
specialized dictionary-like object.  You could do something like:

   "%(module.name|htmlescape|h1)s<p>%(module.docstring|htmlescape)s"

It would parse up the components, get the first term (including
the "." parts) then apply the results through the set of
"|" transformations.

>>> class MyDict:
...     def __getitem__(self, name):
...             print "Asking for", repr(name)
...             return "*"
...
>>> d = MyDict()
>>> "%(module.name|htmlescape|h1)s<p>%(module.docstring|htmlescape)s"%\
...     d
Asking for 'module.name|htmlescape|h1'
Asking for 'module.docstring|htmlescape'
'*<p>*'
>>>


>   I think [__init__] should continue to return the new object
> by default but also allow another return item to be substituted.
> For example:
>
>            def __init__( self ):
>                   if MyClass.aSingleton == None:
>                         MyClass.aSingleton = self
>                   return MyClass.aSingleton

Use a factory function for the general case.  For example
if you need this singleton behaviour you can do:

class MyClass:
  pass
_MyClass_singleton = MyClass()
def MyClass():
  return _MyClass_singleton

or more specifically:

class MyClass:
  pass
_MyClass = MyClass

_MyClass_singleton = None

def MyClass():
  global _MyClass_singleton
  if _MyClass_singleton is None:
    _MyClass_singleton = _MyClass()
  return _MyClass_singleton

I'll grant you that it doesn't look as clean, but it isn't
used all that often and the above is a pretty standard way
to write cached/static data.  Also, there have been a lot
of threads on special support for singletons in Python.

Oh, and __del__ also does not allow a return value.

>4.  Allow dictionaries to respond to sequence operations
> as if .keys() were in the list object.  Examples:

What if the keys in the dictionary are integers?

data = {1: "one", 3: "three"}
print data[1]  # This prints "one"
print data[0]  # Do you want this to be valid or an error?

You can get the behaviour you want with a simple wrapper:
(this is untested)

class IndexedDict:
  def __init__(self, dict):
    self.__dict = dict
  def __getitem__(self, key):
    try:
        return self.__dict[key]
    except KeyError:
        keys = self.__dict.keys()
        try:
            return self.__dict[keys[key]]
        except IndexError:
            raise KeyError, key
  def __getattr__(self, name):
    return getattr(self.__dict, name)


Several implementation choices are possible (I for one do not
want your behaviour).  You can easily implement what you want
from what's already there, so this isn't a needed change to Python.

                    Andrew
                    dalke at acm.org





From kragen at dnaco.net  Sat Dec 30 18:04:27 2000
From: kragen at dnaco.net (Kragen Sitaker)
Date: Sat, 30 Dec 2000 23:04:27 GMT
Subject: Probs with classes: cannot instantiate
References: <20001230.23424514@snafu.btx.dtag.de>
Message-ID: <%Rt36.18$vd1.2480@e420r-atl2.usenetserver.com>

In article <20001230.23424514 at snafu.btx.dtag.de>,
Markus Vogl  <vogl at mavo.de> wrote:
>This is the file [ DirParser.py] stuff:
>class DirParser:
>	def display(self):
>		print "This is the display method."
>
>To test this i use classtester.py in the same directory [which is in 
>path]:
>
>import DirParser
>x = DirParser()
>
>The error messages:
>	x = DirParser()
>...
>TypeError: call of non-function ( type module )

You want x = DirParser.DirParser()

Or possibly you should call the file dp.py and say 
"from dp import DirParser" instead of "import dp".
-- 
<kragen at pobox.com>       Kragen Sitaker     <http://www.pobox.com/~kragen/>
Perilous to all of us are the devices of an art deeper than we possess
ourselves.
       -- Gandalf the White [J.R.R. Tolkien, "The Two Towers", Bk 3, Ch. XI]




From pbjorn at uswest.net  Wed Dec 20 17:53:14 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Wed, 20 Dec 2000 15:53:14 -0700
Subject: ANNOUNCE: soundex.py
References: <LNBBLJKPBEHFEDALKOLCCENFIEAA.tim.one@home.com>
Message-ID: <3A41385A.ABCCAB0E@uswest.net>

Tim Peters wrote:

> [posted & mailed]
>
> [Bjorn Pettersen]
> > Thanks Skip!  I was actually using the soundex module in one of my less
> > frequently used apps so I hadn't realized that it was gone yet :-)
> >
> > Could I get you to put an explicit licence on it (it would be
> > much more usable to me if it had a licence and if the licence wasn't
> > GPL <wink>).
>
> Note too that Sean True wrote a Python implementation of the
> more-sophisticated NYSIIS algorithm, available at
>
>     http://metagram.webreply.com/downloads/
>
> (it's linked to from Parnassus if you forget that).  It's not encumbered.
>
> Whatever stuff of mine Skip reused was released to the public domain.
>
> if-soundex-is-the-answer-the-question-was-ill-advised<wink>-ly
>     y'rs  - tim

Thanks for the info, Tim!  The soundex version found my misspellings of
"potatos" and "appple pie", and the nysiis encoding didn't like "1000 Island"
so I'll probably stay with soundex (after all I'm only trying to be sligthly
more helpful than not returning anything <wink>).

better-than-nothing'ly y'rs
-- bjorn





From aleaxit at yahoo.com  Fri Dec 22 06:24:57 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 22 Dec 2000 12:24:57 +0100
Subject: grammar & spell check
References: <91u20g$pl3$1@bob.news.rcn.net>
Message-ID: <91vdmr028k8@news2.newsguy.com>

"Satheesh Babu" <vsbabu at erols.com> wrote in message
news:91u20g$pl3$1 at bob.news.rcn.net...
> Hi,
>
> Is there any suggestions on what services/software I can use
> to do grammar/spell check? for English language.
>
> requirements:
> 1. given a text string, it should do the checks and return
> errors/suggestions report
> 2. Should be able to interface from Python
> 3. Preferably should work from Win32 systems
>       (that gives an idea, can't I use  MS Word's functions for this?)

Of course, if you can count on Word being present on
your target machine; Python, with win32all (already
present in the ActiveState distribution), is an excellent
COM client.

For example, here's a spell-check program in Python:

import win32com.client

msword = win32com.client.Dispatch("Word.Application")

string = "Here is some texxt with a coupple of problems"

for word in string.split():
    if msword.CheckSpelling(word):
       print "!%s! OK!" % word
    else:
       print "?%s? ->"%word,
       suggestions = msword.GetSpellingSuggestions(word)
       for suggest in suggestions:
           print suggest.Name,
       print

D:\PySym>python spel.py
!Here! OK!
!is! OK!
!some! OK!
?texxt? -> text texts
!with! OK!
!a! OK!
?coupple? -> couple couplet coupler coupled couples
!of! OK!
!problems! OK!

D:\PySym>

You'll need Word's online docs for VBA to learn all the
details about spell-checking and grammar-checking.


For a cross-platform and free solution, I would expect
you could wring something similar out of StarOffice 5.2,
but unfortunately I wouldn't know where to start to
get cross-platform interfacing of Python with SO 5.2
(on Windows, specifically, they probably expose some
COM object model -- but that wouldn't help your programs
run under, say, Linux).  Has anybody managed to drive
StarOffice from Python cross-platformly, btw...?


Alex





From aleaxit at yahoo.com  Mon Dec 18 03:35:18 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 18 Dec 2000 09:35:18 +0100
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <3a36c85e.4409136@News.CIS.DFN.DE> <owen_nospam-1312001131590001@cs204-90.fsmodem.washington.edu> <3A37DD31.C9A5F18C@pacbell.net> <91bhlh$c64u$1@nntp6.u.washington.edu> <91d7270niu@news1.newsguy.com> <91dmda$89ei$1@nntp6.u.washington.edu> <k1e2i3t4h5r6a7y-BD04E8.22073517122000@news>
Message-ID: <91ki9801tjl@news1.newsguy.com>

"Keith Ray" <k1e2i3t4h5r6a7y at 1m2a3c4.5c6o7m> wrote in message
news:k1e2i3t4h5r6a7y-BD04E8.22073517122000 at news...
    [snip]
> > (The Mac issue is also why IBM's smalltalk is out, unless they're
> > planning a port to MacOS X. And whoever said Cincom's GUI is slow on a
> > Mac, I agree, and unattractive, too. I didn't realize it was fast on
> > other platforms. That's very good to know.)
> >
> > -- Russell
>
> I wonder if it would be feasible to wrap YAAF in Python?
>
> YAAF is a c++ cross-platform gui/application-framework class library.
> <http://www.pandawave.com/yaaf1.html>  Works on Mac, Windows, Linux.

Haven't examined it in complete detail, but from a cursory first
look I can see no technical reason that would make this wrapping
unfeasible.  So many other GUI frameworks have been wrapped for
Python, after all.  It _would_ be a substantial amount of work,
of course, and probably need a team of volunteers driven by a
strong desire to make YAAF available to Pythonistas cross-platform,
with Mac-, Linux-, _and_ Windows-equipped and -knowledgeable
volunteers as part of the team.  (I suspect that making wxPython
work perfectly on Mac would be a lesser undertaking, if nothing
else because it only requires the Mac-part, as Linux and Windows
are pretty well covered already).


Alex





From junaftnoon at nospamplzyahoo.com  Sat Dec 30 14:55:48 2000
From: junaftnoon at nospamplzyahoo.com (June Kim)
Date: Sun, 31 Dec 2000 04:55:48 +0900
Subject: Lazy Python usage (hack request)
References: <Pine.LNX.4.10.10012301443330.13756-100000@kev.ioc.ee> <92l4tn$raf$1@panix3.panix.com>
Message-ID: <92lemh$5or$1@news.nuri.net>

"Aahz Maruch" <aahz at panix.com> wrote :

[snip]

> 
> Well, if you're going to stick to it not being an option to make any
> changes, the answer is "no".  However, you can make a little hack that
> *does* work:
> 
> >>> from lazy import lazy
> >>> a = lazy.b
> No name 'b' defined, using default 777
> >>> print a
> 777
> 
> Inside lazy.py, you define a class called lazy with a __getitem__()
> method.

Shouldn't the method be __getattr__() instead?


Best regards,

June



From pearu at cens.ioc.ee  Fri Dec 15 17:41:59 2000
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Sat, 16 Dec 2000 00:41:59 +0200
Subject: __repr__ policy problem: should module be prefixed to output?
In-Reply-To: <Pine.LNX.4.10.10012152140170.10103-100000@kev.ioc.ee>
References: <3A3A3A6D.2C3B1A91@sbc.su.se> <91dfl0014sa@news1.newsguy.com> <Pine.LNX.4.10.10012152140170.10103-100000@kev.ioc.ee>
Message-ID: <Pine.LNX.4.10.10012160011400.14411-100000@kev.ioc.ee>

Hi,

On Fri, 15 Dec 2000, Pearu Peterson wrote:

> There is a solution to this repr string dilemma. For example,
> consider foo.py:
> 
> # begin of file foo.py
> class A:
>     def __init__(self,value):
>         self.value = value
>     def __repr__(self):
>         return '%s.%s(%s)'%(self.__module__,
>                             self.__class__.__name__,
>                             self.value)
> 
> exec 'import %s'%__name__   # this does the trick
>                             # (could it be done also in C extension
>                             #  modules?)
> 
> #eof foo.py

I now see several problems in this solution:
1) It does not work if 'from foo import A' is used
2) If module `foo' has local object with the name `foo', then it will be
   overwritten.
So, it is not useful solution. 2) can be avoided by the module developer
but I don't see how the problem 1) could be avoided. Any ideas?

> And the answer to the original question is Yes.

For completeness, let me give another 'bad' solution where the answer to
the original question is No.
This solution re-wraps builtin function eval (which, I am pretty sure,
will not please you). But anyway, here is the result (sources are given
at the end of this message):

Python session:
>>> import foo
Warning: Built-in function "eval" is re-wrapped
>>> foo.A(1)
A(1)
>>> eval(repr(foo.A(1)))
A(1)

Another Python session:
>>> from foo import *
Warning: Built-in function "eval" is re-wrapped
>>> A(1)
A(1)
>>> eval(repr(A(1)))
A(1)

The main problem to this solution is that different modules may define
different objects with the same name. So, for generality, the answer
should be Yes.

In conclusion, I don't have a good solution. Just these two bad ones that
might be useful only in special cases.
Are there any better solutions to the repr string dilemma?

Regards,
	Pearu

######### File evalhack.py
import sys,string
_eval = __builtins__['eval']
_register = {}
def eval(source,g=None,l=None):
    if g is None: g = globals()
    if l is None: l = g
    try: return _eval(source,g,l)
    except NameError:
        v = sys.exc_value
        o = v[0][string.find(v[0],"'")+1:-1]
        if _register.has_key(o):
            l[o] = _register[o].__dict__[o]
        return _eval(source,g,l)
eval.__doc__ = _eval.__doc__

def register(module,*objs):
    m = __import__(module)
    for o in objs:
        _register[o] = m

print 'Warning: Built-in function "eval" is re-wrapped'
__builtins__['eval'] = eval

######### EOF evalhack.py #######

######### File foo.py
class A:
    def __init__(self,value):
        self.value = value
    def __repr__(self):
        return '%s(%s)'%(self.__class__.__name__,self.value)

import evalhack
evalhack.register(__name__,'A')

######### eof foo.py



From aleaxit at yahoo.com  Sun Dec 24 11:17:09 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 24 Dec 2000 17:17:09 +0100
Subject: classes (was Re: Same again please for OOP)
References: <mailman.977623149.8184.python-list@python.org>
Message-ID: <9257mf0sl1@news1.newsguy.com>

"Rob Brown-Bayliss" <rob at ZOOstation.cc> wrote in message
news:mailman.977623149.8184.python-list at python.org...
> On 21 Dec 2000 20:49:27 -0800, Richard P. Muller wrote:
> > Can someone recommend a decent introduction to functional programming in
    [snip]
> Like wise,  I am trying to get my head around OOP, but all I can find
> are basic howto make a class in C++

OOP is _much_ more 'central' to Python than FP; thus, a good grasp
of it is proportionally more important.  'Howto' is easy, but you're
right in implying it's not as important as 'using it WELL'.


> I understand how to make classes in both Python and C++.  What I want is

Good; so, the easy part is out of the way.

> help in deciding waht would be a good class and what would not,
> preferably in a pythjon contxt, but a generic context would be wonderful

It can help to think of classes as coming in *two* kinds: one models
directly a type of 'objects' that you have identified while thinking about
the domain for which you want to provide solutions ('domain analysis');
the other one is a 'purely utilitarian' device, which just takes advantage
of some OO language mechanism.

The split is not as sharp as this would seem to imply, but this way of
presenting it may still help.  We could call the former kind 'semantic'
or 'problem-space' classes/objects, and the latter 'utility' or 'solution-
space' ones (many classes in real-world programs exhibit both aspects:
_some_ semantic correspondence to domain analysis, _some_ aspects
that are 'purely utilitarian').

Don't think of 'semantic' classes as 'good', and 'utility' ones as 'bad':
they're _both_ 'good' if, and only if, they help you solve your problems
with clarity and ease!  These are general programming issues, only
tangentially related to classes, but well worth emphasizing...:

    Semantic classes will be 'good' if they model those aspects of your
    problem-domain which you need to handle, to the level of detail and
    'faithfulness' which *you actually need* -- a model/program is always
    a _simplification and abstraction_ of the world (the map is not the
    territory)... eschew 'too-close' modeling for its own sake!

    Utility classes will be 'good' if they come in handy, with just the
right
    amount of 'mechanism' for your actual needs.  *Do the simplest thing
    that can possibly work*, in each case -- ready to 're-factor' things as
    the program evolves, often in an 'exploratory' way, to keep simplicity
    and elegance as you converge towards solutions.

    Don't put in extra complexity 'just in case', because *you ain't gonna
    need it* often enough that the effort won't pay for itself.  (Unless
    you are releasing a _framework_, or at least a module/library for
    general use, rather than a specific program; in this case, balancing
    generality, flexibility/reusability, and simplicity, can be even more
    of a design challenge:-).


And coming to classes (remember classes? it's a post about classes)...
it's hard to do them justice within the confines of one post, but, let me
start with some general considerations, and an example to clarify.


Classes let you *group*, "under one roof" so to speak, aspects of
_state_ ("data") and/or aspects of _behavior_ ("code").  In any given
case, unless you have some interest in effecting such a grouping, it's
quite unlikely that a class is the best approach for you in that case;
if you do have such interest, a class is likely to be 'right' -- unless
some other type of object (an already existing one, which you can
find in the Python library or some other available module/package)
meets some specific needs even better.

A simple case is when you just want to group some data items and
give each of them a name -- no specific 'behavior', just some state
with associate naming.  Tuples and lists let you do the grouping,
but the 'names' of the several items are just numeric indices -- often,
this is not the clearest, most readable approach.  Dictionaries would
let you 'name' items by supplying string keys -- the syntax is not
ideal if constant-strings are all you're using as names.  A class with
no behavior (methods) can easily be the best, simples approach here.

For example, suppose our program will need to model CD's.  A CD
object has several attributes -- depending on our purposes, we may
want to model some appropriate subset of them; for example, a
"title", a "price" in Euros, and a sequence of "tracks" (each of which
has its own attributes, say a "title" and a "duration" in seconds).

If we used tuples or lists, we might conventionally agree that the
first item is the title, the second one the price, the third one the
sequence of tracks (each with two items, title then duration).

For example, a Czech CD I bought this summer could be modeled
according to these conventions as (truncating the list of tracks!):
acd = ["Hity Papa Offenbacha", 5.00,
    [["Barkarola", 264], ["Piscu Heleny", 95], ["Pariduv Soud", 247]]]
or similarly with tuples instead of lists.  The problem with this is
that any given task, such as, say, "find the title of the longest
track", becomes not-very-readable and hardly-maintainable...:

def titleLongest(acd):
    tracks = acd[2]
    longest = 0
    for i in range(1,len(tracks)):
        if tracks[i][1] > tracks[longest][1]:
            longest = i
    return tracks[longest][0]

All of those '[2]', '[1]', '[0]' should feel unnerving -- strong hints
that our chosen representation is sub-optimal; and further, our
functions become *strongly* coupled to the physical way we've
chosen to lay out our data!  If we refactor the data representation,
we'll have LOTS of work ahead, redoing all of our functions -- what
an *unpleasant* prospect!

A dictionary may be a bit better...:

acd = {'title':"Hity Papa Offenbacha", 'price':5.00, 'tracks':
    [ {'title':"Barkarola", 'duration':264},
       {'title':"Piscu Heleny", 'duration':95},
       {'title':"Pariduv Soud", 'duration':247} ] }

The repetitiveness of the construction is unpleasant, but let's
see if the code using this representation is more readable...:

def titleLongest(acd):
    tracks = acd['tracks']
    longest = 0
    for i in range(1,len(tracks)):
        if tracks[i]['duration'] > tracks[longest]['duration']:
            longest = i
    return tracks[longest]['title']

Well, yes; at least, this code DOES suggest a little bit of what
we're doing.  But, could classes help...?

class CD:
    pass

class Track:
    pass

acd = CD()
acd.title = "Hity Papa Offenbacha"
acd.price = 5.00
acd.tracks = []

atrack = Track()
atrack.title = "Barkarola"
atrack.duration = 264
acd.tracks.append(atrack)

atrack = Track()
atrack.title = "Piscu Heleny"
atrack.duration = 95
acd.tracks.append(atrack)

atrack = Track()
atrack.title = "Pariduv Soud"
atrack.duration = 247
acd.tracks.append(atrack)

The construction is *quite* unpleasantly wordy.  But what about
the using-code?

def titleLongest(acd):
    tracks = acd.tracks
    longest = 0
    for i in range(1,len(tracks)):
        if tracks[i].duration > tracks[longest].duration
            longest = i
    return tracks[longest].title

Ah, yes, that's better... actually, thanks to the clarity afforded
by the named-attribute syntax, a sightly different approach
suggests itself:

def titleLongest(acd):
    longest = acd.tracks[0]
    for track in acd.tracks[1:]:
        if track.duration > longest.duration
            longest = track
    return longest.title

By looping on the tracks directly, rather than on an index to
them, we achieve a further worthwhile simplification.  It's true
that this approach *could* be taken even with the first of our
proposed representations, but...:

def titleLongest(acd):
    longest = acd[2][0]
    for track in acd[2][1:]:
        if track[1] > longest[1]
            longest = track
    return longest[0]

...the lack of names really kills us here!  This is hardly better
than the form we started out with, and _is_ rather opaque.


Pity about the wordy initialization code... anything we can do
about *that* one...?

Why, sure!  How do you like, for example:

acd = CD(title="Hity Papa Offenbacha", price=5.00, tracks=Tracks(
    ("Barkarola", 264), ("Piscu Heleny", 95), ("Pariduv Soud", 247) ))

Doesn't this look *just right* -- even better than the first attempt,
which had no naming at all, or the second one, which had names
all over the place, because we have names *right where we want to*...?

So how do we get this beauteous init'ing -- is it painful...?

Naah!  The CD class is too easy for words...:

class CD:
    def __init__(self, **kwds):
        for name,value in kwds.items():
             setattr(self,name,value)

Ain't it pretty?  The __init__ method is automatically called by
Python when we call the class object -- the first argument, which
we always name 'self', is the instance object we're building, then
come the others that were passed in the call to the class-object.

Here, we're using the '**somedictionary' notation to receive
whatever arguments were passed with named-form, and then
the handy idioms of 'iterate over dictionary keys/values' and
'set an object attribute' to smoothly, seamlessly translate them
into attributes of the CD object.


For the Tracks idea, as we don't *want* to make 'a sequence of
tracks' into a class-instance of its own (it's one of those cases
where an existing object type -- here, a list -- matches our needs,
so, let's not multiplicate entitites without necessity), we'll just
use a 'factory function' (fancy name for a function which builds
and returns a new object!-).

class Track:
    def __init__(self, title, duration):
        self.title = title
        self.duration = duration

def Tracks(*tracks):
    return [Track(title,duration) for title,duration in tracks]

The list-comprehension plucks and unpacks each 2-items tuple
from the sequence of arguments, and hands the results to the
Track classobject-call, which in turns uses them in __init__ to
initialize the object's attributes.  That's just one way of putting
things, an equally good alternative being:

def Tracks(*tracks):
    return [Track(*track) for track in tracks]

which may be preferable as it has no coupling to what and how
many arguments Track's __init__ wants -- it just passes them on.


So we're smoothly led from 'grouping state' to 'grouping (some)
behavior' -- at least, behavior connected to initialization.  What
could be another example of 'behavioral grouping'...?

What if we often need to refer to "a CD's duration" -- meaning,
the total of its tracks' durations.  Would using a class help here?

Why, sure!  One possibility is to compute 'total duration' in the
initialization, and it's probably the simplest one:

class CD:
    def __init__(self, **kwds):
        for name,value in kwds.items():
             setattr(self,name,value)
        if hasattr(self,'tracks') and not hasattr(self,'duration'):
            duration = 0
            for track in self.tracks: duration += track.duration
            self.duration = duration

As long as nobody changes the .tracks attribute after the
CD object's initialization, we're fine.

What if we're afraid that client-code *might* change the
tracks attribute -- how are we going to keep .duration up
to date then...?

Ah, then we DO need to make .tracks into an object instance
of a special-purpose class... just so we CAN 'hook into' any
modification and report it back to the containing CD object!

I'm not going to show this in detail, as it may get a little bit
involved, but you can look at this sort of things as one first
example of 'utility class' -- although it _does_ have a semantic
aspect, a list-of-tracks would normally suffice, but not if we
have special 'utilitarian' needs related to possible changes to
the sequence, or to some items of the sequence.


Yeah, sigh, doing justice to such considerations *would*
require a book, not just a post.  But the good books I know
about O-O are about statically-typed languages, where it is
all so different... pity, because Python is such a *great*
language in which to learn O-O approaches!  My best advice,
based on books I know, would be 'Structure and Interpretation
of Computer Programs' by Abelson and Sussman, based on
Scheme -- curiously, the same one I suggested for the
purpose of exploring FP (it *is* a great book!-).  Scheme is
not really OO, but then, it *IS* something of a "language
construction kit", and you can build decent OO on top of it
(not as handy as having it built-in, like in Python, but quite
OK to understand its underpinnings!-).  I opine you're NOT
likely to get great understanding, out of a book based on a
statically typed language [particularly one where you never
get to choose between OO and other approaches, such as
Eiffel or Java, where ONLY OO is on offer], about how to best
choose and implement OO in a multi-paradigm dynamically
typed language (such as Python)!


Alex





From eblakes at drwtrading.com  Fri Dec  8 12:15:41 2000
From: eblakes at drwtrading.com (Edward Blakes)
Date: Fri, 8 Dec 2000 12:15:41 -0500
Subject: Implementing IPersistStorage using PythonCOM
References: <t308m58pkecra3@corp.supernews.com> <3A307D8C.10909@ActiveState.com>
Message-ID: <t325urdmtq1t07@corp.supernews.com>

Thanks for responding Mark. I really appreciate your time (as well as your
work with PythonCOM, in the first place). See my comments inline.

"Mark Hammond" <MarkH at ActiveState.com> wrote in message
news:3A307D8C.10909 at ActiveState.com...
> Edward Blakes wrote:
>
> > CATID_PerspectiveAttributeTypeServer = pythoncom.MakeIID("{a guid}")
> > CLSID_PythonSample = pythoncom.MakeIID("{a guid}")
>
> This is probably a problem - it means a new GUID will be created each
> time it is run - definitely not what you want.  Paste generated GUIDs
> directly into the code.
>

My understanding was that this simply converts a string to a PyIID, not
generates a brand new GUID. As you will see below, I directly placed the
string per your suggestion, but to no avail.

> Also - you _really_ should get my book ;-)  The discussion on debugging
> is what you need the most.
>

I have it! That's how I got this far!

> Adding this:
>
I made the changes suggested, got more output but with the same result.
Given your response I believe I clipped important information that, in its
absence, mislead you. So that I won't do that again, I am placing the entire
source here so that you can run it directly. I hate the length but I want
you to have all the information needed.

Begin code.

import struct
import pythoncom
import win32api
import win32ui
from win32com import storagecon
from winerror import S_OK

class DummyAttr:
    SampleProperty = ""

    _public_attrs_ = ["SampleProperty"]
    _public_methods_ = ["DoSomething", "GetClassID", "IsDirty", "InitNew",
"Load", "Save", "SaveCompleted", "HandsOffStorage"]
    _reg_clsid_ =
    _reg_progid_ = "PerspectiveAttribute.PythonSample"
    _reg_catids_ = ["{5A0E5630-78E2-11D2-9D2E-00105A10A8F3}"]
    _com_interfaces_ = [pythoncom.IID_IPersist,
pythoncom.IID_IPersistStorage]

    def __init__(self):
        self.SampleProperty = "This is an initial value"
        win32ui.OutputDebugString("self.SampleProperty = " +
self.SampleProperty + "\n")
    def DoSomething(self):
        import win32ui
        win32ui.MessageBox("Hello from DoSomething", "Message")
    def GetClassID(self):
        return "{D7929F71-133C-4F63-BE0B-BEC0842E7847}"
    def IsDirty(self):
        return 1
    def InitNew(self):
        return S_OK
    def Load(self, pStg):
        try:
            pStream = pStg.OpenStream("ATTRSAVESTREAM",
storagecon.STGM_READ|storagecon.STGM_SHARE_EXCLUSIVE, 0, 0)
            lenData = pStream.Read(struct.calcsize('l'))
            samplePropertyLen = struct.unpack('l', lenData)[0]
            self.SampleProperty = pStream.Read(samplePropertyLen)
            return S_OK
        except:
            win32api.OutputDebugString('Caught exception in
DummyAttr.Load\n')
            pass
    def Save(self, pStg, fSameAsLoad):
        try:
            pStream = pStg.CreateStream("ATTRSAVESTREAM",
storagecon.STGM_READ|storagecon.STGM_SHARE_EXCLUSIVE|storagecon.STGM_CREATE,
0, 0)
            samplePropertyLen = len(self.SampleProperty)
            data = struct.pack('l', samplePropertyLen)
            pStream.Write(data)
            pStream.Write(self.SampleProperty)
            return S_OK
        except:
            win32api.OutputDebugString('Caught exception in
DummyAttr.Save\n')
            pass
    def SaveCompleted(self, pStg):
        return S_OK
    def HandOffStorage(self):
        return S_OK

class LockBytes:
  _public_methods_ = [ 'ReadAt', 'WriteAt', 'Flush', 'SetSize',
'LockRegion', 'UnlockRegion', 'Stat' ]
  _com_interfaces_ = [ pythoncom.IID_ILockBytes ]

  def __init__(self, data = ""):
    import pywintypes
    import time
    self.data = data
    now = pywintypes.Time(time.time())
    self.ctime = now
    self.mtime = now
    self.atime = now

  def ReadAt(self, offset, cb):
    print "ReadAt"
    result = self.data[offset:offset + cb]
    return result

  def WriteAt(self, offset, data):
    print "WriteAt " +str(offset)
    print "len " + str(len(data))
    print "data:"
    #print data
    if len(self.data) >= offset:
      newdata = self.data[0:offset] + data
    print len(newdata)
    if len(self.data) >= offset + len(data):
      newdata = newdata + self.data[offset +  len(data):]
    print len(newdata)
    self.data = newdata
    return len(data)

  def Flush(self, whatsthis=0):
    print "Flush" + str(whatsthis) #+ "Data:" + self.data
    return S_OK

  def SetSize(self, size):
    print "Set Size" + str(size)
    if size > len(self.data):
      self.data = self.data +  "\000" * (size - len(self.data))
    else:
      self.data = self.data[0:size]
    return S_OK

  def LockRegion(self, offset, size, locktype):
    print "LockRegion"
    pass

  def UnlockRegion(self, offset, size, locktype):
    print "UnlockRegion"
    pass

  def Stat(self, statflag):
    print "returning Stat " + str(statflag)
    return (
      "PyMemBytes",
      storagecon.STGTY_LOCKBYTES,
      len(self.data),
      self.mtime,
      self.ctime,
      self.atime,

storagecon.STGM_DIRECT|storagecon.STGM_READWRITE|storagecon.STGM_CREATE ,
      storagecon.STGM_SHARE_EXCLUSIVE,
      "{00020905-0000-0000-C000-000000000046}",
      0,   # statebits ?
      0
      )

def Test():
    import win32com.client
    import win32com.server.util
    import win32com.server.dispatcher

    useDispatcher = win32com.server.dispatcher.DefaultDebugDispatcher

    # Create the server
    origObj = DummyAttr()
    aServer = win32com.server.util.wrap(origObj,
pythoncom.IID_IPersistStorage, useDispatcher=useDispatcher)
    aServerDisp =
win32com.client.dynamic.Dispatch(aServer.QueryInterface(pythoncom.IID_IDispa
tch))

    # Test a server method
    aServerDisp.DoSomething()

    # Test persistence
    aLockBytes = win32com.server.util.wrap(LockBytes(),
pythoncom.IID_ILockBytes, useDispatcher=useDispatcher)
    pStg = pythoncom.StgCreateDocfileOnILockBytes(aLockBytes,
storagecon.STGM_DIRECT|storagecon.STGM_CREATE|storagecon.STGM_READWRITE|stor
agecon.STGM_SHARE_EXCLUSIVE, 0)

    # Show initial property value
    print "SampleProperty = " + aServerDisp.SampleProperty
    aServerDisp.SampleProperty = "1st change"
    print "SampleProperty = " + aServerDisp.SampleProperty

    persistStg = aServer.QueryInterface(pythoncom.IID_IPersistStorage)
    persistStg.Save(pStg, 1)
    origObj.Save(pStg,1)
    aServer.Save(pStg, 1)
    pStg.Commit(storagecon.STGC_DEFAULT)

    aServerDisp.SampleProperty = "2nd change"
    print "SampleProperty = " + aServerDisp.SampleProperty

    persistStg.Load(pStg)
    aServer.Load(pStg)
    print "SampleProperty = " + aServerDisp.SampleProperty

def Register():
    from win32com.server.register import UseCommandLine
    return UseCommandLine(DummyAttr)

if __name__ == '__main__':
    import sys
    if "/test" in sys.argv:
        print "Testing..."
        Test()
    else:
        Register()

When I run the above, I still get the 'member not found' error. While you're
at it, feel free to comment on my use of the struct module. I couldn't
figure out any other way of writing data to streams. Don't know if it will
work b/c of the member problem.

Ed
eblakes at drwtrading.com




From roy at panix.com  Fri Dec 29 15:44:41 2000
From: roy at panix.com (Roy Smith)
Date: Fri, 29 Dec 2000 15:44:41 -0500
Subject: Great New Book
References: <t4o1vr8hktk531@corp.supernews.com> <92ij47$n3p$1@bob.news.rcn.net> <slrn94pnrk.irm.grey@teleute.rpglink.com> <92inl3$a5o$1@panix2.panix.com> <slrn94pp5m.irm.grey@teleute.rpglink.com>
Message-ID: <roy-8A57F2.15444129122000@news.panix.com>

In article <slrn94pp5m.irm.grey at teleute.rpglink.com>, 
morpheus at here.not.there wrote:

> >>    Not to mention 2'5" isn't all that thick.  I have fictional 
> >>    paperbacks
> >>that are that thick.  :)
> 
> >Really?  You have paperbacks more than two feet thick????  (No fair
> >counting all the Robert Jordans as a single book.)
> 
>     Touche'.  2.5".  :)


But I thought in python " and ' meant the same thing? :-)


From monamant at freemail.sx.cn  Thu Dec 21 06:26:09 2000
From: monamant at freemail.sx.cn (monamant at freemail.sx.cn)
Date: Thu, 21 Dec 2000 11:26:09 GMT
Subject: please help
Message-ID: <91spce$tc1$1@nnrp1.deja.com>

Hi,
  I am just thinking about taking Python as my first programming
language for serious learning (after reading "how to become a hacker"),
but some distrusts with python's applicability prevent me from fully
applying myself to it. Would you kindly help clear up some of my doubts
about Python?
1. I want to use IIS 4.0 for my webserver, can I use scripts compiled
by Python in it?
2. What about the ASP? If I want to use ASP, can I use Python to do
such jobs like user authentication? How well can Python be used in Web
programming on Windows platform?
3. I used VB in the past and you can make a '.exe' file after
compiling. But how can I actually execute the Python program in
Windows? It seems that I can not make a .exe file for it to execute. It
is dependent on the interpreter?
4. If I really sit my self to Python, does it reward equally for the
same effort in learning C? In the aspect of the quality of programs.
Thank you in advance for answering my questions.
An eager learner waiting for reply

(P.S if I begin my serious learning process, I will translate the
tutorial into Chinese.)


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


From arcege at shore.net  Fri Dec 22 14:56:52 2000
From: arcege at shore.net (arcege at shore.net)
Date: Fri, 22 Dec 2000 19:56:52 GMT
Subject: Python interpreter in python?
References: <91ndgd$g8d$1@bunyip.cc.uq.edu.au>
Message-ID: <8mO06.1503$JG.176924@news.shore.net>

Chui Tey <teyc at bigfoot.com> wrote:
: Is there a python implementation of the python VM? Basically for
: experimenting?

I would start first by looking at the code module, it is a full Python
interpreter written in Python.  If I remember correctly, there was talk
a while ago that Python 2 (P3K?) would have a Python-coded interpreter
instead of C based.  Guido wrote the Idle interactive shell with this
module (c.f. Tools/idle/PyShell.py).

If you still want to look into emulating the VM, then first look into
the parser modules and how it builds the VM instructions (the "Python
Language Services" in the Library Reference manual).  From there, you
can more easily handle the byte-codes and interpret data.  I would
especially look at section 17.1.6.1 "Emulation of compile()" (while not
a large section, it gives you pointers on where to go).  Then from
there, look at writing eval/exec in Python.

  -Arcege



From db3l at fitlinxx.com  Mon Dec  4 15:54:29 2000
From: db3l at fitlinxx.com (David Bolen)
Date: 04 Dec 2000 15:54:29 -0500
Subject: General default exception handler
References: <lj0n2to9vs8bmv3oo49mqgoc6h7esh8t37@4ax.com> <NZMW5.10915$lF4.165334@e420r-atl1.usenetserver.com> <138n2tojuquau637pm7obq8c4gvtl7hutd@4ax.com>
Message-ID: <uhf4jkiuy.fsf@ctwd0143.fitlinxx.com>

Dale Strickland-Clark <dale at out-think.NOSPAMco.uk> writes:

> Interesting idea but that won't work in this case. I have to trap
> the error inside the COM object or much of the detail is lost.

Then how about just structuring your COM object so that the various
entry points just pass control through some sort of dispatch function
that can provide the single general purpose wrapper for unhandled
exceptions.  Python's ability to pass in a dynamic function reference
and arbitrary parameters through such a dispatch should make this
fairly straight-forward.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From aleaxit at yahoo.com  Thu Dec 14 07:25:51 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 14 Dec 2000 13:25:51 +0100
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <tcmay-FCA22F.14073713122000@news.got.net> <918v7a02scn@news1.newsguy.com> <tcmay-04E80C.16181313122000@news.got.net> <91a1ed02a4j@news2.newsguy.com> <3A389B1A.4F33FFD4@cfmu.eurocontrol.be>
Message-ID: <91ae8s02llf@news2.newsguy.com>

"Ian Wild" <ian at cfmu.eurocontrol.be> wrote in message
news:3A389B1A.4F33FFD4 at cfmu.eurocontrol.be...
> Alex Martelli wrote:
> >
> > I can't find a source any more, but one of the most interesting
> > arguments I've heard for "only inherit from abstract classes"
> [...otherwise...]
> > it would break the idea that your concrete
> > classes partition the universe of discourse, if it were.
>
> Where does this partitioning requirement come from?

>From the fact that an object 'is an instance' of one,
and only one, class -- 'THE class' (singular) 'of the
object'.  If an object 'belongs' to n>1 classes, in
this framing, this is modeled by having the (inevitably
singular) class 'of the object' _inherit_ from the
said n>1 classes (i.e., 'belongs' is a weaker relationship
than 'being an instance of').

I am not aware of any OOA approaches in which is-instance-of
(as here defined) isn't an exhaustive and many-to-one
relationship between objects and (concrete) classes;
in other words, OOA approaches which let some objects
(in the universe-of-discourse that is being modeled)
be an instance of 'no class', or of several classes
at the same time.


> > This is based on the idea of subclassing as asserting IS-A.
>
> ?

IS-A implies several important things, among which is
the fact that the constraints on an object for subclass
membership are at least as strong as those for base
class membership (in Eiffel terms, class-invariants
are at least as strong for subclasses as for base
classes).


> It seems to me you'd need an "(e) None of the above"
> class very frequently if your world doesn't happen to have
> any genuine structural divisions.
>
> Dragging in that old favourite, "Squares and Rectangles",
> your partitioning clause means I'd have to invent a new class,
> "NonSquareRectangle", adding no interesting features, just so
> I can have aSquare behave like aRectangle.  This seems like
> unnecessary work, just to keep some C++ guru happy.

Forget C++ -- we're talking OO _analysis_.  Assume you're
trying to perform OOA on the universe of discourse "immutable
planar 2D geometric polygons", for the sake of definiteness.

If Rectangle and Square are both concrete (instantiable)
classes, with *overlapping* requirements for membership,
how do you classify a given rectangular object which
happens to have equal sides?  Toss a coin?

If you always classify such objects as Square, then you
have a constraint on actual membership to Rectangle _which
is not expressed in your OO model_ -- the fact that the
sides are not equal.  It's inconsistent to claim that
the constraint about equality of sides is 'no interesting
feature' -- if side-equality was of no interest, then
you would not be singling out squares!

You can't place the constraint (on sides being unequal)
on Rectangle itself, which is a baseclass of Square --
because Square would inherit the constraint, which is
inconsistent with _its_ intended constraint (equal sides).

So, you have a feature (constraint) which IS of interest
(you ARE interested in side-equality -- else, forget
squares!) but cannot be in the baseclass -- therefore,
in an OOA view, you must inherit and add the feature
(constraint) in the subclass.  So, we have an abstract
baseclass Rectangle (which knows and cares nothing
about its sides being equal or not) and concrete
subclasses Square (constraint: equal sides) and
NonsquareRectangle (constraint: unequal sides).

(If an object which happens to be a Rectangle AND have
equal sides ends up classified as EITHER a Square OR
a more generic Rectangle depending on accidental issues,
such as what path-of-construction was followed, you'll
get all of the usual darned issues to which CSG is so
prone to -- without even having to go 3D for them!-).


If you're modeling *mutable* objects, which is more
usually the case except in functional-programming
contexts, then the case is even sharper.  Square has
a constraint *on future mutation* -- its sides not
just 'are equal' at a given time, they're constrained
to *remain* equal as the object mutates (else, the
*class itself* of the object would have to switch --
and we've already clarified that we're treating classic
OOA, where objects belong to given classes "for the
duration", so changing class at runtime is not ok:-).

NonsquareRectangle has no such constraint, and will
allow each parallel pair of its sides to vary freely,
independently of the _other_ parallel pair of sides.

This is best modeled by giving the two concrete
classes in question different mutator-methods:
Square will have, say, SetSide, taking a single
positive number argument; NonsquareRectangle will
have SetSides, taking _two_ positive numbers as
arguments.  In this case, a NonsquareRectangle
whose sides happen to be set equal at a given time
will not protest -- it keeps its distinguishing
characteristic from Square, which is, there is
NO guarantee/constraint against *further* changes
preserving the current 'accidental' equality.

(I consider mutability/immutability need to be
addressed pretty early on in any OOA task -- they
are HIGHLY crucial to determine what we're trying
to accomplish, and how we're going to set about
it...!-).

The (clearly abstract!) common superclass Rectangle
may be factored to present methods common (with
different implementation) to both subclasses,
such as a GetSides accessor returning a Pair of
non-negative numbers, and also, no doubt, many
other nice things (accessors/mutators for center,
slant, perimeter, area, etc, etc).  Or, there may
be further abstract classes up the hierarchy, with
Rectangle inheriting from FixedNumberOfSides which
in turn subclasses Polygon (assuming we want to
model some kinds of polygons whose number of sides
is allowed to change during runtime, while others
have a fixed number of sides -- Rectangle will then
clearly fall among the latter) -- the various
methods will be exposed at appropriate levels in
the hierarchy.

Here's where MI peeks in, by the way -- we might
clearly like to have a RegularPolygon abstract
subclass of Polygon, right?  Then, it would be
most handy for Square to inherit, multiply, from
both RegularPolygon AND Rectangle.  Clearly, as
is quite reasonable in an analysis phase, this is
more about inheritance of interface, than of
implementation -- Square, EquilateralTriangle,
and so on, may not in fact use many (or any)
data held in either (or both) of their supers
(RegularPolygon for both, plus either Rectangle
or Triangle); but we do want to be able to pass
a reference to either Square or EquilateralTriangle
to any method requiring a reference to any
RegularPolygon -- and we also want to pass a
reference to Square to any method requiring a
reference to a Rectangle; the interfaces (which
include some possibly-abstracted versions of the
class-invariants) must thus be 'implemented' (as
commonly framed -- in OOA, in Eiffel, in C++ --
'inherited').

When you move to implementation issues, the
possible overheads of inheriting-from-concrete
can be heavy.  A Polygon (if concrete) needs
a generic sequence (array, list, whatever) of
Sides (or, equivalently, Vertices); a concrete
Hexagon needs to fix that sequence's multiplicity
to exactly six; a RegularHexagon would need only
1/6 of that -- either you fudge by not expressing
the 6-sides constraint on Hexagon (which feels rather
funny, doesn't it?-), or you waste 80% of your
store on duplicate data... if most of your polygons
are many-sided regular ones, that can kill you!-)

But that's implementation, and a well known
problem with inheriting-from-concrete in that
realm.  What I'm saying is that the problems with
inheritance-from-instantiable (fully concrete)
classes start much earlier than at implementation
time -- start with conceptual, ontological issues
regarding OOA and related classification tasks.


Alex





From henrik.martensson at frontconsulting.se  Thu Dec 21 07:09:00 2000
From: henrik.martensson at frontconsulting.se (Henrik Mårtensson)
Date: Thu, 21 Dec 2000 12:09:00 GMT
Subject: Q: Is there any good documentation on MAPI using Python
Message-ID: <wpm06.791$AH6.82177@newsc.telia.net>

Hi

I would like to use the MAPI with the MS office suite programming in Python,
but I'm having a little bit of a problem with finding good documentation.

Does anybody have any leads?

/Henrik




From mwh21 at cam.ac.uk  Mon Dec 11 10:26:38 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 11 Dec 2000 15:26:38 +0000
Subject: Modifying running code? [HACK ALERT]
References: <912m65$pu1$1@news.rz.uni-karlsruhe.de>
Message-ID: <m3u28b2d3l.fsf@atrus.jesus.cam.ac.uk>

Bjoern Giesler <un4e at rzstud1.rz.uni-karlsruhe.de> writes:

> Hi,
> 
> I'm writing a kind of mini runtime environment that lets me debug / correct
> buggy Python code "on the fly". This is supposed to work as follows:
> 
> 1) an exception gets thrown by the buggy code and is caught by a toplevel
> exception handler
> 
> 2) the toplevel handler figures out the function that caused the exception
> 
> 3) the function source is loaded into an editor and can be corrected by the
> user
> 
> 4) if the code has been saved, it is recompiled and inserted instead of the
> original function
> 
> 5) the program is then resumed at the point where the original function had
> been called
> 
> Can this be done in Python at all? Can it be done in regular Python or only
> in Stackless Python (which supports continuations and would therefore, I
> believe, make step 5 easy)? Is the whole thing complete bull doodoo? Any
> hints or suggestions?

You'd certainly have to write some C to get round the fact that frames
are immutable in Python & starting with stackless Python is probably
wise and you'd need to write most of it anyway.

I'm not sure what you're trying to do is particularly wise, but that's
your concern, not mine :-)

Cheers,
M.

-- 
41. Some programming languages manage to absorb change, but 
    withstand progress.
  -- Alan Perlis, http://www.cs.yale.edu/homes/perlis-alan/quotes.html


From erno-news at erno.iki.fi  Thu Dec  7 02:08:43 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 07 Dec 2000 09:08:43 +0200
Subject: Bug: Readline getting stuck on Linux and Solaris
References: <lxk8zqwupy2.fsf@ioasun3.epfl.ch> <90m53k$ovh$1@nnrp1.deja.com> <m3u28h48ja.fsf@atrus.jesus.cam.ac.uk> <slrn92u8al.76b.quinn@lira.ugcs.caltech.edu>
Message-ID: <kuk89cbtdw.fsf@lasipalatsi.fi>

In article <slrn92u8al.76b.quinn at lira.ugcs.caltech.edu>,
quinn at lira.ugcs.caltech.edu (Quinn Dunkan) writes:

| Strange.. readline has *never* worked right for me, on IRIX or
| Linux, on any version of python (1.4, 1.5.2, 2.0), and any version
| of readline (3.0, 4.0, 4.1), if python is compiled with threads.
| IRIX (6.4, 6.5, 6.5.2) hangs, Linux (2.0.36, 2.2.12, 2.2.17) exits.

most odd - approximately all linux distributions ship with python compiled
with readline, and i've never seen any problems in those or manual
builds.

  -- erno

ps. anyone know if there's anything new wrt the license story?  last i
heard the respective lawyers were still out on the subject. it would
be nice to be able to distribute readline-enabled binary python builds
legally...


From richard at NOstarfighterSPAM.freeuk.com  Tue Dec  5 14:26:37 2000
From: richard at NOstarfighterSPAM.freeuk.com (Richard)
Date: Tue, 5 Dec 2000 19:26:37 -0000
Subject: Pmw.CounterDialog
References: <90g351$prk$1@nnrp1.deja.com>
Message-ID: <anbX5.16289$R77.1305391@nnrp4.clara.net>

Many thanks for your solution to this problem, I'd given up hope of getting
a reply :-)

Richard

<rhamorim at my-deja.com> wrote in message news:90g351$prk$1 at nnrp1.deja.com...
> Hi, to correct it is simple. You'll have to edit PmwCounter.py. Here it
> is:
>
> - Find the function called _changeNumber() (line 273)
> - On line 279 there's the return statement for this function. Change it
> to return str(value) (and not return str(value)[:-1])
>
> That ocurred because str(long) on Python 2.0 doesn't return a number
> with L on the end like Python 1.5.2 did.
>
> Tell me if you have any further problems.
>
> Roberto Amorim
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.




From Blair.Lowe at compeng.net  Mon Dec  4 18:06:48 2000
From: Blair.Lowe at compeng.net (Blair Lowe)
Date: Mon, 4 Dec 2000 16:06:48 -0700
Subject: SSL with python
In-Reply-To: <3A29DBF6.6A074D47@tridog.com>
References: <9097pc+pspc@eGroups.com>
 <v04220804b64f0e47b847@[192.168.11.33]> <3A29DBF6.6A074D47@tridog.com>
Message-ID: <v04220807b651d381086e@[192.168.11.33]>

Hi again,

I emailed the author of the ssl_wrapper python lib: Val Shkolnikov 
<nvsoft1 at worldnet.att.net>.

Hopefully we can get it from him.

Maybe he reads this list too.

Blair.

At 22:36 -0700 2000/12/02, Ken Kinder wrote:
>  >From a simple search of http://www.vex.net/parnassus/ -- the 
>premier python module
>directory, I found a few. I suggest just going through the results yourself.
>http://www.vex.net/parnassus/apyllo.py?find=ssl
>
>ssl_wrapper 1.1 looks very promising but it's a broken link. Does anyone know
>where a mirror of that one is?
>
>Blair Lowe wrote:
>
>>  -----BEGIN PGP SIGNED MESSAGE-----
>>
>>  Hi,
>>
>>  I am new with python, and I was wondering if anyone has got it to
>>  work with SSL connections.
>>
>>  We are trying to do some socket level manipulation using urllib, and
>>  httplib, and we are getting some messages saying that we are using
>>  straight http over an SSL connection.
>>
>>  Does anyone know of any information about SSL and python (web or 
>>printed work)?
>>
>>  Thanks,
>>  Blair.
>>  Computer Engineering Inc. http://www.compeng.net
>>  Phone: 780 499 5687 (9 - 5 MST) Fax:   780 435 0693 (24 Hours)
>>
>>  -----BEGIN PGP SIGNATURE-----
>>  Version: PGPfreeware 6.5.8 for non-commercial use <http://www.pgp.com>
>>
>>  iQCVAwUBOilgDUZu3Sp72OwBAQFHRQP8CKBpVM8Wll7N2DSjbhtgueqsWohYkjmc
>>  4nymzTfsTPGebMjjE+AnozMe+yzUvpSaXedABw3vGfJME/r807H/uIOrZbMLYhsw
>>  gp0fWL06DJpcNRuAYLHZ6QTD5VjqWxsgm8BZSNA3vzR2I40JYh5BOaVOVjy9S4mA
>>  alP+cyWsFA0=
>>  =38Vk
>>  -----END PGP SIGNATURE-----
>>
>>  --
>>  http://www.python.org/mailman/listinfo/python-list
>
>
>--
>http://www.python.org/mailman/listinfo/python-list

Computer Engineering Inc. http://www.compeng.net
Phone: 780 499 5687 (9 - 5 MST) Fax:   780 435 0693 (24 Hours)



From neelk at alum.mit.edu  Tue Dec 26 19:28:39 2000
From: neelk at alum.mit.edu (Neelakantan Krishnaswami)
Date: 27 Dec 2000 00:28:39 GMT
Subject: One Python 2.1 idea
References: <mailman.977771949.17911.python-list@python.org> <92ab9u$dou$1@nnrp1.deja.com> <92ap0i22b6r@news1.newsguy.com> <slrn94ie9k.anf.neelk@alum.mit.edu> <92bamo$roe$1@panix3.panix.com>
Message-ID: <slrn94ihni.anf.neelk@alum.mit.edu>

On 26 Dec 2000 15:47:36 -0800, Aahz Maruch <aahz at panix.com> wrote:
>In article <slrn94ie9k.anf.neelk at alum.mit.edu>,
>Neelakantan Krishnaswami <neelk at alum.mit.edu> wrote:
>>
>>I'm not sure this would help very much. Tim posted a link to an
>>attempt to add method caches, and reported inconclusive results.  As
>>of late last week, I'm not surprised; the Python interpreter seems to
>>be spending most of its time in bytecode instruction dispatch. Logic
>>suggests that that's where performance enhancements would have the
>>most impact.
>>
>>Fortunately, I don't have to put my money where my mouth is, since
>>someone has already done it for me. :) Vladimir Marangozov has
>>modified the Python interpreter (as of 1.5) to use the GNU computed
>>labels extension and implemented direct threading.
>
> Would making this standard require a Python installation to use a GNU
> compiler?  Or is this just a library that can be linked in?

It's a compile-time option.

The big while-loop in ceval.c has been macrofied, and the definitions
of the macros are different depending on whether gcc is available or
not. If gcc is available, then computed jumps are used, and if not it
expands into the regular while loop. (I understand this is the usual
thing to do when writing portable direct-threaded interpreters in C.) 
After compilation I don't think there is any effect except speed on
how Python behaves for users and C extensions.

A webpage on what threading is (it's unrelated to multithreading, just
a case of overloaded terminology) and to what extent it is a good
thing can be found at:

  http://www.complang.tuwien.ac.at/forth/threaded-code.html

This site is actually where I found the link to Vladimir Marangozov's
webpage. 


Neel


From donn at u.washington.edu  Fri Dec  8 14:46:36 2000
From: donn at u.washington.edu (Donn Cave)
Date: 8 Dec 2000 19:46:36 GMT
Subject: SIGCHLD handle & socket
References: <90q16m$tm2$1@nnrp1.deja.com>
Message-ID: <90rdqs$8bcg$1@nntp6.u.washington.edu>

Quoth redjade at my-deja.com:
| I am writing a forking-type server with python.
| My plan is:
| 1) first, open a socket with port 15151
| 2) if there's a connection request, fork a child
| 3) child will die after 3 sec of connection establishment.
| 4) parent should handle SIGCHLD signal so that it make zombie-child
| disappear.
| 5) parent should allow any connection request and should be fork-able.

As another followup already mentions, an interrupt exception is
inevitable in this scheme.  In C, you'd check for errno == EINTR
after every error return from a socket, pipe or tty I/O function
and just retry.  In Python, you have to trap an exception.

If it's at all possible, you will have better luck if you don't
trap SIGCHLD and find some other way to wait() for the child
processes.  If you don't trap this unusual signal, it won't be
delivered and you won't have those interrupts.  Even in C,
they're rarely worth the hassle, and with Python's extra signal
handling issues plus the exception business, surely not worth it.

If you want to take all the exited child processes on a regular
schedule, you can time out on the bound socket, for example -
use select() to check for connection ready to accept() or time out
after some fraction of a second.  Use posix.waitpid(-1, posix.WNOHANG)
to check for any and all outstanding exited processes.

	Donn Cave, donn at u.washington.edu


From syver at NOSPAMcyberwatcher.com  Sat Dec 23 16:24:40 2000
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Sat, 23 Dec 2000 22:24:40 +0100
Subject: About WebBrowser
References: <91vbr9$eo1q2@SGI3651ef0.iddeo.es>
Message-ID: <9235bj$6e6$1@troll.powertech.no>

"Juan Huertas" <jhg at galdon.com> wrote in message
news:91vbr9$eo1q2 at SGI3651ef0.iddeo.es...
> It's possible to print the content of a Html page using the webbrowser
> module into a python module?

I am not sure what you're thinking about here. But I'll try, if it's
possible to print with the webbrowser COM component from the dispatch
interface it presents it is possible to do it from Python, but is that what
you're asking for?




From news at dorb.com  Mon Dec 11 22:49:50 2000
From: news at dorb.com (Darrell)
Date: Tue, 12 Dec 2000 03:49:50 GMT
Subject: Fault trees structure
References: <D_IY5.49$17.2785@m2newsread.uni2.es> <9128e902uo3@news2.newsguy.com> <z8eZ5.249$YZ4.8868@m2newsread.uni2.es>
Message-ID: <yfhZ5.23310$Ia1.4003773@typhoon.nyroc.rr.com>

Jumping right to code :)
Seemed more fun than reading too deeply about fault trees.

Looks like we need shift registers, flip-flops and ...

--Darrell

class NodeBase:
    def name(self):
        return self._name

    def connectInput(self, name):
        if not hasattr(self, "_input"):
            self._input={}
        self._input[name]=None

    def connectOutput(self, otherNode, name):
        if not hasattr(self, "_output"):
            self._output={}
        self._output[otherNode.name()]= name
        otherNode.connectInput(name)

    def signal(self, collection, name, value=1):
        self._input[name]=value
        self.logic(collection)

    def propagate(self, collection, value):
        for nodeName, inputName in self._output.items():
            try:
                node=getattr(collection, nodeName)
                node.signal(collection, inputName, value)
            except:
                print self._name
                print nodeName
                print collection
                raise

class AndNode(NodeBase):
    def __init__(self, name):
        self._name=name

    def logic(self, collection):
        cc= reduce(lambda x,y: x and y, self._input.values())
        self.propagate(collection, cc)


class OrNode(NodeBase):
    def __init__(self, name):
        self._name=name

    def logic(self, collection):
        cc = reduce(lambda x,y: x or y, self._input.values())
        self.propagate(collection, cc)

class AnnounceNode(NodeBase):
    def __init__(self, name):
        self._name=name

    def logic(self, collection):
        cc = reduce(lambda x,y: x or y, self._input.values())
        if cc:
            print self._name, self._input
        else:
            # Noise on the input, but nothing to report
            print '.',

class SourceNode(NodeBase):
    def __init__(self, name):
        self._name=name

    def __call__(self, collection, value):
        self.propagate(collection, value)


class FaultTree:#(UserDict.UserDict): Not sure about this
    def __init__(self, *nodes):
        for n in nodes:
            setattr(self, n.name(), n)

ft=FaultTree(AndNode("and1"), AndNode("and2"), AndNode("and3"),
AnnounceNode("fail"))

ft.and1.connectOutput(ft.and3,"one")
ft.and2.connectOutput(ft.and3,"two")
ft.and3.connectOutput(ft.fail,"who cares")

    # Setup some signal sources
drivers=[]
for x in range(3):
    s1=SourceNode(x)
    drivers.append(s1)
    s1.connectOutput(ft.and1, x)

    # Setup some signal sources
for x in range(3):
    s1=SourceNode(x)
    drivers.append(s1)
    s1.connectOutput(ft.and2, x)

    # Make it go
for i in drivers:
    i(ft, 1)


#### output
>faulttree.py
. . . . . fail {'who cares': 1}

Each "." is an input being propagated though the graph.
Unitl the output AnnounceNode fires with the "fail" message.
"who cares" is the name of the input connection of the AnnounceNode

"Arturo P?rez Mulas"  wrote:
> Fault Tree are the following:
>
> - They have only ONE top node
>
> - Each node can have unlimited inputs (opposed to two allowed for binary
trees)
>
> - Nodes can be of any of the following different types:
>
> a) 'Gates' : are nodes that admit further inputs; there are different
types
> of gates,  AND, OR, XOR, n/m, NOT but this is not important at this point,
> since  from the tree structure they are all similar (except type n/m
> gates which   must have exactly 'm' inputs)
>
> ....





From donn at u.washington.edu  Wed Dec 20 20:30:39 2000
From: donn at u.washington.edu (Donn Cave)
Date: 21 Dec 2000 01:30:39 GMT
Subject: while true: !!!
References: <3A4024E5.98843CFC@cosc.canterbury.ac.nz> <91qadb$1di$1@plutonium.compulink.co.uk> <3A41539E.DF6977D6@cosc.canterbury.ac.nz>
Message-ID: <91rmfv$oag$1@nntp6.u.washington.edu>

Quoth Greg Ewing <greg at cosc.canterbury.ac.nz>:
| gbreed at cix.compulink.co.uk wrote:
|> 
|> Take any type, and if
|> there's a way of it being nothing, it evaluates false.
|
| Hmmm, let's try that out. Is [[]] true or false? 
| It's a list containing a list containing nothing. 
| Since [] is false, a list containing nothing must be
| nothing. So [[]] is a list containing nothing,
| which is nothing. Ah, it must be false! :-)

I sure hate to prolong the life of this miserable thread,
but that's too silly.  [[]] is not a list containing nothing,
it's a list containing [].  [] is not nothing, it's something
that's false.

	Donn Cave, donn at u.washington.edu


From aleaxit at yahoo.com  Tue Dec 26 04:43:08 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 26 Dec 2000 10:43:08 +0100
Subject: Duplicated List...
References: <929lq7$ds1$1@hiline.shinbiro.com>
Message-ID: <929pan016bf@news1.newsguy.com>

"dyo" <dyo at shinbiro.com> wrote in message
news:929lq7$ds1$1 at hiline.shinbiro.com...
> Hi everyone.
>
> >>> a=[1,2,3]
> >>> b=[a,10,20,30]
> >>> b[0].append('x')
> >>> b
> [[1, 2, 3, 1, 2, 3, 'x'], 10, 20, 30]

No way.  You must be doing something different if this
is the output you observe -- e.g. the first statement
must be

>>> a = [1,2,3] * 2

or something similar.


Anyway, after these statements, the reference named
'a' and the one at index 0 in list b are currently bound
to the same object.  Besides modifying that object, a
simple way to check is with the id function:

>>> a=[1,2,3]
>>> b=[a,4,5]
>>> b
[[1, 2, 3], 4, 5]
>>> id(a)
7930812
>>> id(b[0])
7930812

The id that comes out for you will be arbitrary, of course,
but it will be the same for both references -- as they're
bound to the same object.  OK so far?  Whatever mutators
you call on that object, it keeps its identity, and those
mutator methods do NOT affect *the references*, but
rather *the object*:

>>> a.append('x')
>>> a
[1, 2, 3, 'x']
>>> b
[[1, 2, 3, 'x'], 4, 5]
>>> id(a)
7930812
>>> id(b[0])
7930812

.append is a mutator, the object's current state changes,
but of course its identity doesn't, and nor do the references
that currently happen to be bound to it.


Mutating an object of a given identity is an utterly different
and unrelated issue from binding and rebinding references
to objects...:

> >>> a=[1,1000]

This rebinds the named-reference (aka 'variable') "a"
to something completely different, severing previous
bounds of that reference.

>>> a=[1,1000]
>>> id(a)
7930556
>>> id(b[0])
7930812

See?  *a*, the named-reference, is affected: it is now
bound to an object with a different identity.  The object
that was previously referred-to by a is supremely
unaffected and indifferent, as are any other references
to it that were previously around -- they all still are.


> >>> a
> [1, 1000]
> >>> b
> [[1, 2, 3, 1, 2, 3, 'x', 'xxx'], 10, 20, 30]
> >>>
>
> Why do not change List 'b' ?
> What happen List 'b' ?

b[0] still refers to whatever it was referring to before a
was re-bound.

If you think of '=' as "assignment", as the docs call it,
and know about 'assignment' from such languages as
C, Fortran, Pascal, you're never going to make heads or
tails of languages with reference-semantics, such as
Python or Java. That's why I prefer to call '=' "rebinding".

[although THAT irks people whose backgrounds are in
other languages such as LISP -- I'll freely admit that
Python's "=" is close to a LISP setq, which is normally
called 'assignment', not to a LIST let...]

There are objects, and there are references to objects
(which live in various 'slots', some named, some indexed;
no real difference).  Mutator methods affect objects;
rebinding ('assignment') affects references.  Keep this
distinction in mind, and everything will eventually fall
into place...


Alex





From grey at despair.rpglink.com  Thu Dec 14 16:08:43 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Thu, 14 Dec 2000 21:08:43 -0000
Subject: Spinning off time intensive procedures to speed CGI
References: <916bpn$a6i$1@nnrp1.deja.com> <spenrose-A548F5.21552413122000@news.dnai.com> <slrn93ic42.cg1.quinn@dinar.ugcs.caltech.edu>
Message-ID: <slrn93idmr.ss9.grey@teleute.rpglink.com>

On 14 Dec 2000 20:41:38 GMT, Quinn Dunkan <quinn at dinar.ugcs.caltech.edu> wrote:
>Sending email doesn't so much depend on the speed of your server as the speed
>of the dns lookup, the internet between you and the remote host, and the
>remote host.  

    Uhm, if anyone is implementing an MTA in their CGI they deserve any
latency they get.  Otherwise none of these are an issue when it comes to
opeming the mta and having it send the message; a simple system call.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From heiland at ncsa.uiuc.edu  Thu Dec 21 20:12:40 2000
From: heiland at ncsa.uiuc.edu (Randy Heiland)
Date: Thu, 21 Dec 2000 19:12:40 -0600
Subject: JPython embed question
Message-ID: <3A42AA88.D9BD321A@ncsa.uiuc.edu>

In the JPython demo/embed example, we find how to create an instance of
a class, e.g.:

  interp.set("foo1", new Foo());
  interp.set("bar1", new Bar());

I'd like to be able to dynamically create an instance of a class - e.g.,

  createInstance(String instName, String className)

thanks,
--Randy



From wware at world.std.com  Mon Dec 11 19:11:27 2000
From: wware at world.std.com (Will Ware)
Date: Tue, 12 Dec 2000 00:11:27 GMT
Subject: Python-like compiled language
References: <913jr3$dmv$1@sylvester.vcn.bc.ca>
Message-ID: <G5FHv5.7Hy@world.std.com>

Alexander Lawhead (alawhead at vcn.bc.ca) wrote:
> I'm a new programmer...
> trying to find a Python-like compiled language that
> would be faster than Python+NumPy.

It usually turns out with software that most of the time is spent
in one, or a very few, time-critical portions. This means that the
rest of the code is relatively insensitive to your efforts to speed
it up, and those efforts (outside the narrow time-critical regions)
will be time wasted. So programmers have developed tools called
profilers to determine where code is spending its time. Before you
undertake a Herculean challenge of learning a new language and
rewriting your entire program in it, you might want to profile the
code you're using now, and look at the parts of it that are taking
the most time to see if they can be sped up.

I haven't studied NumPy and I don't know anything about your application
but I suspect that the authors of NumPy have gone to great pains to
optimize the things that are likely to take up time in your code.
There might not be much to gain by throwing away Python and replacing
it with compiled code. This is actually not such bad news: as long as
you stick with Python, you have all the comforts and conveniences
Python offers you, many of which will disappear if you go to a fully
compiled situation.

It might make sense to contact the authors of NumPy to discuss where
further optimizations could be fruitful. Konrad Hinsen and Paul Dubois
spring to mind as likely suspects.

As far as a language recommendation, mine would be either C or C++.
Not too long ago, Fortran compilers were still significantly better
than available C compilers, but I think C/C++ has been catching up
(beware of ignorance; I never use Fortran). In any case, C and C++
are wildly popular languages throughout the computer science world,
whereas Fortran has retreated to various niches such as scientific
calculation, and its days are probably numbered even there. And if
none of those reasons interests you, Python's source code (and the
source code for many other non-C languages) is written in C.

-- 
# - - - - - - - - - - - - - - - - - - - - - - - -
# Resistance is futile. Capacitance is efficacious.
# Will Ware	email:    wware @ world.std.com


From cookedm at physics.mcmaster.ca  Sat Dec 30 22:13:52 2000
From: cookedm at physics.mcmaster.ca (David M. Cooke)
Date: 30 Dec 2000 22:13:52 -0500
Subject: HTML Parser
References: <mailman.978224102.28797.python-list@python.org> <0gw36.45$9s1.9126@e420r-atl2.usenetserver.com>
Message-ID: <qnkito1p9jj.fsf@arbutus.physics.mcmaster.ca>

At some point, kragen at dnaco.net (Kragen Sitaker) wrote:

> In article <mailman.978224102.28797.python-list at python.org>,
> Voitenko, Denis <dvoitenko at qode.com> wrote:
> >HTMLtags=re.compile('<.*>')
> 
> In a string like "x<a>b<c>d", this will match "<a>b<c>", because the .*
> matches "a>b<c".  This explains your problem.
> 
> Fixing it is harder.

Not that hard: use the pattern '<.*?>'. This exact pattern is used as
an example in the documentation for the regular expression syntax for
the re module. Here's the relevant paragraph:

*?, +?, ??
  The "*", "+", and "?" qualifiers are all greedy; they match as much
text as possible. Sometimes this behaviour isn't desired; if the RE
<.*> is matched against '<H1>title</H1>', it will match the entire
string, and not just '<H1>'. Adding "?" after the qualifier makes it
perform the match in non-greedy or minimal fashion; as few characters
as possible will be matched. Using .*? in the previous expression will
match only '<H1>'

-- 
|>|\/|<
/--------------------------------------------------------------------------\
|David M. Cooke
|cookedm at mcmaster.ca


From sdm7g at virginia.edu  Sat Dec 23 00:25:40 2000
From: sdm7g at virginia.edu (Steven D. Majewski)
Date: Sat, 23 Dec 2000 00:25:40 -0500 (EST)
Subject: A Suggestion for Python Colon Syntax
In-Reply-To: <LNBBLJKPBEHFEDALKOLCEEDDIFAA.tim.one@home.com>
Message-ID: <Pine.NXT.4.21.0012230024000.233-100000@localhost.virginia.edu>


On Fri, 22 Dec 2000, Tim Peters wrote:
> 
> I expect the PEP will be rejected, but that doesn't mean it will be.  Guido
> did surprise me once, in about 1995 <wink>.
> 

Tim, 
 Please save me from searching thru a years worth of python-list archives
 and tell me what that 1995 surprise was! 

--- Steve Majewski <sdm7g at Virginia.EDU>





From mwh21 at cam.ac.uk  Tue Dec 12 18:53:56 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 12 Dec 2000 23:53:56 +0000
Subject: Bug: Readline getting stuck on Linux and Solaris
References: <lxk8zqwupy2.fsf@ioasun3.epfl.ch> <90m53k$ovh$1@nnrp1.deja.com> <m3u28h48ja.fsf@atrus.jesus.cam.ac.uk> <slrn92u8al.76b.quinn@lira.ugcs.caltech.edu> <m3r93k4kqk.fsf@atrus.jesus.cam.ac.uk> <slrn932eoj.d1j.quinn@ngwee.ugcs.caltech.edu>
Message-ID: <m3aea12o2z.fsf@atrus.jesus.cam.ac.uk>

quinn at ngwee.ugcs.caltech.edu (Quinn Dunkan) writes:

> It's not tricky on an Indy :)  In any case, on the irix machine, I've been
> using 2.0-for-real and it still locks.  It apparently works ok if I
> hit control C quickly: I get 'import site' failed, presumably because it died
> of a KeyboardInterrupt.  However, even a 'while 1: pass' ^C locks.  I got a
> backtrace by suspending the locked python, and killng it with a SIGILL to get
> a core.
> (gdb) bt
> #0  0xfa3ab88 in __ioctl () at lio.c:201
> #1  0xfa434f0 in ioctl () at nftw.c:184
> #2  0x5ffd2a74 in rl_restart_output ()
> #3  0x5ffdd0f4 in _rl_clean_up_for_exit ()
> #4  0x5ffdd860 in rl_cleanup_after_signal ()
> #5  0x5ffdd304 in _rl_erase_entire_line ()

Can I see some more of this, please?

Cheers,
M.

-- 
  In general, I'd recommend injecting LSD directly into your temples,
  Syd-Barret-style, before mucking with Motif's resource framework.
  The former has far lower odds of leading directly to terminal
  insanity.                                            -- Dan Martinez


From lale at fotonation.com  Tue Dec  5 15:27:41 2000
From: lale at fotonation.com (Ilariu Raducan)
Date: Tue, 5 Dec 2000 20:27:41 -0000
Subject: CGI POST binary file upload?
References: <90j8j0$cmt$1@nnrp1.deja.com>
Message-ID: <p1dX5.3468$Er5.2143@news.indigo.ie>

In class _sockobject in the file socket.py change the 'r' mode with 'rb'

  def makefile(self, mode='r', bufsize=-1):
        #MODIFIED   return _fileobject(self._sock, 'rb', bufsize)
        return _fileobject(self._sock, 'rb', bufsize)

I tried the POST method with the same results as you and I got the same
results.
After I changed to 'rb' is working

And remember to open any binary file with 'rb' mode in windows

Regards,

Lale




"Bill Seitz" <fluxent at yahoo.com> wrote in message
news:90j8j0$cmt$1 at nnrp1.deja.com...
> I've tweaked some code which someone else posted a while back, to allow
> a CGI (behind Apache on Win2K) accept a POST with a file upload, and
> write it out to the file tree.
>
> It works fine for text/html files (though the byte count is slightly
> off: line endings?). But for binary files it's a total mess. For GIFs I
> generally only get ~100 bytes, for JPEGs ~500bytes. MsWord and PDF
> files also don't work, though I usually get a bigger end result.
>
> I don't think the problem is the file writing, it's the
> receiving/reading. I say this because I also output in my response page
> len(data['image'].value)
> and it agrees with the size of the output file.
>
> I tried adding
> msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
> up at the top of my code, but (a) it doesn't seem to matter, and (b)
> I'm not sure whether I'm doing it right.
>
> thx
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.




From pbjorn at uswest.net  Fri Dec 22 22:56:20 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Fri, 22 Dec 2000 20:56:20 -0700
Subject: A Suggestion for Python Colon Syntax
References: <mailman.977446446.16022.python-list@python.org> <slrn947q64.jkg.zeitlin@seth.lpthe.jussieu.fr>
Message-ID: <3A442264.36499605@uswest.net>

Vadim Zeitlin wrote:

>  BTW, what was the real rationale for adding composed assignment operators
> (+=) to Python 2.0 (which I'm really extremely glad to have now)? Wasn't it
> meant to make Python more attractive to the people coming from C-lie
> languages? And, if by chance it was so, why not continue making it even more
> simple to use to the very same people?

The decision to implement augmented assignment is described in PEP 203
(http://python.sourceforge.net/peps/pep-0203.html). I quote:

   There are two main reasons for adding this feature to Python:
   simplicity of expression, and support for in-place operations.

If we wanted to make C/C++ programmers attracted to Python we'd have added curly
braces <wink>.

-- bjorn



From aleaxit at yahoo.com  Wed Dec 13 08:16:13 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 13 Dec 2000 14:16:13 +0100
Subject: Probleme with locale
References: <917pjb$ii2$1@reader1.imaginet.fr>
Message-ID: <917sre02fh5@news2.newsguy.com>

"Christophe Prevost" <ommadawn at club-internet.fr> wrote in message
news:917pjb$ii2$1 at reader1.imaginet.fr...
> Hi,
>
> I need locale for my re in french. And there's something strange...
>
> import locale
> this work ->
> >>>locale.setlocale(locale.LC_ALL, "")
> >>>locale.getdefaultlocale()
> ('fr_FR', 'cp1252')
>
> this doesn't work ->
> >>>locale.setlocale(locale.LC_ALL, "fr_FR")
>
> Tell me Why?

Because the string 'fr_FR' is not a valid
locale descriptor.

>>> locale.setlocale(locale.LC_ALL, "French_France.1252")

This will work.

I'm not sure what, exactly, .getdefaultlocale() is
returning -- I can't find this function in the docs...


Alex





From aleaxit at yahoo.com  Mon Dec 25 20:00:24 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 26 Dec 2000 02:00:24 +0100
Subject: classes (was Re: Same again please for OOP)
References: <mailman.977623149.8184.python-list@python.org> <9257mf0sl1@news1.newsguy.com> <roy-E5B925.13324224122000@news.panix.com> <925khi019hs@news1.newsguy.com> <kubsu1t81x.fsf@lasipalatsi.fi> <92708d0g4g@news1.newsguy.com> <kuy9x4seef.fsf@lasipalatsi.fi>
Message-ID: <928qqa02n69@news1.newsguy.com>

"Erno Kuusela" <erno-news at erno.iki.fi> wrote in message
news:kuy9x4seef.fsf at lasipalatsi.fi...
    [snip]
> | Giving accessor methods uniform names is hardly a minus, and
> | starting them with something suggesting 'get' is particularly good.
>
> even easier than __get_location(self) is get_location(self) ;)

No!  You have to type __get_location(self) *ONCE*, when
implementing your class; client-code has to type foo.get_location()
a ZILLION times, whenever it needs to access that attribute.

Saving two leading underlines ONCE, and paying with 6 extra
characters PLUS the possible error of omitting the '()':-), a
zillion times -- 'penny-wise, pound-foolish' comes to mind.

Just let clients use foo.location and be happy...


> | example is: obj.price *= 1.07 # 7% increase versus
> | obj.setPrice(obj.getPrice() * 1.07) # 7% increase
>
> | Why force client code to go through the latter, more complicated
> | form, when the simpler first form is available so cheap?
>
> because, the latter is simpler. yes, it is more tedious to type. i
> count this against augmented assignment :)

"I want to multiply the object price by 1.07" is simpler when
expressed just as such.


> | Why force client-code to know and remember which attributes
> | are actually 'in' the original object (and thus must be accessed
> | through accessor methods, in your style) and which ones aren't (and
> | thus must be accessed directly, unless one goes to the serious
> | trouble of building up an accessor method and sticking it on
> | on-the-fly -- blah!)...?
>
> interesting point. i've never done this, seems like dubious practice
> at first sight.

It's very handy & Pythonic.  And Javascript'ers do it all the time,
so as they discover Python you'll see it happen more and more.


> | These are pragmatical considerations that make direct, simpler
> | notation preferable.  On a _conceptual_ plane, the key issue is not
> | confusing state and behavior.  "Methods BEHAVE -- they DO stuff" and
> | "properties JUST ARE -- I can 'read' them, and some of them I may
> | also be allowed to 'write'" is a good pair of rules-of-thumbs to use
> | as guidelines in deciding how to model something -- method, or
> | property (attribute)?  It's a very similar distinction to that
> | between functions and variables.
>
> so, you'd prefer most of the string methods that take no args in
> python 2.0 were actually attributes? i don't know, somehow
> doesn't seem pythonic to me.

Nah -- I don't think of a string as 'carrying around' versions of
itself in all kinds of funky cases, or severally stripped, &c -- I
think of each of this as computational behavior, not *state*, of
the string-object.


Alex





From jwbnews at scandaroon.com  Sun Dec 24 18:55:37 2000
From: jwbnews at scandaroon.com (John W. Baxter)
Date: Sun, 24 Dec 2000 15:55:37 -0800
Subject: <> vs != was Re: A Suggestion for Python Colon Syntax
References: <2kM06.120$Lm2.7518@news.clarkson.edu> <LNBBLJKPBEHFEDALKOLCKEDEIFAA.tim.one@home.com> <mailman.977594956.22001.python-list@python.org> <OR616.67063$x6.31192247@news2.rdc2.tx.home.com> <925ipj$b3d$1@nnrp1.deja.com>
Message-ID: <jwbnews-A319ED.15553724122000@news.olympus.net>

I prefer <> over != for not equal. 

That's probably because I used Pascal and Pascal-like languages for a 
long time before subjecting myself to C.

I'd prefer still to be using Pascal over C, but the C crowd has captured 
the world.

 (For that matter, I prefer "not equal" and "is not equal" over != for 
not equal.)

Others feel differently about the matter...I won't try to convert you 
unless you try to convert me.  ;-)

   --John  (who, sometime late in the 1990s, had to stop saying he had 
been around computers more than twice as long as C had been outside AT&T)

-- 
John W. Baxter   Port Ludlow, WA USA  jwbnews at scandaroon.com


From phd at phd.pp.ru  Tue Dec 12 07:23:01 2000
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 12 Dec 2000 15:23:01 +0300 (MSK)
Subject: dumbdbm bug fixed
In-Reply-To: <a05001901b6584e55adbf@[38.26.127.234]>
Message-ID: <Pine.LNX.4.30.0012121521570.26062-100000@fep132.fep.ru>

Hi!

   Any chance you'll create and post a patch to python.sourceforge.net?

On Sat, 9 Dec 2000, Amos Newcombe wrote:
> In dumbdbm.py (v 2.0) it says
> >- seems to contain a bug when updating...
>
> The bug I found was that the length of an item in the index (.dir)
> file doesn't get updated when the record in the .dat file does.
>
> The fix is simple. Just call self._commit() at the end of
> _Database.__setitem__(), so it looks like this.

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




From peter at engcorp.com  Wed Dec  6 10:01:05 2000
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 06 Dec 2000 10:01:05 -0500
Subject: How can i use a DLL in python ?
References: <t2sjn0aji5dkef@corp.supernews.com>
Message-ID: <3A2E54B1.DDDD1731@engcorp.com>

Jose Antonio martin wrote:
> 
> I am trying to port the fdftk.dll  and fdftk.lib from adobe acrobat to work
> in python with it.

> How can use this Dll and .lib from Python ?
> Can i use this modules  ?

I would guess the .LIB file is used to statically link during
compilation with languages such as C, so you would not use that with
Python, Perl, or Java.

The .DLL file can be used from Python, but to do so you might need to
get the 'calldll' package and "wrap" the .DLL routines with your own
routines from Python.

I don't believe there is any way you can *easily* do what you want, so
be prepared for a little learning and thinking.  With luck, someone else
has done this already and the search of the web that you are about to do
will find it.

(I await rebuttal. :-)


From neelk at alum.mit.edu  Tue Dec 19 19:17:30 2000
From: neelk at alum.mit.edu (Neelakantan Krishnaswami)
Date: 20 Dec 2000 00:17:30 GMT
Subject: Python in "math through programming" curriculum
References: <2g2t3t06r2agrrjmfnr41eu4lc2u39a8qi@4ax.com> <3A3EFF47.67DE69CF@gte.net> <91n9i30pr7@news2.newsguy.com>
Message-ID: <slrn9402bt.fa6.neelk@alum.mit.edu>

On Tue, 19 Dec 2000 10:24:56 +0100, Alex Martelli <aleaxit at yahoo.com> wrote:
>
> People who 'know' the very opposite -- that multiple paradigms widen
> one's mind, and are typically needed for rich, complex programming
> tasks in the 'real world' -- may still be rightly reluctant to give
> multi-paradigm exposure to students at the price of using a rich,
> composite, complicated language.  But Python lets you have your
> paradigms and eat them too...!-)

I disagree with you here about Python's appropriateness. Conceptually,
the strongly typed FPLs are the best languages there are for learning
about the mathematics of computer programming. The combination of
declaring algebraic datatypes and using pattern-matching to describe
functions is a brilliant leap forward, because it makes the
mathematical structure of the program very very explicit. (This is
amazingly apparent if you try to write an interpreter in ML.)

Python is weak in this regard because it makes sum types rather hard
to use: one must declare a set of classes and then give them all the
appropriate methods. This is rather scattered, and when you have
mutually recursive datatypes the relationships become much harder to
see. It's much better than C, or even Java, but still rather awkward
vis-a-vis any of the FPLs.

> Caml, IMHO, has rather cluttered syntax when compared to Haskell (or
> Python!); it only gets worse when the Object-oriented extensions are
> also considered.  Admittedly, a minor consideration, but, _for the
> specific issue of programming-newbies_, not a negligible one IMHO.

This is the *big* weakness of the ML-family languages (including
Haskell and Clean): they all have gnarly syntaxes. Python is proof
that a simple syntax is extremely important; it has essentially the
same semantics as Perl, yet I would gnaw off my arm to escape
programming in Perl and quite enjoy programming in Python. Random
syntactic clutter is really bad for novices; their brains are fully
occupied with the ideas, and trivia like the distinction between "and"
vs. "andalso" and when you should parenthesize expressions will only
distract them and make their lives harder. (Both are real examples
from SML.)

I think that if someone were interested in designing a beginner's
language, a Hindley-Milner typed FPL with a simple syntax (could be
parsed with an LL(1) grammar, say) would be a rather good target to
aim for. Python wins big in this category, of course. 


Neel


From kirschh at lionbioscience.com  Thu Dec 28 03:56:45 2000
From: kirschh at lionbioscience.com (Harald Kirsch)
Date: 28 Dec 2000 09:56:45 +0100
Subject: split hangs
References: <slrn94l1pu.4so.jp@localhost.localdomain> <92e3lm$pni$1@panix3.panix.com> <slrn94l4d3.550.jp@localhost.localdomain> <yv2n1dhufo2.fsf@lionsp093.lion-ag.de> <slrn94lua3.67q.jp@localhost.localdomain>
Message-ID: <yv2ito4vs8i.fsf@lionsp093.lion-ag.de>

jp at ulgo.koti.com.pl (Jacek Pop?awski) writes:

> Harald Kirsch wrote:
> >Yes, something is very strange with this type of regex. Run the
> >following program. It shows strange behaviour with 1.5.2 on Linux as
> >well as with 2.0 on IRIX.
> 
> wow, looks like every step is few times slower than latest, so it
> doesn't hang, it just compute and compute... but why regex are _so_
> slow ?

I checked the same regex with tclsh8.4 and it does not hang. But Tcl,
afaik, has its own perl-compatible regex package. I just downloaded
pcre-3.4 to check how this behaves.

  Harald Kirsch

-- 
----------------+------------------------------------------------------
Harald Kirsch   | kirschh at lionbioscience.com | "How old is the epsilon?"
LION Bioscience | +49 6221 4038 172          |        -- Paul Erd?s



From etsang at my-deja.com  Wed Dec  6 13:42:03 2000
From: etsang at my-deja.com (etsang at my-deja.com)
Date: Wed, 06 Dec 2000 18:42:03 GMT
Subject: import site error!!: in building python 2.0
Message-ID: <90m19p$lhr$1@nnrp1.deja.com>

Hi, when I build the python 2.0 with resource modules turned on, python
compliles and I get the executable. However, I encountered the follwing
error when I test run the executable:

./python -v
'import site' failed; traceback:
ImportError: No module named site
Python 2.0 (#1, Dec  6 2000, 10:33:11)
[GCC 2.8.1] on sunos5
Type "copyright", "credits" or "license" for more information.

>>> >>> import resource
import resource # builtin
>>> ^D
# clear __builtin__._
# clear sys.path
# clear sys.argv
# clear sys.ps1
# clear sys.ps2
# clear sys.exitfunc
# clear sys.exc_type
# clear sys.exc_value
# clear sys.exc_traceback
# clear sys.last_type
# clear sys.last_value
# clear sys.last_traceback
# restore sys.stdin
# restore sys.stdout
# restore sys.stderr
# cleanup __main__
# cleanup[1] signal
# cleanup[1] exceptions
# cleanup[1] resource
# cleanup sys
# cleanup __builtin__
# cleanup ints
# cleanup floats

What is going on???

Also when I try insering the line
expect expect.a in Setup
and try to make again (after make clean and deleting buildno,
getbuildinfo.o and those .a files), it does ot link expect.a?
anyone tried bulding python 2.0 with expect??

Thanks


Sent via Deja.com http://www.deja.com/
Before you buy.


From jvanar1 at my-deja.com  Fri Dec 29 11:48:34 2000
From: jvanar1 at my-deja.com (jvanar1 at my-deja.com)
Date: Fri, 29 Dec 2000 16:48:34 GMT
Subject: Need help with string method
Message-ID: <92if8v$iu1$1@nnrp1.deja.com>

I have a pretty dumb question here.  I can't seem to get any string
methods working from *within* a script.  here is a super-simple example
that yields an error

>>import sys
>>import string
>>theString = 'This is a string'
>>print theString.find('string')

now this works fine from the python command line but gives an error
when run as a script. Here is the error message:

Traceback (innermost last):
  File "./clean", line 18, in ?
    print theString.find('string')
AttributeError: 'string' object has no attribute 'find'

Am I missing an 'import' somewhere.  Thanks for any help...
-Jim



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


From slhath at home.com  Thu Dec 28 14:18:09 2000
From: slhath at home.com (Scott Hathaway)
Date: Thu, 28 Dec 2000 19:18:09 GMT
Subject: how do i check if a file exists with python?
Message-ID: <RlM26.98766$yR4.2639444@news1.rdc1.tx.home.com>

I have searched Deja, looked in two pyton books, and searched the python.org
site trying to find out this simple question.

Can someone please give me the call and the module it lives in?

Thanks,
Scott




From erno-news at erno.iki.fi  Thu Dec 14 22:56:37 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 15 Dec 2000 05:56:37 +0200
Subject: Is Stackless and uthread.py SMP friendly?
References: <91br9v$qa4$1@nnrp1.deja.com>
Message-ID: <ku1yva5ocq.fsf@lasipalatsi.fi>

In article <91br9v$qa4$1 at nnrp1.deja.com>, jbalding at hotmail.com writes:

| will Stackless and uthread.py make my SMP
| experience with Python a happy one?

it should work just fine, since microthreads don't use os threads.

  -- erno


From rturpin at my-deja.com  Fri Dec 29 16:08:26 2000
From: rturpin at my-deja.com (rturpin at my-deja.com)
Date: Fri, 29 Dec 2000 21:08:26 GMT
Subject: Is 100,000 entries big for a dictionary?
Message-ID: <92iug4$vqb$1@nnrp1.deja.com>

I'm in the process of architecting an application with medium sized data
sets. Python dictionaries are tempting as part of the representation
mechanism. How do they perform when they have 100,000 entries? Can you
offer any other experience with this size data set in Python?


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


From loewis at informatik.hu-berlin.de  Wed Dec 27 19:21:36 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 28 Dec 2000 01:21:36 +0100
Subject: Python as a first language (Was: A Suggestion for Python Colon Syntax)
References: <mailman.977446446.16022.python-list@python.org> <9207a3$av6$1@hecate.umd.edu>
Message-ID: <p6qwvcle6pb.fsf_-_@informatik.hu-berlin.de>

William Djaja Tjokroaminata <billtj at y.glue.umd.edu> writes:

> At the present time, correct me if I am wrong, I think it is still
> rare that Python is one's first programming language.  People who
> start to use Python may already use some other popular programming
> languages.

You are probably right, but many of us hope that Python becomes common
as a language to teach to children; much as Logo was at some earlier
time. The CP4E project was working towards that, and I know a few
highschool teachers are interested, as well.

So I would not throw-away to easily advantages of the language when it
comes to newcomers.

Regards,
Martin



From paul.magwene at yale.edu  Wed Dec 13 00:13:00 2000
From: paul.magwene at yale.edu (Paul Magwene)
Date: Wed, 13 Dec 2000 00:13:00 -0500
Subject: Dijkstra's Shortest Path algorithm
References: <Pine.GSO.4.21.0012090042370.15110-100000@y.glue.umd.edu> <dvrd3t40u9l7atgdro59m5ttg7dsfv7c4i@4ax.com> <Pine.GSO.4.21.0012122249480.19682-100000@y.glue.umd.edu>
Message-ID: <3A37055C.D831CACF@yale.edu>

Roy Katz wrote:
> 
> On Wed, 13 Dec 2000, Andy Robinson wrote:
> 
> > Yes, there is.
> >
> > Aaron Watters kjBuckets collections library include a graph type
> > written in C, which has this and several other useful algorithms.
> > www.chordate.com
> >
> > - Andy Robinson
> >
> 
> Wonderful! That's important for my sanity.
> But that sort of takes the fun out of making something people will
> use. Unless something else needs to be made? Is there anything that people
> need implemented? I'm finally ending my semester next week, so it'd be
> great to get back into serious Python programming (as opposed to this
> semester's C++).
> 


If you're interested in graphs and programming, how about a clique
finding algorithm for undirected graphs?  

I implemented a VERY slow clique finding algorithm myself, and then ran
across Andrew Dalke's web page which  has some C code based on the
Bron-Kerbosch algorithm (see
http://starship.python.net/crew/dalke/clique/ ).  I'm still trying to
wrap my mind around the algorithm in this paper (I'm not a computer
scientist), but I'd love to see a `pure python' version (I think this
would help me to understand what's going on), or better yet a fast
version which utilizes Numeric.



Paul Magwene
paul.magwene at yale.edu


From amk at mira.erols.com  Sat Dec 30 17:17:32 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 30 Dec 2000 22:17:32 GMT
Subject: Great New Book
References: <t4o1vr8hktk531@corp.supernews.com> <92ico9$jde$1@panix2.panix.com> <92ig6k02apk@news2.newsguy.com> <hYe36.204059$U46.6288939@news1.sttls1.wa.home.com>
Message-ID: <slrn94soh7.hd.amk@207-172-111-203.s203.tnt1.ann.va.dialup.rcn.com>

On Sat, 30 Dec 2000 06:07:09 GMT, Greg Jorgensen <gregj at pobox.com> wrote:
>I don't know about books published in Italy, but here in America books in
>general, and technical books especially, are sold by weight rather than
>quality of the content. For years now Que, SAMS, Sybex, et al. have

Philip Greenspun's story about writing his book on databased-backed
Web sites, at http://philip.greenspun.com/wtr/dead-trees/story.html,
is relevant here, particularly the section titled "Why Computer Books
Suck".  It's notable that most of the really good books I've read
lately have been quite slim.  ("Into the Wild", "Extreme Programming
Explained", "Telling", "Ghost Children", "Magnificent Corpses"...)
This isn't to say that thick books are automatically bad, especially
when dealing with fiction, but for technical books thickness may
indicate that the author can't boil things down to their essence very
well.

--amk


From sdm7g at virginia.edu  Mon Dec 25 16:31:33 2000
From: sdm7g at virginia.edu (Steven D. Majewski)
Date: Mon, 25 Dec 2000 16:31:33 -0500 (EST)
Subject: A Suggestion for Python Colon Syntax
In-Reply-To: <LNBBLJKPBEHFEDALKOLCKEDGIFAA.tim.one@home.com>
Message-ID: <Pine.A32.3.90.1001225162905.17387A-100000@elvis.med.Virginia.EDU>

On Sat, 23 Dec 2000, Tim Peters wrote:

> So what do *you* want for Christmas, Steven?  We have to do *something* to
> get you back from that XLISP crap.  How about optional colons, or mandatory
> parentheses around 2-letter variable names?
> 
> hoping-you'll-settle-for-a-fruit-cake-ly y'rs  - tim

I'll settle for a couple more pyobjc developers on sourceforge,
although, as I noted in another thread, I wouldn't mind if someone
rewrote the python runtime in objective-c. 


---|  Steven D. Majewski   (804-982-0831)  <sdm7g at Virginia.EDU>  |---
---|  Department of Molecular Physiology and Biological Physics  |---
---|  University of Virginia             Health Sciences Center  |---
---|  P.O. Box 10011            Charlottesville, VA  22906-0011  |---
		"All operating systems want to be unix, 
		 All programming languages want to be lisp." 




From aleaxit at yahoo.com  Fri Dec 22 12:03:27 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 22 Dec 2000 18:03:27 +0100
Subject: __init__ like in java
References: <91vosi$19h$1@reader1.imaginet.fr>
Message-ID: <9201hg0cjr@news2.newsguy.com>

"Gilles Lenfant" <glenfant at equod.com.nospam> wrote in message
news:91vosi$19h$1 at reader1.imaginet.fr...
> Hi,
>
> Is there an *elegant pythonic*  way to have several constructors for one
> class ?

Quoting a post of mine from earlier this month:

"""
class A:
    def __init__(self, *args):
        m=getattr(self,"_init_%d"%len(args),None)
        if m is None:
            raise TypeError, "wrong # of args (%d)"%len(args)
        m(*args)
    def _init_1(self, arg):
        print "One Arg"
    def _init_2(self, arg1, arg2):
        print "Two Args"

I think the if/elif is simpler and clearer, but, hey, whatever
floats your boat; this one can surely be more concise if you
have widely-varying possible numbers of arguments.

Note that you can (and probably should, if you do this often)
get this special dispatching-init from a mixin class:

class InitDispatchMixin:
    def __init__(self, *args):
        m=getattr(self,"_init_%d"%len(args),None)
        if m is None:
            raise TypeError, "wrong # of args (%d)"%len(args)
        m(*args)

so the various classes need just inherit from this and define
the several _init_N methods they desire.
"""

This 'overloads' the constructors based on how many arguments
are given -- how elegant (and how Pythonic...!) this is, being
of course *debatable*.  Overloading on *types* would be less
elegant *and* less Pythonic, though you could easily extend
this idea to do it -- I would discourage it even more strongly.


Alex





From joseamartin at venmex.com  Thu Dec  7 09:48:32 2000
From: joseamartin at venmex.com (Jose Martin)
Date: Thu, 7 Dec 2000 10:48:32 -0400
Subject: DCOracle Need Help
Message-ID: <t2v9fu2pht2d69@corp.supernews.com>

I am trying to use DCOracle on Python 2.0 under Windows or Linux
but when i try to make my scritp :

The first line is
from DCOracle import *

when i try to run:
Python say me:

File "d:\python20\DCOracle\__init__.py", line 52, in ?
    import sys, Buffer
ImportError: No module named Buffer

I was searched my python path for a module called Buffer, but i dont have it
Whats happennig ?

Thanks...







From root at rainerdeyke.com  Tue Dec 12 22:07:17 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Wed, 13 Dec 2000 03:07:17 GMT
Subject: [Q] Are Exceptions used that much in practice?
References: <3A35D546.A4731B5B@enteract.com> <3A36596A.89EEA210@alcyone.com> <3A36C6E3.26181E3F@enteract.com>
Message-ID: <FJBZ5.19773$x6.11272365@news2.rdc2.tx.home.com>

"Jerome Mrozak" <goose at enteract.com> wrote in message
news:3A36C6E3.26181E3F at enteract.com...
> I believe I've been misunderstood.  The problem is not "can I write code
> without handling exceptions" but rather "are exceptions used
> voluntarily, rather than by compulsion".

The Python interpreter will throw exceptions.  You are not obligated to
handle them.  In some cases, you can prevent the exception from being thrown
in the first place.

For example, suppose you are trying to calculate the average of a sequence.

def average(l):
  total = 0
  for item in l:
    total += item
  total /= len(l)

When the length of the sequence is 0, this throws a ZeroDivisionError.  You
can handle this with a try clause:

def average(l):
  total = 0
  for item in l:
    total += item
  try:
    total /= len(l)
  except ZeroDivisionError:
    return 0

You can handle it with a guard condition:

def average(l):
  total = 0
  for item in l:
    total += item
  if len(l) == 0:
    return 0
  else:
    return total / len(l)

Or you can ignore it altogether, assuming that you will never be passed an
empty sequence:

def average(l):
  total = 0
  for item in l:
    total += item
  total /= len(l)

This last option is actually the best in most cases.  If the caller wants to
pass an empty sequence, the caller still has the option of catching the
ZeroDivisionError:

try:
  avg = average([])
except ZeroDivisionError:
  print 'Hello world!'


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From carsten.dot.geckeler at gmx.dot.de  Tue Dec  5 17:42:40 2000
From: carsten.dot.geckeler at gmx.dot.de (Carsten Geckeler)
Date: Tue, 5 Dec 2000 23:42:40 +0100 (MET)
Subject: Random string
In-Reply-To: <200012042351.QAA24052@mail.cybermesa.com>
Message-ID: <Pine.LNX.4.10.10012052340230.507-100000@erde.kosmos.all>

On Mon, 4 Dec 2000, Jay O'Connor wrote:

> 	Is there a good algorithm or existing function for creating a random string such
> as "AFYGADFGWETOQW6DSJD", suitable as a 'session id' for a web session

A very straight forward way is:

import whrandom
str = ""
for i in range(20):
	str = str + chr(whrandom.randint(ord("A"), ord("Z")))

Not very fancy, but it works.  But you still have to check if it's unique.

Cheers, Carsten
-- 
Carsten Geckeler:  carsten dot geckeler at gmx dot de
To get proper email-address replace `dot' and `at' by the corresponding symbols.




From  at home.com  Wed Dec 20 22:57:31 2000
From:  at home.com (John Q. Public)
Date: Thu, 21 Dec 2000 03:57:31 GMT
Subject: Overriding sys.stdout
Message-ID: <3A417FA8.29EBC549@home.com>

I would like to be able to take a snapshot of stdout while a script is
running, but still allow stdout to be written to the terminal.  My
current, working, solution is to hack IDLE but I would like a generic
solution so we would not be bound to a specific IDE.  I would like to
use PythonWin on NT, if I can, or do the journaling in batch mode.

My plan is to implement a class/module/whatever that I can used to
overwrite sys.stdout.  This class/module/whatever would write data to a
file and send it to the old stdout.  My mail questions are these:

    Does such a class/module/whatever exist?
    If not, does anyone have an idea where I can start for info?  As a
last resort, I will start wading through IDLE and PythonWin, but I would
like some ideas first.

Thanks!

Jonathan Polley
jwpolley(a)collins(d)rockwell(d)com




From tage at cs.uit.no  Sun Dec 10 03:20:05 2000
From: tage at cs.uit.no (Tage Stabell-Kulo)
Date: 10 Dec 2000 08:20:05 GMT
Subject: Creating variables in a class definition
Message-ID: <90vebl$683$1@news.uit.no>

Here is how I would like to globally suck in the environment, and then
use that throughout the modules; notice the way I derive from DEMO so
that I can use self.ROOT.


class DEMO:
	# Here are the variables we want to update
	var = { "PATH" : "/bin", "ROOT" : "/var" }
	for i in in a.keys():
		tmp = os.getenv(i)
		if tmp:
			a[i] = tmp
	# Create variables
	PATH = var["PATH"]
	ROOT = var["ROOT"]

class DoIt(DEMO):
	def __init__(self):
		os.chdir(self.ROOT)

The problem with this that the variables in DEMO are named twice; it 
is a ticking bomb waiting for someone to add one in the dictionary and
forgetting to create a variable for it.

The question: How to I create variables (from a dictionaly in this
case) in the class as I initialize?

I known getattr will do it for me, but if I use getattr in
DEMO have I not inhibited myself from using getattr in DoIt (or
ended up with even more complexity than I had in the first place)?

Yours cincerely,

	[TaSK]




From mfletch at tpresence.com  Fri Dec  8 10:40:56 2000
From: mfletch at tpresence.com (Mike Fletcher)
Date: Fri, 8 Dec 2000 10:40:56 -0500 
Subject: My first attempt at python
Message-ID: <CC59868A0A76D311B1F50090277B321C1F372E@VRTMAIL>

Minor note, you might want to make the tokeniser use RE's
"word"/"whitespace" abstract character class to make the tokenising a little
more robust:

>>> import re
>>> re.split( '\W+', 'this. that() is a {} test',)
['this', 'that', 'is', 'a', 'test']

Enjoy,
Mike

-----Original Message-----
From: David Porter [mailto:jcm at bigskytel.com]
Sent: Friday, December 08, 2000 8:30 AM
To: Jason Stewart
Cc: Python List
Subject: Re: My first attempt at python
...
import re
def tokenize(s):
    r = re.compile(' ')
    return r.split(s)
...
Of course, the best way is:

import string
tokens = string.split(s)
...



From tim.one at home.com  Mon Dec 25 14:18:45 2000
From: tim.one at home.com (Tim Peters)
Date: Mon, 25 Dec 2000 14:18:45 -0500
Subject: One Python 2.1 idea
In-Reply-To: <925lp7$d2c$1@nnrp1.deja.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEHFIFAA.tim.one@home.com>

[rturpin at my-deja.com]
> Making Python faster doesn't require the creation of an
> optimizing compiler, complete with data and control flow
> analysis, loop unrolling, expression optimization, code
> unfolding, etc.
>
> There likely would be a lot of bang for the buck from
> small improvement to the existing infrastructure.

Except that everyone who has been interested in pursuing such things has
been doing so for a decade.  CPython is probably about 4x faster today than
it would have been otherwise.  There don't appear to be many *simple*
changes left that are clearly "worth it".

> Can the code for list comprehensions be made faster?

Sure.  "bang for the buck" is limited by how often a particular trigger is
pulled, though, and listcomps are currently barely used.  This isn't a
technical problem, it's a problem of finding someone motivated enough to
bother doing something about it.  You're not -- and neither am I.

> Does the existing front-end pre-execute all constant (sub)
> expressions?

Python does no source optimizations whatsoever today:  WYSIWYG.  If you
write

    x = 1 + 2 + 3

then at runtime it will load 1, load 2, add them, load 3, add that to the
sum of 1+2, then store the result into x.  It doesn't much matter, though:
this isn't C, and there's no preprocessor generating gobs of constant
expressions to *be* folded.  Nobody writes "x = 1 + 2 + 3".  They *may*
write

    X = 1 + 2 + 3

at top level to define a module-level (conceptual) constant, but in that
case again it's not C, and that isn't a macro -- the computation is a
one-time module startup cost, and so speeding it would be an academic
exercise of no practical value.

> Does it pre-construct fields and methods that are NOT
> conditional in class and constructor definitions?

Nope.  Although I'm not sure how you're going to determine whether a thing
is conditional without some flow analysis <wink>.

> Does it bind constant references to these as tightly as possible?

Ditto.

> Are there extensions to the byte code that would increase
> efficiency?

That's been looked at extensively.  The most promising idea along those
lines was to switch from a stack-based VM to a 3-address register-based one
(Skip Montanaro's "Rattlesnake" project).  It's not a huge win, though --
the base problem is that, e.g., in "x + y", Python has no idea at compile
time what "+" may mean, and so regardless of bytecode model it's going to
pump out *something* that says "ok, it's a binary add, now you figure out
the types of x and y, whether coercions are necessary, whether they're class
instances that supply __add__ and/or __radd__ methods, and do something
appropriate with all that".  Breaking that mess into smaller bytecodes would
only slow things down, via forcing that many more trips around the eval
loop.  Moving to a register-based VM would yield a small speedup via
allowing to skip a few trips around the eval loop (the loads and the store
would be specified in the "+" opcode directly, instead of being spread
across 3 other bytecodes).

Before you suggest it <wink>, note that Guido has always been repelled by
the idea of generating more than one bytecode sequence to do a thing and
letting a runtime test decide which sequence would be cheaper.

> (And maybe I'm wrong. Maybe most of the low hanging fruit has
> been plucked.)

Mostly, yes.  There are a number of patches out there that buy 10-15% by
rewriting the eval loop in excruciating-- and usually
platform-dependent --ways, but they exceed Guido's pain threshold.  That is,
they would make the code harder to maintain than they're worth, in his
opinion (mine too, but only his counts ...).

> I guess this is my fault, for saying "speed." And yeah, I
> would love a full, optimizing, JIT compiler with portable
> code generation, so it lives on all platforms. But we all
> know that isn't happening any time soon, unless some
> company bellies up to the bar with money to make it happen.

Note that it's not a universal desire.  I made my living for 15 years
working on optimizing compilers, and one of the things I loved from the
start about Python is that I have never burned any of my time tracking down
a Python optimization bug (it's hard to introduce an optimization bug when
you don't do any <wink>).  OTOH, Python developers have lost countless hours
(and movies, and reunions with friends, and trips, and sleep ...) to
tracking down optimization bugs in the C compilers used to compile Python.
And a bug in a JIT makes that look like a pleasure.  Like everything else,
"optimization" has real downsides.

> I was pushing merely for modest efforts in this direction.
> I can only push, since my time is devoted elsewhere.

Alas, I'm not sure you're going to find anyone on the receiving end of the
nudge.

my-python-code-runs-5x-faster-this-month-thanks-to-dumping-$2K-
    on-a-new-machine-ly y'rs  - tim




From donn at u.washington.edu  Fri Dec 29 12:27:28 2000
From: donn at u.washington.edu (Donn Cave)
Date: 29 Dec 2000 17:27:28 GMT
Subject: How to import from within a module
References: <mailman.978059610.14528.python-list@python.org>
Message-ID: <92ihi0$jj8$1@nntp6.u.washington.edu>

Quoth Le Snelson <le_snelson at transtechinc.cc>:
...
| I am unable to get __import__( xxx ) to actually add xxx to any visible
| dictionary. However, in the same file if I uncomment a line at the top
| (global scope) that reads:
| import xxx
| Sure enough the xxx is in the list from dir(init).
|
| I have tried all combinations of __import__(xxx, globals(), locals())
| __import__(xxx, globals(), globals()) etc. I have also tried the "import
| xxx" within the function I'm working on.  I expected syntax error but
| got none as well as no discernible result.
|
| Do I need to use the imp module to get this module found and loaded?

I hope you will laugh at how simple it is:

      xxx = __import__('xxx')

Now, I don't use this commonly and may have missed something, but
that seems to work for me.  If you type __import__('string')
interactively, the interpreter prints out its function return, so
it would have been easy to see there.

	Donn Cave, donn at u.washington.edu


From gogo at bluedynamics.com  Wed Dec 20 19:12:13 2000
From: gogo at bluedynamics.com (Georg Gogo. BERNHARD)
Date: Thu, 21 Dec 2000 00:12:13 GMT
Subject: How to make a GUID with plain python
References: <3a413a17.3416402@192.168.0.1>
Message-ID: <3a414ad8.7705216@192.168.0.1>

Well, the thing is that we want to use the mechanism with a ms-sql
server and its "uniqueidentifier" datatype. This GUID is some sort of
standard and is used within DCOM or RPC algorithms and seems to be
very relyable. 

There is a big disadvantage if I use IP adresses since they can be
"unrouteable" as you called, which could mean that they are within a
LAN but using a firewall. The MAC adresses are unique, as I
understood. The Windows CoGreateGUID COM API function combines a
timestamp and a MAC adress. I want to do the same thing in python,
which is maybe not possible.

Thank you for your help,
Gogo.






From moshez at stimpy.scso.com  Tue Dec 26 08:56:43 2000
From: moshez at stimpy.scso.com (Moshe Zadka)
Date: 26 Dec 2000 13:56:43 -0000
Subject: One Python 2.1 idea
In-Reply-To: <20001226140032.C25820@xs4all.nl>
References: <20001226140032.C25820@xs4all.nl>, <mailman.977771949.17911.python-list@python.org> <9292tn$jka$1@nnrp1.deja.com> <929n9c013m3@news1.newsguy.com>
Message-ID: <20001226135643.13218.qmail@stimpy.scso.com>

On Tue, 26 Dec 2000 14:00:32 +0100, Thomas Wouters <thomas at xs4all.net> wrote:

> doing 'x.foo = x.foo' 
...
> broke the pickle modules (because bound methods
> can't be pickled.)

Without getting into the cache vs. no cache debate, that simply sucks.
I've run into this problem a couple time, and had very ugly workarounds
for that.

I think it can be fixed, and I will try to patch that if I ever get the
chance. (Pickle a method as a pair: im_self and the method's name.
Then upon unpickling, get im_self.name)



From billtj at my-deja.com  Wed Dec 27 09:26:45 2000
From: billtj at my-deja.com (billtj at my-deja.com)
Date: Wed, 27 Dec 2000 14:26:45 GMT
Subject: A Suggestion for Python Colon Syntax
References: <mailman.977546107.21808.python-list@python.org>
Message-ID: <92cu71$9m4$1@nnrp1.deja.com>

> So does Guido, but it's rare, and in this case you're ignoring that he
> *already decided*.

Tim, people opinion can change over time, and hopefully Python does not have
the principle that once a feature is already decided, it can never be
changed.  I think to survive, a language has to be flexible to follow
(rapidly) changing technology and therefore modify some of its own
characteristics.

> Yes; I agreed to that in my first msg.  It doesn't matter, though, in the
> sense that Python is not about minimizing keystrokes (not even so much as
> Perl is, let alone extreme one-liner languages like APL or J).  It's much
> more about maximizing readability.  It may be that you're in a minority who
> does not find that the trailing colon increases readability?  It increases
> readability for me, and that matches the testimony of most people who've
> chimed in on this issue over the years.

I think there is some inconsitency here.  If Python is really not about
minimizing keystrokes but more about maximining readability, then I have
at least one counter-point:  Why don't add a pair of parentheses after the
clauses such as:

    if (xxxx xxxx xxxx):
        xxxx xxxx xxxx
        xxxx xxxx xxxx

I think this is even more readable.  Also Python is not consistent in
parentheses:  No parentheses are required after if, for, etc., but parentheses
are required for function argument in a def statement:

    def func (arg1, arg2):

Just like in Tcl, why not make everything consistent?  Like

    def func arg1 arg2:
        xxxx xxxx xxxx

    if xxxx xxxx xxxx:
        xxxx xxxx xxxx

It will then be easier to memorize the basic syntax of Python.  Because every
time I have to switch between C/C++, Perl, etc., I have to look up the
books just for syntax.

> ABC was never popular.  And I'm not sure what more people being familiar
> with Tcl has to do with this.  That *may* influence a decision to add
> semantic features (like, say, file events) someday, but is unlikely to have
> any influence on future syntax -- the languages are worlds apart
> syntactically.

The relationship between Tcl and Python, to me at least is, after the first
keyword such as if, for, etc., then we just add words after that without
needing any parentheses, etc., except the colon in Python.  Yes, in Tcl we
need braces after if, but that is because of basic syntax of Tcl, and not
because it is for "if" or for "for" needs.  I myself programmed in Tcl before
I used Python.	For a more general example, the book Tkinter assumes that
some people have programmed in Tcl/Tk before they start using Tkinter.	I
always see things like "Python for Tcl programmer", and not vice versa. 
Therefore, in the world of Tk, I bet there are a lot people migrating from
Tcl to Python and probably almost none from Python to Tcl.

>
> > I can see clearly the big difference (as I have experienced) between
> > using indentation and curly braces; it makes really consistent coding
> > style.  However, I don't see the colons as falling into the same
> > category.
>
> Guido does.  Now what?
>
Guido does, but shouldn't we take also the opinions of the community?  I
don't like to make this analogy, but in the history, there were/are
communities ruled by kings, and there communities ruled by
democracy.  Which one works better for the welfare of the community?

> > Yes, Guido wants everybody to use colons at the end of the clauses,
> > but right now what prevents someone to put semicolon at the end
> > of every statement?
>
> Some people do.  When they do it public, we ridicule them.  Then they
> usually stop <wink>.

Tim, if you don't allow people to, why don't you just make it part of the
language syntax?  Then everybody will be happy.  People who use semicolon can
be stopped by the interpreter and immediately fix the problem, and you don't
have to "police" the people and you will have more time for other fun things
:).

The problems with C++ is a lot of software construction is based on
conventions, and not part of the language itself.  But because it is based on
conventions, people can abuse them.
>
> >  Just compare the two or three formats:
> >
> >     xxxx xxxx xxxx:
> >         xxxx xxxx xxxx
> >         xxxx xxxx xxxx
> >
> >     xxxx xxxx xxxx:
> >         xxxx xxxx xxxx;
> >         xxxx xxxx xxxx;
> >
> >     xxxx xxxx xxxx
> >         xxxx xxxx xxxx
> >         xxxx xxxx xxxx
> >
> > Which one do you think is the most consistent in layout?  To me, it
> > is not the first one.
>
> The most consistent would be this:
>
>     xxxx xxxx xxxx
>     xxxx xxxx xxxx
>     xxxx xxxx xxxx
>
> That is, consistency is a red herring.

No, because then you lose the indentation which without question has been
advocated since the time of structured programming.
> > ...
> > But the semicolon at the end of the statement break the Pythonic
> > rule.
>
> If you put them there, yes.  Happily, nobody does; for example, you won't
> find a semicolon at the end of any code stmt in any .py file in the standard
> distribution.  When conformity is voluntarily and universally achieved,
> there's no need to legislate it.
>

Python could also have achieved the indentation through voluntarity, but it
forced it through the syntax.  Again, I think it is better if everything is
consistent.  If something is to be universal, force it through the syntax. 
If syntax allows something to happen, then people, in my opinion, must be
allowed to use it.  By having part rule and part voluntarism, it creates
extra work, especially in the voluntarism part.  And if it has been
universally accepted, then I guess nobody will complain when you force it
through the syntax, such as from now on no semicolon is allowed at the end of
the line, unless it is followed by another statement on the same line.

> > ...
> > I am not too familiar with the parsing stuff.  However, in my
> > simplistic opinion, Python is not a free-form language like C
> > or Perl.  Therefore, probably it is reasonable for any parser
> > to breaks a Python code first into lines, even in backward
> > parsing, instead of parsing it token by token first.  In parsing
> > backwards, can then it just detect first that the line is at
> > different indentation level rather than try to detect the colon?

> When you're writing code and hit the ENTER key, good editing environments
> try to *suggest* sensible indentation for the new line.  In part, that
> requires guessing whether the statement just ended opens a block (in which
> case the new line should be indented more) or not (in *most* of which cases
> the indentation should be duplicated from the statement that just ended).
>
> Regular expressions don't suffice for this determination, so it's a lot of
> painful character-at-a-time parsing.  The trailing colon is a great hint:
> if the stmt that just ended doesn't end with a colon, there's no need to
> endure the expense of further analysis (the line that just ended can't
> possibly open a new block without that colon).  Speed is important here,
> because the user expects the response to the ENTER key to appear
> instantaneous, and the parsing code is usually written in an interpreted
> language.  The Emacs Python mode is greatly helped by some parsing
> primitives supplied by elisp and coded in C.  IDLE (and by inheritance, also
> PythonWin, which shares IDLE's auto-indent code) has a much rougher time of
> it, being coded in pure Python, and having the added speed burden of needing
> to talk to a Tk text widget thru the Tkinter interface layer to find out
> anything about what's in the buffer (even worse, that's indirected in Python
> code too, so that PythonWin can talk to the Scintilla text widget instead).
>
> Every quirk of the syntax is exploited mercilessly to reduce processing
> time, and that's a hard job (I wrote both of those parsers, so I'm not just
> guessing about that); the use of colons to open blocks is one of the quirks
> that can be exploited a lot.  The IDLE code would need to be redone from
> scratch without it (IDLE doesn't preserve any of the alphanumeric characters
> now:  it squashes all runs of alphanumerics into a single "x" character,
> because it can do that quickly, and then chew over far fewer total
> characters at Python speed).

I'm sorry, Tim, my background is not too strong in this parsing department. 
My question is, can you just detect the first word of the line?  I think
there are only a finite number of them, such as class, def, if, for, etc.  If
the first word is any of these, then indent the next line, without regard of
the colon.  Is detecting any of these words at the beginning of line much
harder (apart from more CPU time) than detecting a colon at the ending of
line?  (I don't mean programmatically.	Yes, of course there will be more
code to detect the various words than detecting the single colon.  But at
least, in principle, the two are equivalent, right?)

>
> At this point, if you want to keep pushing this the way to do it is to open
> a PEP:
>
>     http://python.sourceforge.net/peps/
>
> While a PEP needs an implementation before it can become final, the PEP
> author need not write the implementation, and the PEP can be accepted before
> an implementation is even started.  What you would need to do is make a
> compelling (to Guido) case, and identify all the consequences and how
> they'll be dealt with.  The process is covered in more detail in
>
>     http://python.sourceforge.net/peps/pep-0001.html
>
> I expect the PEP will be rejected, but that doesn't mean it will be.  Guido
> did surprise me once, in about 1995 <wink>.

OK, I will do it.  Thanks.  Hopefully Guido will be more open and generous
during the holiday season :) .

Regards,

Bill


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


From urner at alumni.princeton.edu  Wed Dec 20 13:36:27 2000
From: urner at alumni.princeton.edu (Kirby Urner)
Date: Wed, 20 Dec 2000 10:36:27 -0800
Subject: Python in "math through programming" curriculum
References: <2g2t3t06r2agrrjmfnr41eu4lc2u39a8qi@4ax.com> <3A3EFF47.67DE69CF@gte.net> <513v3toslkjkggh9oumlf20qc9bi1gdtif@4ax.com> <3A3FC677.FDB19937@gte.net> <CeP%5.521$AH6.45315@newsc.telia.net> <3A401C2E.1FF789A8@gte.net>
Message-ID: <n1t14ts1somvhn7cmqfvu5q8ie1dvkrque@4ax.com>

Cliff --

It's fine to have a difference of opinion regarding what
computer languages to use for what purposes e.g. w/ 
regard to 8th grade math learning.  I have no experience
with the languages you mention (except some with 
Mathematica) and am quite willing to accept that 
brilliant and appropriate curriculum writing for 
K-12ers might incorporate them -- I just haven't seen
a lot of that yet and was wondering if you'd done any
yourself.

The only point of my earlier question was to inquire as
to how much personal experience you have with the 
languages you mention (I know you've used Mathematica
and Ada), and to encourage you to share from personal
experience, as that's what tends to be more valuable.
Anyone can read web pages hyping this or that language,
but only through actual use and experimentation to we
put ourselves in a position to form personal judgements.

I am quite happy to attach weight to your opinions IF 
you have some hands-on experience to share.  Just saying
that these languages are worth considering because 
they're free (vs. Mathematica) isn't sufficient data 
on which to base judgements IMO.  Relative cost is 
just one of a great many criteria we need to consider,
and not necessarily the most important one.  QBASIC is
free on most DOS machines yet I wouldn't recommend it
as a first language in this day and age.

I agreed with you that the Mac implementation of IDLE
might be rather disappointing, have asked for more 
feedback on whether it really lacks color coding of 
keywords.  The high school I'm aware of which so far
does the most with Python (Yorktown High in Arlington
VA) is Linux-based.  I'm wondering if Macworld, which
is what many schools have invested in, is going to 
end up being somewhat incompatible with Pythonworld,
at least at the K-12 level. I don't actually know enough
to have an answer to that question, which is why I'm 
asking it.

I don't think you should fly off the handle just 
because I prodded you to share from personal experience
about the claims you're making for this or that 
computer language over Python.  I think to discuss
this issue intelligently, you should have more to 
say than a few negative words about using it on a
Mac, without any specificity.  Now you're echoing my
own critique re the IDLE interface, but I'm the one 
who brought that up, from my own personal experience
installing it at my daughter's school after downloading
it over the internet.  Did you have the same complaint?
Did you know that IDLE had color-coded key words on 
these other platforms?  So have you written programs
in Python on these other platforms then?  What did 
you think of it, compared to what you've done with
Clean?  All I was asking was for more details.

Those who study my website will see I've put a lot 
of work into curriculum writing that features Python.
Plus I'm a long time student of the philosophy you 
mention (synergetics) and I have my own take on it,
which I share via my websites.  You're quite free to 
offer your criticisms and negative comments, obviously
(you do it a lot), but the question naturally arises 
as to what you would offer the world in place of what 
I've contributed.  In the so-called free market, to 
compete is to have an alternative, better product, and 
to advertise its compelling advantages.  

In your case, I'm not aware of any alternative product, 
just the negative comments, plus some Mathematica 
worksheets which are not even attempting to interpret
that whole philosophy.  So again, rather than indulge
in negative comments, how about you advertise your 
own positive contributions, i.e. your own curriculum 
writing making use of these other computer languages
you've been posting about (the DrScheme team has 
certainly set a high standard, with their high quality
materials).  

Tell us specifically what's wrong with Python from 
your own experience -- because even if the IDLE GUI 
isn't as fully developed for the Mac as it is for 
other platforms (something I'm asking about -- I don't 
know for absolute certainty), this is a rather 
superficial shortcoming that might be addressed 
(may have to do with limitations of Tk on the Mac?).  
You should be more specific, and not so quick to get 
emotional as a substitute for dealing with substance.

I apologize if saying I wasn't sure if I should 
attach any weight to your opinions was so incredibly
insulting that you had to flame, but on the other 
hand I don't think you should just assume that you 
can wade into whatever debates you see on whatever 
topic and just assume that you'll be taken seriously
no matter what.  You have to do the homework.  If 
you've done it, no problem, share your results.  If
you haven't, how about a little honesty on that score?
-- that would goes a long way towards giving people 
a better sense of what you DO have well-informed 
opinions about.  I'm quite willing to accept that 
you're more than a loudmouthed know-nothing.  I'm 
giving you the benefit of the doubt.

Kirby



From tim.one at home.com  Thu Dec 28 14:24:16 2000
From: tim.one at home.com (Tim Peters)
Date: Thu, 28 Dec 2000 14:24:16 -0500
Subject: Good introduction to functional programming with Python?
In-Reply-To: <3A4B4A3F.99A4E018@schneider-kamp.de>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEMCIFAA.tim.one@home.com>

[Tim]
> if-we-don't-act-fast-i'm-afraid-python-may-lose-the-industrial-
>     control-market-ly y'rs  - tim

[Peter Schneider-Kamp]
> You'd have to write program verification tools for distributed
> python applications...

No problem -- tabnanny.py can run on any number of machines simultaneously!

> maybe-that's-why-erlang-is-erlang-ly y'rs

feh!-erlang-doesn't-even-have-a-tabnanny-ly y'rs  - tim




From root at rainerdeyke.com  Tue Dec 12 14:34:08 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Tue, 12 Dec 2000 19:34:08 GMT
Subject: [Q] Are Exceptions used that much in practice?
References: <3A35D546.A4731B5B@enteract.com> <914mfd02ibt@news2.newsguy.com>
Message-ID: <Q4vZ5.19457$x6.10649942@news2.rdc2.tx.home.com>

"Alex Martelli" <aleaxit at yahoo.com> wrote in message
news:914mfd02ibt at news2.newsguy.com...
> (Some might prefer a wider 'except: return 0' clause, but
> most often it's better to be specific in catching exceptions).

I would go further and say that 'except: return 0' should never be used,
since it would catch and ignore MemoryError, KeyboardInterrupt, SystemExit,
ComputerExplodingError, or any number of other exceptions that are totally
unrelated to the issue at hand and require careful handling and/or program
termination.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From sragsdale at my-deja.com  Tue Dec 12 18:22:05 2000
From: sragsdale at my-deja.com (sragsdale at my-deja.com)
Date: Tue, 12 Dec 2000 23:22:05 GMT
Subject: Bug: Readline getting stuck on Linux and Solaris
References: <lxk8zqwupy2.fsf@ioasun3.epfl.ch> <90m53k$ovh$1@nnrp1.deja.com> <m3u28h48ja.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <916bum$a9k$1@nnrp1.deja.com>

In article <m3u28h48ja.fsf at atrus.jesus.cam.ac.uk>,
  Michael Hudson <mwh21 at cam.ac.uk> wrote:
> sragsdale at my-deja.com writes:
>
> > In article <lxk8zqwupy2.fsf at ioasun3.epfl.ch>,
> >   Jan Kybic <Jan.Kybic at epfl.ch> wrote:
> > > Hi,
> > > 	I am using Python 2.0, I downloaded it in the RPM form and I
> > > also compiled it from source on both Linux RedHat 6.2 and Solaris.
In
> > > both cases, when I have the readline module enabled, the keyboard
> > > input gets stuck.
> >
> > I'm having similar problems with Python 2.0 hanging in Irix.  What's
> > most annoying is that this happens on the Python command line when
you
> > hit control-C.  In Python 1.5.2 this is a KeyboardInterrupt event.
In
> > Python 2.0 this hangs the interpreter which starts using 100% of the
> > processor.  I can't kill it with more control-C's.  If I disable the
> > readline (4.1) module this problem goes away.  This problem does not
> > happen in Linux 2.2.15.  I can't find this bug listed on the
sourceforge
> > bug list.
>
> Hmm.  This *used* to happen on Linux (with slightly different
> symptoms; control-C would kill the interpreter, not hang it), but the
> bug got fixed.  A couple of things to try: does control-C do what it
> should when you're not sitting at the ">>> " prompt?  E.g. if you type
>
> >>> while 1:
> ...    pass
> ...
>
> and whack control-C, what happens?

control-C behaves normally in that case.  Python2 is able to exit from a
while loop via control-C without problems.  However if you hit control-C
*again* (after you've exited to the '>>>' prompt) you're back in the
locked-up mode.

Haven't tried building without threads, but needless to say this
shouldn't happen even with threads.


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


From fredrik at effbot.org  Thu Dec  7 00:58:19 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Thu, 07 Dec 2000 05:58:19 GMT
Subject: UnboundLocalError: Local variable 'Tee' referenced before assignment
References: <90mru0$cv4$1@nnrp1.deja.com>
Message-ID: <%FFX5.2390$O%.371404@newsc.telia.net>

jschmitt at vmlabs.com wrote:
> I'm getting an error message that I don't understand.  I'm using this
> handy class that someone kindly on comp.lang.python:
>
> class Tee:
>     """
>     Think of the unix command-line utility
>     """
>     def __init__( self, *fileobjects ):
>         self.fileobjects = fileobjects
>
>     def write( self, string ):
>         for fileobject in self.fileobjects:
>             fileobject.write( string )
>
> and I'm using it kind of like this:
> reportfile = open( os.path.join( workingdir, "report.txt" ), "w" )
> sys.stdout = Tee( reportfile, sys.stdout )    # THIS LINE IS THE PROBLEM
> # do some print's here...
> del Tee
> reportfile.close()

assuming the last code is inside a function:

> del Tee # THIS LINE IS THE PROBLEM

why do you think you have to remove the Tee class
from the local scope?

when you do this, Python thinks Tee is a local variable, which
means that you cannot use it *before* it has been assigned
inside the function.

just remove the "del" statement, and it'll work as expected.

> UnboundLocalError: Local variable 'Tee' referenced before assignment
>
> Does anyone have a spare clue?

did you look in the manual?

</F>




From gmayer at cs.bgu.ac.il  Tue Dec  5 23:42:31 2000
From: gmayer at cs.bgu.ac.il (Mayer Goldberg)
Date: Wed, 6 Dec 2000 04:42:31 GMT
Subject: newbie to python: can't load tkinter -- please help
Message-ID: <90kg3p$n2p$0@pita.alt.net>

Hello:

I'm very new to python. I'm running python 2 on windows 95 US. I'm getting
the following error message when trying to use tkinter:

-------------------------------------------------
  File "d:\python20\lib\lib-tk\Tkinter.py", line 34, in ?
    import FixTk # Attempt to configure Tcl/Tk without requiring PATH
  File "d:\python20\lib\lib-tk\FixTk.py", line 1, in ?
    import sys, os, _tkinter
ImportError: DLL load failed: One of the library files needed to run this
application cannot be found.
-------------------------------------------------

I realise that I need to set up PATH correctly, but I'm not sure to what...
Also: This may not be the only environment variable that needs to be set...

Can someone please help?

Thanks,

Mayer


From scarblac at pino.selwerd.nl  Tue Dec 12 12:45:45 2000
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 12 Dec 2000 17:45:45 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu>
Message-ID: <slrn93cotl.80c.scarblac@pino.selwerd.nl>

Roy Katz <katz at Glue.umd.edu> wrote in comp.lang.python:
> On Tue, 12 Dec 2000, Fredrik Lundh wrote:
> 
> > The real problem is that many Python tutorials fail to explain
> > how things work (or hide it somewhere in the exercises), but
> > that's not really Guido's fault...
> 
> Tough for the new Python programmer, then.  It's nt fair to them. 
> 
> > (Hint: Everything is a reference to an object.  Variables are
> > named references, not actual objects.  Some objects can be
> > modified in place.  Some cannot be modified.  That's all)
> 
> Okay! tell me, which are the objects which cannot be modified in place?

Integers, longs, complex numbers, strings, functions, tuples, types, floats
and None are immutable. Most of these are obvious; the number 3 is the
number 3, why would you change that?!

Lists, classes, instances, dictionaries and modules are mutable.

Did I miss any?

So there are only three mutable things: lists, dictionaries, and namespaces.

Anyway, the important ones of these are in any Python book. You sure you've
been programming Python for three years?

> assume these are int's, long's, float's, etc.  Strings, lists, objects and
> dictionaries are call-by-reference.  Pardon my choice of diction, but I am
> against inventing new terminiologies (modified-in-place??).

Mutable and immutable. And *everything* is by reference, including
assignment. No new terminology here.

> I want to know where this is documented.  Specifically, where is the
> dichotomy of ints/longs/floats vs. objects/strings/lists/dictionaries.

Strings are not mutable.

There is no real dichotomy here. Some types have methods that can change
the type, other types don't. Some types are namespaces (modules, classes
and instances), and they can be changed.

> While I'm brainstorming with Python grammar changes,

Not a very useful idea 10 years after the language was released, but
whatever... 

-- 
Remco Gerlich


From mak at mikroplan.com.pl  Wed Dec 20 05:56:48 2000
From: mak at mikroplan.com.pl (Grzegorz Makarewicz)
Date: Wed, 20 Dec 2000 11:56:48 +0100
Subject: BUG in SWIG refcouting
Message-ID: <NDBBIKNLJKPLOLAJJJAPOECJFFAA.mak@mikroplan.com.pl>

Every SWIG version, which generates python code, has a bug.
This bug is in function which takes Python's self variable.
For SWIG-883 it is a SWIG_GetPtrObj function.

Below is the patched version:

SWIGSTATICRUNTIME(char *)
SWIG_GetPtrObj(PyObject *obj, void **ptr, char *type) {
  PyObject *sobj = obj;
  char     *str;
  if (!PyString_Check(obj)) {
    sobj = PyObject_GetAttrString(obj,"this");
    if (!sobj) return "";
//test by commenting NEXT line!
    Py_DECREF(sobj);
  }
  str = PyString_AsString(sobj);
  return SWIG_GetPtr(str,ptr,type);
}


How to check this?

Create the files:

1. swigbug.i which contains:
//
%module swigbug

%{
#include "swigbug.h"
%}

%include "swigbug.h"
// EOF

2. swigbug.h which contains:
class SwigBug {
public:
	SwigBug(){}
	~SwigBug(){}
};
//EOF

3. d_xrange.py which contains:

import sys
import swigbug
def demo():
	bug=swigbug.SwigBug()
def main():
	print 'xrange:'
	rc0=sys.gettotalrefcount()
	rc1=sys.gettotalrefcount()
	for i in xrange(100):
		demo()
		rc1=sys.gettotalrefcount()
		if rc1-rc0:
			print 'loop:',i,rc0,'+',rc1-rc0
			rc0=rc1
main()
# EOF

4. Compile the debug version of module swigbug and execute in the debug
version of Python
	command line for swig: swig -python -dnone -c++ -shadow -o swigbugc.cpp
swigbug.i

5. Have you patched the file generated by SWIG?
	NO - at the end of every loop you should see the message, that reference
counter is increased
	YES - there are messages about increasing the reference counter for numbers
in range 86-99 (WHY???)

6. For curious: you could check what will you get if you change the loop
from:
	for i in xrange(100)
to:
	for i in [0 , 1 , ..., 100]:
or:
	for i in range(100):
or:
	while i<101:

Grzegorz Makarewicz




From abrahams at mediaone.net  Sat Dec  9 12:04:39 2000
From: abrahams at mediaone.net (David Abrahams)
Date: Sat, 09 Dec 2000 17:04:39 GMT
Subject: Python wrapper for C++ STL?
References: <3A31760E.D157EB54@erols.com> <C0gY5.126062$U46.4232859@news1.sttls1.wa.home.com> <90s3fk$j8p$1@nnrp1.deja.com> <3A31A05C.6A853969@erols.com> <90taaj02qjj@news2.newsguy.com>
Message-ID: <HCtY5.88851$751.2103138@typhoon.ne.mediaone.net>

"Alex Martelli" <aleaxit at yahoo.com> wrote in message
news:90taaj02qjj at news2.newsguy.com...

> Hmmm, yes, but, take care: the 'loop-through' feature had better be
> exposed through a specially-named method or attribute -- I think it
> would be better if .keys() returned a normal Python list, as a caller
> would expect from normal polymorphic behavior, and while it's nice
> that this list is sorted rather than randomish'ly unordered, it can be
> a substantial overhead to build it.
>
> And you can't just use:
>     for key in myFuncyObject:
>         whatever(key)
> as that would require the object to be indexable by 0, 1, ..., n with
> the meaning of 'loop through it', interfering with the normal meaning
> (as these integers can be perfectly normal dictionary-keys!-).
>
> The design pattern to implement the indexability thus requires an
> auxiliary object (and a little bit of care in it, since Python for-loop
> protocol is based on a growing int index, and you have to mutate
> that, internally, into a sequential 'current iterator'/'increment it'
> pattern).  No rocket science, mind you, but it's not quite as obvious
> as _just_ 'wrapping a std::map'...:-).

The Boost Python Library to be released this weekend under that moniker
(formerly py_cpp) has a feature in the pipeline allows easy wrapping of STL
containers. Even non-random-access containers can be used efficiently with

    for element in container.cursor():
        whatever(element)

This ingenious feature was designed by Ullrich Koethe. It won't be in the
first release, but probably will appear within the next few weeks
thereafter.

Regards,
Dave




From SBrunning at trisystems.co.uk  Tue Dec 12 04:35:49 2000
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Tue, 12 Dec 2000 09:35:49 -0000
Subject: compiled or interpreted?
Message-ID: <31575A892FF6D1118F5800600846864D5B15B6@intrepid>

> From:	Matt Frey [SMTP:freydaddy at kricket.net]
> Is Python a compiled language that can be run on a platform like c++, or
> is
> it interpreted like java, or is it in a class of its own?
 
It's interpreted, but not like really like Java. And it's *certainly* in a
class of its own!

See <http:\\www.python.org> for information on Python.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk





-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From tcmay at got.net  Sat Dec  9 02:02:54 2000
From: tcmay at got.net (Tim May)
Date: Fri, 08 Dec 2000 23:02:54 -0800
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu> <slrn92tu7t.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com>
Message-ID: <tcmay-8BCFAB.23025408122000@news.got.net>

In article 
<slrn92tu7t.lk.amk at 207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com>, 
akuchlin at mems-exchange.org wrote:

> On Wed, 06 Dec 2000 17:06:55 -0800, Nathan Gray <nospam at caltech.edu> 
> wrote:
> >It's really upsetting how often I've seen this sort of response to a 
> >polite
> >question on this newsgroup lately.  Tell me, what's so terrible with 
> >asking a
> >group of experts for pointers when researching a subject?  It's a 
> >perfectly
> >legitimate and highly effective way to learn about something.  There's 
> >plenty
> 
> I've noticed this trend, too, though T.C. May, the person you're
> replying to, has always had a low tolerance for cluelessness, as I
> noted when lurking on the cypherpunks mailing list in the early 90s;
> May's posts were often informative and entertaining, but not often
> placid.  (I was pleased to turn up here, though I'm not sure why he's
> taken up Python.)

I used to use Lisp when I ran an AI project for Intel. Later, I took up 
Smalltalk. A bunch of friends of mine are doing crypto and digital cash 
sorts of apps in a mixture of C and Java but with Python as the 
scripting/glue language. The first anonymous remailers were done in 
Perl. 

I'm very interested in Python vs. Ruby vs. Smallscript (a forthcoming 
scripting language) for some projects I'm interested in.

Does this explain my interest?

I did some work on simulated annealing while at Intel, which is why I 
commented in the SA thread. (A friend of mine worked for Scott 
Kirkpatrick, the "father of simulated annealing," at IBM Yorktown 
Heights.)

As to my tolerance for cluelessness, it's not so much impatience as 
simple advice: nearly _every_ garden variety inquiry about "what's the 
history of Python?" and that sort of thing should first be submitted to 
a search engine. Thus was it so five years ago, and thus is it so in 
spades today.

--Tim May

-- 
--Tim May, Corralitos, California

Language interests: Smalltalk (Squeak), Lisp, Mathematica
Machine: Macintosh G4/G3 Powerbook
Project interests: evolutionary programming, games (Go), agents, cryptology


From donn at u.washington.edu  Thu Dec 28 12:34:07 2000
From: donn at u.washington.edu (Donn Cave)
Date: 28 Dec 2000 17:34:07 GMT
Subject: Some Python 2.1 ideas
References: <3A452881.7686E717@uswest.net> <%vb16.187233$U46.5801214@news1.sttls1.wa.home.com> <3A464D5D.B541C04E@uswest.net> <mailman.977688488.25438.python-list@python.org> <20001225150250.6FA06A84F@darjeeling.zadka.site.co.il> <mailman.977760789.10224.python-list@python.org> <roy-EA8C3B.11361325122000@news.panix.com> <3A479C29.2631F27B@uswest.net> <mailman.977876470.23568.python-list@python.org> <3A4A50E6.11BBE0FD@uswest.net> <mailman.977962871.16241.python-list@python.org> <3A4AB05C.D41F4022@uswest.net>
Message-ID: <92ftif$17fc$1@nntp6.u.washington.edu>

Quoth Bjorn Pettersen <pbjorn at uswest.net>:
| D-Man wrote:
|
| > Yeah, it is a problem if you don't know where the file came from.  How
| > about creating a standard text file format?  All files could be opened
| > in binary mode on all platforms, then EOL is handled by each
| > application (or better yet make a standard module that does the
| > implementation correctly once).
|
| Feeling like fighting windmills today, eh <wink>? I think the current wish is to
| do it from the other end, i.e. instead of relying on the text file format to be
| "correct", implement a method on string object that is smart enough to handle all
| text file format line endings intelligently.

Without the context of the I/O stream, to decide whether '\r' is
data or separator, isn't that string function going to be the opposite
of smart?  How does a string handle '\r' intelligently?

	Donn Cave, donn at u.washington.edu


From abouf066 at aix2.uottawa.ca  Wed Dec 27 17:52:36 2000
From: abouf066 at aix2.uottawa.ca (abouf066 at aix2.uottawa.ca)
Date: Wed, 27 Dec 2000 17:52:36 -0500
Subject: removing directories
Message-ID: <Pine.A41.3.96.1001227175036.37924A-100000@aix2.uottawa.ca>

Greetings,

	Is there a way to force removal of directories even if they
contain files and/or directories. I just tried 

os.rmdir( path ) and removedirs ( path)

both functions complain about existence of files !?

Thanks a lot
_____________________________________________________________
							
					LaBoufarikoise




From trotts at llnl.gov  Sun Dec 31 02:10:17 2000
From: trotts at llnl.gov (Issac Trotts)
Date: Sat, 30 Dec 2000 23:10:17 -0800
Subject: How can I conveniently print an expression?
Message-ID: <20001230231017.A30045@llnl.gov>

Suppose I want to print the expression len(sys.argv).
The most straightforward way to do this is to incant

print 'len(sys.argv) = '+`len(sys.argv)`

Naturally this chafes at my sense of hackerly laziness.
I would prefer to have something in python akin to my 
favorite C++ macro:

#define COUT(expr) cout<<#expr<<" = "<<expr<<endl;

which allows me to write the concise statement

COUT(argc);

rather than the needlessly verbose

cout<<"argc = "<<argc<<endl;

In Python, one might naively try something like

def printexpr(expr_string): 
  expr = eval(expr_string)
  print expr_string+' = '+`expr`

which would be used to say things like

printexpr('len(sys.argv)')

but of course this does not always work because when Python executes
eval(expr_string), it assumes that the elements of the expression
specified by expr_string are to be found in the global scope.  So 
when we write 

def do_stuff():
  i = 2
  printexpr('i')

i = 1
printexpr('i')
do_stuff()

we obtain the following output:

i = 1
i = 1

whereas we were hoping for:

i = 1
i = 2

If there were some mechanism that allowed us to access the variables
of a calling function from within a called function, then the problem
would be solved.  Does anyone know of such a mechanism?

Thanks,
Issac





From gogo at bluedynamics.com  Wed Dec 20 20:02:00 2000
From: gogo at bluedynamics.com (Georg Gogo. BERNHARD)
Date: Thu, 21 Dec 2000 01:02:00 GMT
Subject: How to create GUID UUID like with pythoncom.CreateGuid
Message-ID: <3a4136bb.2555949@192.168.0.1>

Hello, everybody!

I want to  create global unique identifiers with python1.5.2

The algorithm should be platform independend which means that  a
python only solution is needed. I am starting to think that that is
impossible without extending python with some C / C++ functions that
have to be compiled on every platform they are needed, because the MAC
adress of a compuiter is part of the GUID, and there is no way to get
it in python.

There are easy ways get a GUID in WIndows, the COM/RPC API has
functions for that that can be called with pythoncom. But windows is
just one of many platforms, I want to have a more flexible solution.

Maybe someone has already found a good algorithm for this problem,
because it would be useful not only for indexing a database, but also
for XML/RPC and similar techniques.

Ideas are welcome, regards, 
gogo at bluedynamics.com




From amk at mira.erols.com  Thu Dec 28 18:27:31 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 28 Dec 2000 23:27:31 GMT
Subject: A Suggestion for Python Colon Syntax
References: <mailman.977962510.16011.python-list@python.org>
Message-ID: <slrn94njsd.jk.amk@207-172-57-128.s128.tnt2.ann.va.dialup.rcn.com>

On Wed, 27 Dec 2000 19:14:38 -0500, Tim Peters <tim.one at home.com> wrote:
>Guido is often called "the BDFL" here.  That stands for Benevolent Dictator
>For Life.  Python design is in fact a dictatorship, and that's widely
>perceived as one of Python's greatest strengths:  it gives the language an

Also note that there's nothing preventing anyone from forking the
Python source, and this could happen if Guido makes enough bad
decisions.  However, so far most of his bad decisions --

* the __ privacy hack
* the "print >>" syntax
* wanting to deprecate the string module

-- have been fairly minor ones that can be ignored, and people don't
remain BDFL of a project for 10 years because they're stupid, so the
community is naturally reluctant to jettison such a person. (The team
of fanatically-devoted and highly-trained assassins helps.)
Introducing case-insensitivity, optional static typing, or some other
large change, may someday be the breaking point causing a fork, or it
might not.

--amk


From fredrik at effbot.org  Wed Dec  6 15:55:50 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 06 Dec 2000 20:55:50 GMT
Subject: CGI POST binary file upload?
References: <90j8j0$cmt$1@nnrp1.deja.com> <p1dX5.3468$Er5.2143@news.indigo.ie> <zodX5.1943$O%.340817@newsc.telia.net> <kJuX5.3700$Er5.2139@news.indigo.ie>
Message-ID: <qJxX5.1852$Kd1.283280@newsb.telia.net>

Ilariu Raducan wrote:
> It's true mode='rb' should be the default parameter in the function call
> only when the OS is windows.

What are you talking about?

makefile is a method in the standard library, which takes
an optional mode argument.  If you don't like the default
value, to pass in a better value (e.g. "rb") when you
*call* the method.

In other words, replace:

    myfile = mysocket.makefile()

with

    myfile = mysocket.makefile("rb")

(All network modules in the standard library do this when
necessary, so why cannot you?)

Your so-called "fix" changed makefile to completely ignore the
mode argument.  As a result, you'll break any code that passes
in something other than "rb" (like any code using the Socket-
Server's wfile attribute, including BaseHTTPServer, and others).

</F>






From greg at cosc.canterbury.ac.nz  Mon Dec 18 23:08:24 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Tue, 19 Dec 2000 17:08:24 +1300
Subject: Error evaluating numbers starting in zero
References: <mailman.977137021.11074.python-list@python.org>
Message-ID: <3A3EDF38.3F994B6A@cosc.canterbury.ac.nz>

Mike Brenner wrote:
> 
> So, that may also explain the failure with
> 
>         i = 08
> 
> the python interpretor must be using EVAL

Actually, it's the other way around -- eval() is
using the Python interpreter! In other words,
eval() evaluates an *expression* (not just a
number!) according to the rules of the Python 
programming language.

> and with reading in numbers,

The input() function uses eval() to evaluate
the string which is read in. If you don't want
that, use raw_input() and process the string
yourself.

Generally it's a bad idea to use eval() or
input() unless you're sure it's *really* what
you want. Most of the time it's massive overkill,
and can even lead to security problems. For
instance, consider that if someone types
"foo()" in response to an input() statement,
it will cause the foo() function IN YOUR PROGRAM
to be called. Or if they typed

   os.unlink("/home/fred/really_important_file.txt")

the consequences could be unpleasant!

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


From glenfant at equod.com.nospam  Wed Dec 13 17:15:48 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Wed, 13 Dec 2000 23:15:48 +0100
Subject: smtplib
References: <918mdo$7d9$1@nnrp1.deja.com>
Message-ID: <918s6q$qjv$1@reader1.imaginet.fr>

Your machine is not in the list of allowed clients in the SMTP server
config.
Your problem has no connection with Python.
Try to use this SMTP with an usual mail client and you'll get the same
result.

"Brad" <blosh at tsbrass.com> a ?crit dans le message news:
918mdo$7d9$1 at nnrp1.deja.com...
> anyone know what this means...I am running Python 1.6 on HPUX10.20
>
> smtplib.SMTPHeloError: (504, 'DNS verification of sending machine failed:
no
> mail will be accepted')
>
>
> Sent via Deja.com
> http://www.deja.com/



From sav at ulmen.mv.ru  Mon Dec 25 03:47:51 2000
From: sav at ulmen.mv.ru (Alexander Semenov)
Date: Mon, 25 Dec 2000 11:47:51 +0300
Subject: Is there any analog of perl's pack()/unpack() in Python?
Message-ID: <9271lj$2b5u$1@news1.simtel.ru>

Hi!

Is there any way to read/write file with binary coded numbers in it?
I'm just tring to port this perl functions:
sub getlong {
        my($t);
        sysread FIL, $t, 4;
        return unpack('l',$t);
}

sub getfloat {
        my($t);
        sysread FIL, $t, 8;
        return unpack('d',$t)
}

etc...

I know, it's platform-dependent, but python has platform-dependent
modules and IEEE format is rather common.

Any suggestions?

Excuse me for my English, my native is Russian.

Thanks, Best Regards,
Alexander Semenov <sav at hey.to>





From dnew at san.rr.com  Fri Dec  8 12:56:46 2000
From: dnew at san.rr.com (Darren New)
Date: Fri, 08 Dec 2000 17:56:46 GMT
Subject: Python legalities
References: <3A2FC63D.652101A@accessone.com> <3A308881.7FBEBAD4@engcorp.com>
Message-ID: <3A3120DF.9A9F3725@san.rr.com>

Peter Hansen wrote:
> Somehow I doubt the distinction between "free" and commercial software
> is really that clear.  Is it definitely, legally, contractually the case
> that you are completely safe licensing software for money from a third
> party? 

Generally, that's one of the things that goes into the contract, yes. Not
shrinkwrap contracts, mind, but real contracts where both parties agree.
Someone can still come after you, but it's a much cheaper response to say
"No, go talk to Joe" than to fight it yourself.

The other choice, of course, is to buy your own insurance. I would imagine
someone would be glad to insure you against that sort of stuff for a price.

-- 
Darren New / Senior MTS & Free Radical / Invisible Worlds Inc.
San Diego, CA, USA (PST).  Cryptokeys on demand.
Personal malapropism generator free with purchase!
Steganography: The manual is hidden in the source code.


From katz at Glue.umd.edu  Tue Dec 12 10:50:46 2000
From: katz at Glue.umd.edu (Roy Katz)
Date: Tue, 12 Dec 2000 10:50:46 -0500
Subject: Inheriting the @ sign from Ruby
In-Reply-To: <914v6l02ouv@news2.newsguy.com>
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com>
Message-ID: <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu>

On Tue, 12 Dec 2000, Alex Martelli wrote:

> 'self' is not "pasted in front of a name" -- 'self' is an object,
> and it is used with perfectly regular syntax for object access.

Yes, we understand this.  Passing 'self' (or whatever people choose to
call it) to an instance method is automatic upon invocation, and manual
for anything else.  So we have:

   Class X:
      def __init__( self, c ): 
          @c = c     # @c binds to 'self'.c
      def example( self ):  X.__init__( self )
      def same_example( self ): X.__init__( self )
      def another_example( myself, y ): @x = y     # binds to 'myself'.y
   

To me, grammatically and practically, self *is* pasted in as the first
argument.  I'm bringing up a grammatical question.  My question was
whether or not the shorthand was practical. 

> So, you would like '@identifier' to be a special-case
> syntax meaning exactly the same as 'self.identifier' (if
> the first argument, as per normal convention, is named
> 'self').

I would like the @identifier to denote <self-object>.identifier, where
<self-object> is the first parameter to a method. Regardless of whether it
is 'self' or 'myself';  the 'def bind(myself, x)' was just an example. 


> such a violation of
> normal expectations will make your code harder to read
> and understand by any maintainers.

Why should there be any violation if all instance identifiers are labelled
with a @? To be honest, I believe that the @identifier convention would
decrease the number of variant programming styles.  We all want that
'self' be used (instead of any other name).  Fine! the @identifier
scheme would at least reduce this, then. You wouldn't see 'myself' (or
some other deviant) pasted in front of every instance variable; you'd
see a common @ sign. So for reading someone else's code, that's a good
thing. 

> > What faults or strengths do you see with this?
> 
> I see no strength in this supplementary syntax form
> offering a second, different way to express what can
> be perfectly well expressed explicitly today.

shorthand??

> 
> I see it as a fault that all this does is offer a
> second, different way to express what can be perfectly
> well expressed explicitly today.

Save for being able to call the first parameter of a method whatever you
like, wouldn't this reduce the number of programming conventions?

> It would also be a (small) step along the road of

I disagree; Like I mentioned above, this would enforce
standardization. 

> turning the currently crystal-clear syntax of Python
> into that morass of special-purpose, ad-hoc, 'convenient'
> kludges that make certain other languages' typical
> programs look so peculiarly similar to line noise.

Python's language is not crystal-clear, imho.  If we're so tough on
perl for having a kludgy interface, why do we have {} for
defining dictionaries instead of [ key:value ]?  Why do we have to
explain to new Python programmers how the label-value system Python uses
is different than traditional call-by-reference and
call-by-value?  (I've been programming in Python for three years now 
and I understand only that passing an int or float is call-by-value,
otherwise it is call-by-reference.  At the risk of naivete, I would
like a simple & referencing operator so that I can explicitly,
finally, have control over this.) When would I ever use print>> ?
Isn't there a less-ugly way of generating a raw string, instead of placing
an 'r' in front of it? 

Good that you mentioned that the language is deteriorating. IMHO, from 
the above, it's seemed like this from the start. (It's still my favorite 
language). 




But that's just my opinion. 

Roey. 





From ahamm at sanderson.net.au  Thu Dec  7 01:44:00 2000
From: ahamm at sanderson.net.au (Andrew Hamm)
Date: Thu, 7 Dec 2000 17:44:00 +1100
Subject: Attention NYC Shoppers & Visitors - Beware of Rip-off SY Stores!!! ...... kHP2pr9I
References: <66FF28553F3.hlfeano@wmwnkhkpvfv.gov> <3a2f1968_4@corp.newsfeeds.com>
Message-ID: <3a2f30aa$1@news.iprimus.com.au>

Joshua Muskovitz wrote in message <3a2f1968_4 at corp.newsfeeds.com>...
>So, aside from this being basic usenet spam, does anyone know what the text
>at the bottom means?

The whole thing seems like your common, garden variety racial vilification
to me.





From rob at ZOOstation.cc  Sat Dec 23 13:13:49 2000
From: rob at ZOOstation.cc (Rob Brown-Bayliss)
Date: 24 Dec 2000 07:13:49 +1300
Subject: Same again please for OOP
In-Reply-To: <3A42DD57.9BCF7FA1@wag.caltech.edu>
Message-ID: <200012231813.eBNIDoh20571@ZOOstation.cc>

On 21 Dec 2000 20:49:27 -0800, Richard P. Muller wrote:
> Can someone recommend a decent introduction to functional programming in
> Python? I haven't found any suitable intro, and I'd like to learn more
> about map(), apply(), lambda, and the like. I understand the syntax, but
> I'd like to learn something of the functional programming context these
> commands come from.


Like wise,  I am trying to get my head around OOP, but all I can find
are basic howto make a class in C++

I understand how to make classes in both Python and C++.  What I want is
help in deciding waht would be a good class and what would not,
preferably in a pythjon contxt, but a generic context would be wonderful
as well.

Hope you can help.  

thanks


-- 

  Rob Brown-Bayliss
 ---======o======---
  www.ZOOstation.cc



From echuck at mindspring.com  Wed Dec 27 18:19:51 2000
From: echuck at mindspring.com (echuck at mindspring.com)
Date: Wed, 27 Dec 2000 23:19:51 GMT
Subject: Python vs PHP (& Java?)
References: <mailman.977897649.4877.python-list@python.org>
Message-ID: <92dtei$415$1@nnrp1.deja.com>

Ben,

When I started doing serious server-side web development, I switched
from Java to Python for a serious boost in productivity. There is
already plenty of material out there on web sites and c.l.p. archives
regarding this topic so I'll leave it at that.

I also checked out PHP, but found it less complete and less flexible
than Python which is built from the beginning to be easy to use,
thorough and productive.

I naturally checked out Zope, but didn't like it for various reasons. I
ended up starting the Webware project, which provides PHP-like
capabilities as well as other features. Webware is broken into
components so you can use as little or as much as you choose.

The project is active and growing and we'll be making our fifth major
release this January.

http://webware.sourceforge.net


-Chuck



In article <mailman.977897649.4877.python-list at python.org>,
  "Ben Ocean" <zope at TheWebsons.com> wrote:
> This is a multi-part message in MIME format.
>
> ------=_NextPart_000_0037_01C06F88.0218F7C0
> Content-Type: text/plain;
> 	charset="Windows-1252"
> Content-Transfer-Encoding: 8bit
>
> Hi;
> I?m an intermediate-level programmer that?s new to Python (via Zope)
and I
> like it a lot. However, it appears to me that, while Python is a good
> replacement for Perl, it cannot replace much of the functionality of
PHP,
> which can be scripted into HTML (whereas Python, it appears, must be
> called). So, I have to wonder about the speed of Python vs. PHP for
> Internet-based applications. Python has the advantage of byte-code and
> compilation, but it?s server-side. PHP might not be as fast on the
server,
> but it doesn?t have to make as many trips back and forth between
server and
> client. Plus, I?m curious as to why Python wasn?t written in such a
way that
> it could be written directly into HTML documents like PHP. On a
similar
> subject, someone once informed me that Java was by far the best
language for
> scripting communication programs over the Internet (like chat, IRC,
etc.).
> How does Python compare? Your comments are appreciated.
> BenO
>


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


From pehr at linfoo.eink.com  Wed Dec  6 16:53:55 2000
From: pehr at linfoo.eink.com (Pehr Anderson)
Date: Wed, 6 Dec 2000 16:53:55 -0500
Subject: [pygtk] Python and Orbit Examples
In-Reply-To: <200012060632.eB66WIK27614@ZOOstation.cc>; from rob@ZOOstation.cc on Tue, Dec 05, 2000 at 06:32:17PM -1200
References: <Pine.LNX.4.21.0012050828550.24722-100000@tupelo.daa.com.au> <200012060632.eB66WIK27614@ZOOstation.cc>
Message-ID: <20001206165355.B12869@linfoo.eink.com>

Dear Folks,

I would highly recommend using XML-RPC as a minimally complicated
way to communicate data between your apps.

This might sound complicated but it really only means
	1. specify XML encodings for your data, use these throught your code
	2. open a socket on the target
	3. send an XML query and wait for an XML response
	4. parse the response

The beautiful thing about XML-RPC (remote proceedure call) is 
that it is just dirt simple. You open a socket, dump in 
human readable text, and wait for a human-readable response.
Everything gets completely defined somewhere and servers and
clients could be coded in *anything*! 
No complicated tools are required to implement any parts of
this system though you might want to look at how others 
parse XML rather than doing it all by hand.

And the best part is you'll have an app that anyone can understand.
No secret church of CORBA, COM or other "special tools".
XML is directly comprehensible by HUMAN BEINGS and that makes it
a sustainable encoding with simple tcp sockets for transport.

	-pehr


On Tue, Dec 05, 2000 at 06:32:17PM -1200, Rob Brown-Bayliss wrote:
> >or getting it to work with gnome-python, that is a bit more difficult.
> 
> Sounds a little ominous...
> 
> What I am doing (or planning at this stage) is to write a POS and Stock 
> management system for a small retail chain.
> 
> Would I be better of giving Pyhton a miss, or ditching the Corba idea? 
> 
> I have very little knowledge of linux programming, though I do have 
> years of basic programming on the Amiga (like 18 I think if include vic 20 and c64)
> and a few months VC++ on windows.  I have used Linux for 18 months now, and played
> a bit with C and a very very little with Python. 
> 
> I am confident I can tackle the task in either language, but I know python will be
> quicker to develop with.
> 
> I had looked to corba as a means of communicating between the stores, I could always
> use some form of file transefer, but it's not as elegant or nearly as real time.
> 
> 
> 
> 
> -- 
> 
>   Rob Brown-Bayliss
>  ---======o======---
> 
>   www.ZOOstation.cc
> 
> _______________________________________________
> pygtk mailing list   pygtk at daa.com.au
> http://www.daa.com.au/mailman/listinfo/pygtk



From aleaxit at yahoo.com  Tue Dec  5 15:29:42 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 5 Dec 2000 21:29:42 +0100
Subject: SAP R/3 automation with win32com
References: <90flpu$gpu$1@nnrp1.deja.com> <kl5n2tsfk3hud1c37gjo56290sueccuegk@4ax.com> <90i6n6$hg0$1@nnrp1.deja.com> <90icc6022q9@news2.newsguy.com> <90j81i$c4a$1@nnrp1.deja.com>
Message-ID: <90jjgi010ib@news1.newsguy.com>

<obou at my-deja.com> wrote in message news:90j81i$c4a$1 at nnrp1.deja.com...
    [snip]
> I abuse, but is there a way to find the "Makepy" name of a particular
> COM object? I have tons of SAP objects in Makepy and cannot find one
> that correspond to the objects I'm using...
> Do somebody know where Makepy looks for the titles it shows in it's
> listbox?

In the type-library's "helpstring" propriety.  But that is the same
place from which Ole View, AND Visual Basic's "Project/References..."
dialog, also get it, so, if you can find them with any one of these
tools, the others should be no problem.

Alternatively, the EnsureDispatch method of gencache can (probably)
work on an object you already have instantiated (if that object is
reasonably well-behaved) and perform the same work as makepy
would.  Starting with a clean-slate gen_py, you can just then open
with your favourite text-editor the only long-named .py file you'll
find there.

Or, you can do it by studying the registry.  From a ProgId (such as
"feefie.foofum", in the registry under HKEY_CLASSES_ROOT you'll
find the CLSID; under HKEY_CLASSES_ROOT/CLSID/thatclsid, you'll
find the type-library's own GUID, which you can in turn use as a
key under HKEY_CLASSES_ROOT/Libraries to find out everything
about the location of the specific type-library of interest.  OleView
does that for you, by the way, if you start your browsing with an
object rather than going directly for the libraries.


Alex





From jerry_spicklemire at my-deja.com  Fri Dec  8 13:50:50 2000
From: jerry_spicklemire at my-deja.com (jerry_spicklemire at my-deja.com)
Date: Fri, 08 Dec 2000 18:50:50 GMT
Subject: Parallel Computing
References: <3A300B1D.223630F3@bioeng.ucsd.edu>
Message-ID: <90rai7$ugg$1@nnrp1.deja.com>

Curtis Jensen wrote:
> I haven't had much luck on finding anything about parallel computing
> via Python.  Does Python support this?  Or do we just need to write a
> C> module to run parallel code?  Thanks.

Look here:

 http://www.vex.net/parnassus/apyllo.py/126307487.549099265

Have Fun!
Jerry S.


Sent via Deja.com http://www.deja.com/
Before you buy.


From mwh21 at cam.ac.uk  Thu Dec  7 07:52:37 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 07 Dec 2000 12:52:37 +0000
Subject: Printing modules contents
References: <3A2F84CC.3056C755@hotmail.com>
Message-ID: <m3itow4cmi.fsf@atrus.jesus.cam.ac.uk>

joonas <keisari_ at hotmail.com> writes:

> How can I make a script that would print modules contents like below,
> 
> sys
>   sys.__name__ = "sys"
>   sys.stdout = "<PyShell.PseudoFile instance at 7fd530>"
>   sys.winver = "1.5"
>   sys.etc = "spam"

Think "dir" & "repr".  Shouldn't be hard.

Cheers,
M.

-- 
39. Re graphics:  A picture is worth 10K  words - but only those 
    to describe the picture. Hardly any sets of 10K words can be 
    adequately described with pictures.
  -- Alan Perlis, http://www.cs.yale.edu/homes/perlis-alan/quotes.html


From fredrik at effbot.org  Wed Dec 13 12:37:24 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 13 Dec 2000 17:37:24 GMT
Subject: PIL binary for Python 2.0 for Windows ?
References: <3a2f472f$0$235@hades.is.co.za> <KZOX5.2015$Kd1.294314@newsb.telia.net> <917v4j$it3$1@news.fas.harvard.edu>
Message-ID: <otOZ5.3825$Kd1.435667@newsb.telia.net>

William Ryu wrote:
> I've downloaded the PIL Python Modules from eefbot
> (http://w1.132.telia.com/~u13208596/pil/index.htm) and have loaded the
> _imaging20.dll into my DLLs folder.

That page also says "save it as _imaging.dll".  If it's called _imaging20.dll,
Python won't find it.

If you get rid of the "20", it should work.

</F>




From aleaxit at yahoo.com  Tue Dec 19 09:05:24 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 19 Dec 2000 15:05:24 +0100
Subject: missing HtmlBuilder
References: <91lsh8$9p7$1@nnrp1.deja.com>
Message-ID: <91npvu018ie@news2.newsguy.com>

<ennisr at my-deja.com> wrote in message news:91lsh8$9p7$1 at nnrp1.deja.com...
> I am a relative newbie to Python, using Python 2.0 under Windows.
> 4DOM from Fourthought, Inc. has also been installed.
>
> I have run into a problem trying to run the DOM examples from David
> Mertz article "Charming Python: The dynamics of DOM"
> (http://gnosis.cx/publish/programming/charming_python_2.txt).
>
> Modules referenced in the following import statements seem to be
> missing.
>
>     from xml.dom import core
>     from xml.dom.html_builder import HtmlBuilder
>     from xml.dom.utils import FileReader
>
> Do I have the wrong software loaded?  Where can I find the missing
> modules?

I think you need the PyXml stuff, e.g. from
    http://sourceforge.net/projects/pyxml
probably in some older version.

4DOM may have equivalent or better functionality, but
I suspect it will not match what you read in that
online book, which appears to target older PyXml stuff.


Alex





From gustav at morpheus.demon.co.uk  Wed Dec 20 16:07:04 2000
From: gustav at morpheus.demon.co.uk (Paul Moore)
Date: Wed, 20 Dec 2000 21:07:04 +0000
Subject: Events in win32com
References: <Y1g=Op6IfWSoJn1ISFtOyuvqwBb3@4ax.com> <FY0=OlZFLy71DtnK62u5TEPlhOts@4ax.com> <3A40060B.10504@ActiveState.com> <cYxAOmxi9FgNERgBo29zuCl5c2h1@4ax.com>
Message-ID: <ik724tc2v3cpcqfbt7i1ookvaona0mk5ql@4ax.com>

On Wed, 20 Dec 2000 11:50:22 +0100, Paul Moore
<paul.moore at uk.origin-it.com> wrote:

>Grrr. I'm having problems getting events fired from any of the MS
>applications (IE, Excel), either in VBScript or in Python. Looks like
>it's something stupid I am doing... We'll ignore that for now...

Got it working (on a different machine). Looks like it's either a
config problem or an intermitent one. Everything looks great now.
Thanks for all the help.

One further question - another way of doing the same thing (in WSH) is
to assign to the "onclick" property of the HTML document in IE.
Something like

ie.document.onclick = function() { WScript.Echo("Click!") }

Is it possible to do anything like this in Python? If not, can I do
the same thing another way? (I have created ie via
DispatchWithEvents(), but the HTML document is a property of this. Can
I rewrap it somehow, say as
  doc = DispatchWithEvents(ie.document, DocEvents)
?

Thanks,
Paul


From Bill.Scherer at VerizonWireless.com  Wed Dec 27 15:58:32 2000
From: Bill.Scherer at VerizonWireless.com (Bill Scherer)
Date: Wed, 27 Dec 2000 15:58:32 -0500
Subject: HTTPServer, performance
Message-ID: <3A4A57F8.CCE41022@VerizonWireless.com>

Hello eveyone, hope you had a nice holiday...

Question: Is the code below the proper way to create a multi-threaded
http server in Python, and what kind of performance should one expect
out of it?

The code below works.  It's an extremely stripped down verison of my
current project which really runs under J(p)ython. I was unsatisfied
with performance, so I stripped out all that wasn't web server and tried
it under CPython and JPython.  

With the code below, CPython does about 12 tps on a dual 733 Mhz machine
under RedHat 6.2. JPython is only slightly slower (when using a JVM that
uses native threads and therefore both processors).  Apache on the same
box does over 70 tps with a greater number of simulated users (100 vs
25).

One thing I do see with both CPython and JPython running this code is
random stalling.  ie. The server will be happily serving requests, then
for no apparent reason, it justs stops.  CPU load drops during the
stall, although it never gets very high. Two, maybe three, seconds
later, it resumes serving.  Pings during this time show no change to
network latency (0.4ms).

Any and all help is appreciated.

Thanks!

--------------------------------------------------------------------------------

import SocketServer, BaseHTTPServer

PORT = 9777

class HTTPServer(SocketServer.ThreadingMixIn,
BaseHTTPServer.HTTPServer):
    pass

class RequestHandler(BaseHTTPServer.BaseHTTPRequestHandler):
    def do_GET(self):
        self.send_response(200)
        self.send_header("Content-type", 'text/plain')
        self.end_headers()
        self.wfile.write('hello world')
        return

if __name__ == '__main__':
    httpd = HTTPServer(('', PORT), RequestHandler)
    print 'serving on port', PORT
    httpd.serve_forever()


-- 
William K. Scherer
Sr. Member of Applications Staff - Verizon Wireless
Bill.Scherer_at_VerizonWireless.com



From jinsongliang at yahoo.com  Thu Dec 14 03:08:42 2000
From: jinsongliang at yahoo.com (Jinsong Liang)
Date: Thu, 14 Dec 2000 01:08:42 -0700
Subject: newbie question: python vs. Matlab
Message-ID: <3A38800A.6CE1C201@yahoo.com>

Hi every body,

I am a control engineer and familiar with Matlab. It seems to me that
python is similar to Matlab in some sense. Who can tell me what the
advantage of python over Matlab is?

Thanks.

Leo



From cnelson9 at gte.net  Sat Dec 23 02:11:46 2000
From: cnelson9 at gte.net (Clifford J. Nelson)
Date: Sat, 23 Dec 2000 07:11:46 GMT
Subject: Python in "math through programming" curriculum
References: <2g2t3t06r2agrrjmfnr41eu4lc2u39a8qi@4ax.com> <3A3EFF47.67DE69CF@gte.net> <kj624toonmgugtrgrpkea9b01a56hgogam@4ax.com> <3A413A14.54DF58D1@gte.net> <laj24tcfio6ldfqktnehkck93j7v6iciiv@4ax.com>
Message-ID: <3A445754.E88D0A01@gte.net>


Kirby Urner wrote:
> 
> "Clifford J. Nelson" <cnelson9 at gte.net> wrote:
> 
> >> I went to this website.  It doesn't work for the Mac at all --
> >> Windows only.  So what's your beef with Python on the Mac?
> >> At least it exists for the Mac.
> >
> >That's why I don't attach ANY weight to what you write!
> >Clean used to be a "Mac first product", and Version 1.3.x
> >soon to be updated to Version 2.0 runs on the Mac. You would
> >know that it runs on the Mac if you could just read.
> >
> 
> The latest version with the fancy games library doesn't
> run on the Mac yet (or Linux either).  Older versions
> do, yes, I see that.
> 
> I thought maybe you were advocating Clean for kids because
> of the games angle -- what I asked you about.  Apparently
> not so, given the older Clean suits your purposes (what
> do you do with it??).
> 
> >> Do you think this emphasis on game programming is better
> >> than my "math through programming" approach?  It seems to
> 
> So then... what's so attractive about Clean, from your
> personal experience?  I'm just asking.  You haven't said
> anything about that yet (older versions run on the Mac...
> ... ... so?).
> 
> Python runs on the Mac too (the latest version even).
> 
> http://www.python.org/download/download_mac.html
> http://www.python.org/doc/current/mac/mac.html
> 
> << repeated text snipped >>

I'll just repeat it again. Maybe it will sink in eventually.

I take it personally when Bucky Fuller's Synergetics (see a)
is made to seem like some far out religion, by you Kirby
Urner, or anyone else, or when computer programming in a
language (Python) that can't apparently help itself well
enough to use it to make a good interactive programming
environment for the Mac, is promoted over much easier
languages, like Clean (see b) or Object Caml 3.0 (see c) or
even the hard to learn J programming language (see d).

All of those languages (including Python) are garbage
compared to Mathematica of course, because programming (OOP)
is all just really pattern matching and replacement (see e),
but they're free and they work on both the Mac and Windows,
so they are worth looking at.

a) http://www.rwgrayprojects.com/synergetics/synergetics.html
b) http://www.deene.ufu.br/clean/clean1.html
c) http://pauillac.inria.fr/caml/
d) http://www.jsoftware.com/
e) http://www.stephenwolfram.com/publications/talks/IMS/imstalk.html

Of course, when I say "look at them", I mean download the
version that runs on your computer, read all the
documentation, and try programming with them for a year or
two like I have. Then you won't have to ask so many simple questions.

  Cliff Nelson 
[snip]


From brian_zhou at agilent.com  Sat Dec  9 17:58:16 2000
From: brian_zhou at agilent.com (Brian Zhou)
Date: Sat, 9 Dec 2000 14:58:16 -0800
Subject: Newbie : Unable to use Jython under Win98
References: <OylY5.47189$%j3.534830@news6.giganews.com>
Message-ID: <976402693.883545@emperor.labs.agilent.com>

I just tried it on a win98 + JDK w/o any problem.

Possible cause:
1. I'm using JDK1.3 vs JRE, do you have to use JRE?
2. When I invoke the installer, I used full path c:\jdk1.3\bin\java
3. It tried to install into C:\windows\jython-2.0a1, I changed it to
C:\jython-2.0a1.

You might try jython-user/jython-devel mailing list, providing more info and
ask for help.

Good luck,

-Brian

"Patrick" <patrick.fichou at galactica.it> wrote in message
news:OylY5.47189$%j3.534830 at news6.giganews.com...
> Hi,
>
> I can't use Jython (Jython-20a1.class) under windows whith the JRE from
Sun
> (Jre1_3_0_win). With the outdated Microsoft virtual machine (Version
> 5.00.3319) it seems to run, but with an "old" Java 1.x engine !
> Is there something important to do before I can use Jython ?
>
> * Autoexec
> * Path
> * Install in a special directory (I've tried \Jython and \Windows\Jython)
> * Reg modification
> *...
>
> Thanks
>
>
>




From jepler at inetnebr.com  Tue Dec 12 22:47:40 2000
From: jepler at inetnebr.com (Jeff Epler)
Date: Wed, 13 Dec 2000 03:47:40 GMT
Subject: Fault trees structure
References: <D_IY5.49$17.2785@m2newsread.uni2.es> <9128e902uo3@news2.newsguy.com> <z8eZ5.249$YZ4.8868@m2newsread.uni2.es> <yfhZ5.23310$Ia1.4003773@typhoon.nyroc.rr.com> <91521d02r3g@news2.newsguy.com> <915pu2$6gf$1@news.kodak.com>
Message-ID: <slrn93dsas.9tj.jepler@potty.housenet>

On Tue, 12 Dec 2000 13:18:58 -0500, Darrell Gallion
 <darrell at dorb.com> wrote:
>Not sure that values will be constrained to boolean.
>
>When I tried this something odd happened.
>None isn't valid for and_ while it is for "and"
>
>>>> and_(1, 1)
>1
>>>> and_(None, 1)
>Traceback (most recent call last):
>  File "<stdin>", line 1, in ?
>TypeError: bad operand type(s) for &
>>>> None and 1
>>>>

Well, operator.and_ corresponds to writing
	x & y
what you're looking for is the function in operator which means
	x and y
but that doesn't seem to exist.  :(

Jeff


From howard at eegsoftware.com  Tue Dec 12 13:29:35 2000
From: howard at eegsoftware.com (howard at eegsoftware.com)
Date: Tue, 12 Dec 2000 18:29:35 GMT
Subject: mmap problem with dll
Message-ID: <3a366959.3497335@news-server.socal.rr.com>

I am trying to back the mmapmodule.cpp into 1.5.2.

I know I know but, I am using PIL,Pmw,Blt,Imaging,Numeric, and serial
and have NOT been able to find a working combination under 2.0

After changing the few error message outputs back to some compatible
format, I can build a DLL containing the modified mmapmodule (using a
new name of course).

However, it doesn't work.

I get the message 
     EnvironmentError: [Errno 2] No such file or directory
on the mmap object creation call.

I traced it back to the _get_osfhandle call trying to get the
operating system equivalent of the run-time library handle.

So, I rebuilt the dll to use the Python 2.0 includes and tried it
under 2.0.  It still doesn't work.  The built-in mmap DOES work.

In the dll, I printed out the result from _get_osfhandle....sure
enough it returned -1.  In the Python code, I printed out
msvcrt.get_osfhandle and it works correctly.

My test code (fragment):

        hfile=win32file.CreateFile( filename,\
                    win32con.GENERIC_READ|win32con.GENERIC_WRITE,\

win32con.FILE_SHARE_READ|win32con.FILE_SHARE_WRITE,\
                    secur_att,\
                    win32con.OPEN_ALWAYS,\
                    win32con.FILE_ATTRIBUTE_NORMAL , 0 )
        rslt=win32api.GetLastError()
        if rslt == 0:
            # here if file was CREATED
            #mmap to this file
            fd=msvcrt.open_osfhandle(hfile.handle,0)
            if mstyle == 20:
                map=mmap.mmap(fd,self.filesize)
            else:
                map=mmapnew.mmap(fd,self.filesize)
            # other stuff
        elif rslt == winerror.ERROR_ALREADY_EXISTS:
            # here if it already existed
            #mmap to this file
            fd=msvcrt.open_osfhandle(hfile.handle,0)
            print hfile.handle,fd,msvcrt.get_osfhandle(fd)
            if mstyle== 20:
                map=mmap.mmap(fd,filesize)
            else:
                map=mmapnew.mmap(fd,filesize)
            # other stuff
       else:
           print 'Some other error',rslt


Any clues on why the dll version won't get back a correct os file
handle?

Thanks

Howard Lightstone
EEGSoftware
howard at eegsoftware.com


From gherman at darwin.in-berlin.de  Sun Dec 17 15:08:41 2000
From: gherman at darwin.in-berlin.de (Dinu C. Gherman)
Date: Sun, 17 Dec 2000 21:08:41 +0100
Subject: Singleton vs Proxies DP (was Re: Solution: Direct access to Printer  
 I/O lines)
References: <917k9b$9i9$1@nnrp1.deja.com> <slrn93hrjn.2vp.ssthapa@localhost.localdomain> <91cnp3$g59$1@nnrp1.deja.com> <91d23f0jk7@news1.newsguy.com> <91gfnp$7r4$1@animus.fel.iae.nl> <91gkh801e49@news2.newsguy.com> <3A3CB83D.45C8775E@darwin.in-berlin.de> <91iu95013ta@news2.newsguy.com>
Message-ID: <3A3D1D49.1A453B51@darwin.in-berlin.de>

Alex Martelli wrote:
> 
> I agree that unique id is part of the Singleton Pattern:
> that is what makes it problematic.  One does not really
> _care_ about that identity, only about the state and
> behavior it represents.

For me it sounds like being part of the essence of that
pattern... but you're free to disagree.

> No!  I'm not saying two clients can't hold their
> 'handles' (lw proxies) at the same time.  It IS
> pretty typical.  But it's not at all typical for client
> code to be interested in ensuring _identity_ --
> state and behavior are typical.

In the context of Singletons I would even reduce that to
state only, like with counters of some system-wide entities
of I-don't-know-what. A client is "typically" not able to
compare *bahaviour*, but state, yes. In all cases where
this comparison is difficult or time-consuming, comparing
or, probably more often in fact, *accessing* the current
state via a "globally identical" object is what a Singleton
is about - if I understood anything at all about this
pattern.

Regards,

Dinu

-- 
Dinu C. Gherman
................................................................
"The only possible values [for quality] are 'excellent' and 'in-
sanely excellent', depending on whether lives are at stake or 
not. Otherwise you don't enjoy your work, you don't work well, 
and the project goes down the drain." 
                    (Kent Beck, "Extreme Programming Explained")


From dale at out-think.NOSPAMco.uk  Tue Dec  5 04:25:31 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Tue, 05 Dec 2000 09:25:31 +0000
Subject: Python programs (or scripts) as drop targets
References: <hv5k2tg9cdf33j8v333qpkk7jo45ht596n@4ax.com> <90h0gq$cu0$1@troll.powertech.no> <1mao2tojod77d92ri841ptdvm6t3e0evqu@4ax.com> <3A2C7968.4050205@ActiveState.com>
Message-ID: <j3dp2t0o2p9d5i4hfak9ttl3selngjqcgq@4ax.com>

Mark Hammond <MarkH at ActiveState.com> wrote:

>Dale Strickland-Clark wrote:
>
>> Mark has suggested  that it could go in win32com.shell - although I can't figure out what he means
>> by that - win32comext.shell, perhaps?
>
>Yes - it is imported as "win32com.shell", but exists in a different tree 
>as it is not part of the "COM core".  Such a distinction is a little 
>pointless...
>
>
>> Is the source for this available?
>
>With the rest of the sources.  See my win32all pages.
>
>Mark.

Thanks.
--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants




From jdries at mail.com  Tue Dec 26 20:33:22 2000
From: jdries at mail.com (Jan Dries)
Date: Wed, 27 Dec 2000 02:33:22 +0100
Subject: sniffing btw my web browser and a server
References: <92aush$rhc$1@news.nuri.net> <270767A74DE6B86D.ED3C49CF7CD80055.4014344C6ECF85B8@lp.airnews.net>
Message-ID: <3A4946E2.3F0ADE37@mail.com>

>I'm doing some sort of web client programming, and have to observe the
>messages that my browser and web-servers transmit, being tired of guessing
>what the messages'd be and looking up the references and the specifications.
>
>Anyone has done it in Python or knows how it could be done?
>

If all you want is to see the trafic between a browser and server, you
basically need a proxy server. If it's for testing purposes, and if it's
just to see the trafic (i.e. without interpreting/altering it), this
isn't too hard to do. I did it in C once, and it took me no more than 2
or 3 pages of code. It should be shorter in Python. 
If you want to monitor trafic to one particular server, it's even
easier, since you don't have to extract the name/ip address of server to
connect to from the info sent by the browser.
Your program must just listen on a port the browser connects to (e.g.
8080), and connect to a port the server is listening on (e.g. 80) and it
just sends al the info it receives on port 8080 to the connection on
port 80, and sends everything it receives back from port 80 to the
connection on port 8080. And in the mean time it logs everything it
sends and receives. Keep doing this till either the sender or the
receiver closes the socket.
And that's more or less it. 

Regards,
Jan



From redjade at my-deja.com  Fri Dec  8 02:04:55 2000
From: redjade at my-deja.com (redjade at my-deja.com)
Date: Fri, 08 Dec 2000 07:04:55 GMT
Subject: SIGCHLD handle & socket
Message-ID: <90q16m$tm2$1@nnrp1.deja.com>

I am writing a forking-type server with python.
My plan is:
1) first, open a socket with port 15151
2) if there's a connection request, fork a child
3) child will die after 3 sec of connection establishment.
4) parent should handle SIGCHLD signal so that it make zombie-child
disappear.
5) parent should allow any connection request and should be fork-able.

my code is below
----------------
#!/usr/local/bin/python

import socket, os, signal, time

def handle(signum, frame):
        print 'CHILD terminated'
        os.waitpid(-1, os.WNOHANG)

HOST = ''
PORT = 15151

s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
s.bind( (HOST, PORT) )
s.listen(6)

signal.signal(signal.SIGCHLD, handle)

while 1:
        (conn, addr) = s.accept()
        print 'Connected by', addr
        pid = os.fork()
        if pid == 0:
                s.close()
                conn.send('\n\nConnection Established\n\n\n')
                time.sleep(3)
                conn.close()
                os._exit(0)
        else:
                conn.close()
---------

and error occurred.

$ python rjcall.py
Connected by ('147.46.102.27', 59760)
CHILD terminated
Traceback (most recent call last):
  File "rjcall.py", line 19, in ?
    (conn, addr) = s.accept()
socket.error: (4, 'Interrupted system call')


I can't figure out what is wrong. i think code above follows basic
c-style forking-type daemon. working environment is python2.0 and
FreeBSD-4.2


Sent via Deja.com http://www.deja.com/
Before you buy.


From victor at prodigy.net  Fri Dec 29 00:23:09 2000
From: victor at prodigy.net (Victor Muslin)
Date: Fri, 29 Dec 2000 05:23:09 GMT
Subject: fcgi.py
References: <3a4a865c.1325844015@localhost> <3a4a90ab.1328483109@localhost> <92f20t$re4$4@newsserver.rrzn.uni-hannover.de>
Message-ID: <3a4c1e99.60463812@newnews.prodigy.net>

Well, my most important need was not to have to open database connections every
time CGI executed. Currently I don't need multiple instances of the script
running simultaneously. If and when I do, I plan to add multithreading to my
FastCGI library (I actually believe it had already been done).

I got it running so far with cgi-fcgi.exe bridge. Though it has to fork/exec
cgi-fcgi program every time, it is still an order of magnitude faster than
opening ODBC connection. To get it running on NT, I modified (very slightly)
Robin Dunn's fcgi.py and built fcgi2.py on top of it that uses a socket instead
of a named pipe. You can get my modules at
http://pages.prodigy.net/victor/fcgi2_py. Please report any bugs to me (and
perhaps how to fix them since I am still rather new to Python).

I am yet to attempt to use mod_fcgi (I am new to Apache as well). This is my
next step once I figure out how to configure Apache on WinNT for it (any help is
appreciated :-). Once I get this working I plan to try either mod_python or
mod_snake (any help for configuring Apache with those is appreciated as well :).

On 28 Dec 2000 09:43:57 GMT, bernhard at abnoba.intevation.de (Bernhard Reiter)
wrote:

>In article <3a4a90ab.1328483109 at localhost>,
>	victor at prodigy.net (Victor Muslin) writes:
>
>> I just found an old posting that says that socket.fromfd() is not
>> implemented on Windows platform (though Python documentation does not
>> say so). Is there a FastCGI solution for Python for WindowsNT???
>
>Not that I know of. :(
>I ran into the same problem and the solution to fix fcgi.py to use
>named pipes under WindowsNT is not trivial. Just peek at the mod_fastcgi
>sources themself to see what tricks they have to pull under WindowsNT.
>(Another problem is that the versions of apache I have tried under
>WindowsNT all were buggy regarding fcgi. They did not even run with
>the fastcgi c examples and error reporting was lousy.)
>
>The part of fcgi and apache that you want of course is the
>dynamic capabilities of apache to create as many fcgi instances as
>needed. Fortunatly there seems to be another solution, which still
>stands to be tested here:
>
>Use mod_python with apache and program your python script in a way
>that it will do most of its stuff when importing the modules onces.
>Then subsequent calls in the same apache thread will be in the same
>name space of the python subinterpreter. So you can use the apache
>thread load balancing of the server and answer multiple requests
>fast.
>
>Let me know if you find out more!
>	Bernhard



From aleaxit at yahoo.com  Wed Dec 20 07:28:14 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 20 Dec 2000 13:28:14 +0100
Subject: Python's scope rules
References: <3A40813B.C04D21C@et.dtu.dk>
Message-ID: <91q8lk01ift@news2.newsguy.com>

"Brian Elmegaard" <be at et.dtu.dk> wrote in message
news:3A40813B.C04D21C at et.dtu.dk...
    [snip]
> On usenet I have now learned from skilled scientists, that Python has some
> deficiencies regarding scope rules, and I am not capable of telling them
why it
> has not (or that it has).

Some people strongly dislike Python's very simple scope rules.  I
happen to like them, although they may get richer (and, alas, less
simple) pretty soon.


> In Aaron Watters 'The wwww of python' I have read that Python uses lexical
> scoping with a convenient modularity. But, other people tell me it does
not.
> Some say it uses dynamic scoping, some say it uses it own special
> 'local-global-builtin' scope. What is right?

The only totally-wrong idea here is that Python uses dynamic scoping:
it doesn't!  It's lexical, but *not nested*; which equates to the
local-global-builtin idea.


> The above mainly is a theoretical question. A more practical example which
I
> agree seems a bit odd is:
>
> Python 1.5.2 (#1, Mar  9 2000, 17:40:34)  [GCC 2.95.2 19991024 (release)]
on hp-
> uxB
> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>> x=24
> >>> def foo():
> ...     y=2
> ...     print x+y
> ...
> >>> def bar():
> ...     x=11
> ...     foo()
> ...
> >>> bar()
> 26
> >>>

bar's "x=11" defines its x as a local variable.  Remember
Python does not have 'declarations': the 'assignment
statement' is what defines the variable's existence.

foo doesn't _define_ x, just _uses_ it, and thus, since
there is no x in local scope, the global scope is then
searched, and an x (bound to 24) is found there.

If bar wants to affect ('re-bind') the _global_ variable
x, rather than defining its own local x, it must be
explicit about it:
    def bar():
        global x
        x = 11
        foo()

However, relying on the modification of global variables
is more often a bad idea rather than a good one; if you
avoid it wherever you can do so without excessive hardship,
you'll be quite happy about the several maintenance issues
you'll eschew (and this is not a specifically Python issue;
mutable global variables can be troublesome in any language
that supports them).


> Well, I can live with it. In Fortran I would always put all the variables
to be
> passed in the parameter list. Other people say that Fortran is not easy,
and
> this example shows how stupid python is. It does?

In Fortran, you were QUITE wise to use parameter lists for
their intended purpose, rather than (ab)using the COMMON
(or named-common) area(s).  Having had to maintain not a
little Fortran in my time, I wish all Fortranists were as
wise as you:-).  Follow similar strategies in Python, and
you'll have similar advantages.

Python offers you another easy possibility, which is to
bind values to *instance attributes* -- and THAT is often
easiest and most powerful.

If anybody thinks that Python's requirement of an _explicit_
'global' statement, when global variables must be modified,
is 'stupid', then I suspect they have not thought deeply
enough about the issues, or else lack real-world experience
to have taught them why global variables NEED to be handled
with care and restraint.

This has relatively little to do with Python scoping rules,
by the way; those enter the issue when you have a function
nested inside another function, etc.  The fact that global
variables are accessed implicitly in functions that only
'read' them, but need an explicit 'global' statement in
functions that want to *re-bind* them, is a separate issue,
and, I dearly hope, one that will be maintained if and
when nested scopes are introduced: _accidental_ rebinding
of variables in outer scopes would otherwise be a danger!


> Hopefully not, but probably, this is a faq, but I have not been able to
find
> anything on the matter.

http://www.python.org/doc/FAQ.html, question 4.36, is
probably the most-relevant one -- because this is not
really a scope issue.  The 'local scopes do not next'
issue is addressed in question 4.5.


Alex





From wsryu at is07.fas.harvard.edu  Wed Dec 13 20:55:06 2000
From: wsryu at is07.fas.harvard.edu (William Ryu)
Date: 14 Dec 2000 01:55:06 GMT
Subject: PIL binary for Python 2.0 for Windows ?
References: <3a2f472f$0$235@hades.is.co.za> <KZOX5.2015$Kd1.294314@newsb.telia.net> <917v4j$it3$1@news.fas.harvard.edu> <otOZ5.3825$Kd1.435667@newsb.telia.net>
Message-ID: <91999q$t2$1@news.fas.harvard.edu>

In article <otOZ5.3825$Kd1.435667 at newsb.telia.net>,
Fredrik Lundh <fredrik at effbot.org> wrote:
>William Ryu wrote:
>> I've downloaded the PIL Python Modules from eefbot
>> (http://w1.132.telia.com/~u13208596/pil/index.htm) and have loaded the
>> _imaging20.dll into my DLLs folder.
>
>That page also says "save it as _imaging.dll".  If it's called _imaging20.dll,
>Python won't find it.
>
>If you get rid of the "20", it should work.

Yes. Indeed it does, right on top of the page. My apologies and thanks for
the information.

-wr




From root at rainerdeyke.com  Tue Dec 19 16:15:31 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Tue, 19 Dec 2000 21:15:31 GMT
Subject: while true: !!!
References: <B43D149A9AB2D411971300B0D03D7E8B0FBCBB@natasha.auslabs.avaya.com> <v04220801b663769e1800@[192.168.11.33]> <mailman.977147642.18708.python-list@python.org> <91lclt02qr3@news1.newsguy.com> <slrn93st0m.6c1.qrczak@qrnik.zagroda> <91o3l501i59@news2.newsguy.com> <jJN%5.52075$x6.23325177@news2.rdc2.tx.home.com> <91of2t01um1@news2.newsguy.com>
Message-ID: <TdQ%5.52154$x6.23511368@news2.rdc2.tx.home.com>

"Alex Martelli" <aleaxit at yahoo.com> wrote in message
news:91of2t01um1 at news2.newsguy.com...
> If you mean 'char is a numeric type' (in C but not Python),
> I wouldn't dispute that

I mean that 'char' is a numeric type in the integer family, i.e. a numeric
type for which operators |, &, ^, >>, <<, ++, and -- are defined, and (a +
b) - b always equals a in the absence of overflow.  Numeric types in the
integer family are collectively refered to as integers; 'int' is one
specific type in this family.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From fredrik at effbot.org  Tue Dec 19 19:36:05 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 20 Dec 2000 00:36:05 GMT
Subject: xml-rpc module question
References: <dmC%5.75166$yR4.2151642@news1.rdc1.tx.home.com> <j7L%5.879$VF3.1528@news.more.net> <%3T%5.76381$yR4.2212181@news1.rdc1.tx.home.com>
Message-ID: <V9T%5.651$Qb7.44185@newsb.telia.net>

Scott Hathaway wrote:
> I made the changes you suggested, but I am still having trouble.

what trouble?

jeff's example (based on your code) isn't syntactically
correct, but you should at least get a SyntaxError.

if that's what happens, try identing the offending line:

import xmlrpcserver
import SocketServer

class myRequestHandler(xmlrpcserver.RequestHandler):

    def CubeIt(self, x):
        return x**3

    def call(self, method, params):
        if method == 'CubeIt':
            return self.CubeIt(params)

server = SocketServer.TCPServer(('', 8000), myRequestHandler)
server.serve_forever()

</F>




From chris at araidesign.com  Thu Dec 21 17:06:23 2000
From: chris at araidesign.com (Chris Arai)
Date: Thu, 21 Dec 2000 14:06:23 -0800
Subject: IDLE: Why doesn't the GO button work in debug?
Message-ID: <3A427EDF.A3AB092E@araidesign.com>

Hi,

I am trying to use the debugger in IDLE.  On a Win2K platform, when
trying to debug the "GO" button doesn't actually run until exit or
breakpoint, it sort of singlesteps.  The "Step" button doesn't always
descend into functions either. Is this a bug or just something I'm doing
incorrectly?

Thanks

--
Chris Arai
Arai Design
chris at araidesign.com
http://www.araidesign.com
707 431 1299
FAX 431 1599

-------------- next part --------------
A non-text attachment was scrubbed...
Name: chris.vcf
Type: text/x-vcard
Size: 224 bytes
Desc: Card for Chris Arai
URL: <http://mail.python.org/pipermail/python-list/attachments/20001221/4ac443e1/attachment.vcf>

From smoriano at sicon.net  Tue Dec 19 05:29:12 2000
From: smoriano at sicon.net (Sergio Boix Moriano)
Date: Tue, 19 Dec 2000 11:29:12 +0100
Subject: CGI on Medusa server
Message-ID: <3A3F3878.4ECF623B@sicon.net>

Hi all,

    I've a medusa http server configured on WinNT, and i want to run a
cgi.

    I'm triyng to do a POST command from an html form, but i get the
response HTTP 400. I suspect that i have to configure the http server
for execute cgi scripts, bu t i don't know how to do it. Any idea?

    I send you the codes of the form and the cgi script

######### FORM.HTML ########

<FORM method="post" action="http://umbriel.pauta.es/cgi/simple.py">
<INPUT TYPE=HIDDEN NAME="key" VALUE="process">
Your name:<BR>
<INPUT TYPE=TEXT NAME="name" value="nombre" size=60>
<BR>
Email: (optional)<BR>
<INPUT TYPE=TEXT NAME="email" value="emilio" size=60>
<BR>
Favorite Color:<BR>
<INPUT TYPE=RADIO NAME="color" VALUE="blue" CHECKED>Blue
<INPUT TYPE=RADIO NAME="color" VALUE="yellow">No, Yellow...
<INPUT TYPE=RADIO NAME="color" VALUE="swallow">What do you
        mean, an African or European swallow?
<P>
Comments:<BR>
<TEXTAREA NAME="comment" value="comentarios" ROWS=8 COLS=60>
</TEXTAREA>
<P>
 <INPUT TYPE="SUBMIT" VALUE="prueba">
</FORM>

######### SIMPLE.PY #####

#! d:/python/python.exe
import re
import cgi

# specify the filename of the template file
TemplateFile = "c:/pysicon/ftp/html/template.html"
# specify the filename of the form to show the user
FormFile = "c:/pysicon/ftp/html/form.html"

# Display  takes one parameter - a string to Display
def Display(Content):
    TemplateHandle = open(TemplateFile, "r")  # open in read only mode
    # read the entire file as a string
    TemplateInput = TemplateHandle.read()
    TemplateHandle.close()                    # close the file
    # this defines an exception string in case our
    # template file is messed up
    BadTemplateException = "There was a problem with the HTML template."

    #SubResult = re.subn("<!-- *** INSERT CONTENT HERE *** -->",
Content,TemplateInput)
    SubResult=[0,1]
    print SubResult
    if SubResult[1] == 0:
        raise BadTemplateException
    print "Content-Type: text/html\n\n"
    print SubResult[0]

### what follows are our two main 'action' functions, one to show the
### form, and another to process it

# this is a really simple function
def DisplayForm():
    FormHandle = open(FormFile, "r")
    FormInput = FormHandle.read()
    FormHandle.close()
    Display(FormInput)

def ProcessForm(form):
    # extract the information from the form in easily digestible format
    try:
        name = form["name"].value
    except:
        # name is required, so output an error if
        # not given and exit script
        Display("You need to at least supply a name. Please go back.")
        raise SystemExit
    try:
        email = form["email"].value
    except:
        email = None
    try:
        color = form["color"].value
    except:
        color = None
    try:
        comment = form["comment"].value
    except:
        comment = None

    Output = ""  # our output buffer, empty at first

    Output = Output + "Hello, "

    if email != None:
        Output = Output + "<A HREF=\"mailto:\"" + email + "">"" + name +
"</A>.<P>"
    else:
        Output = Output + name + ".<P>"

    if color == "swallow":
        Output = Output + "You must be a Monty Python fan.<P>"
    elif color != None:
        Output = Output + "Your favorite color was " + color + "<P>"
    else:
        Output = Output + "You cheated!  You didn't specify a color!<P>"


    if comment != None:
        Output = Output + "In addition, you said:<BR>" + comment + "<P>"

    Display(Output)


###
### Begin actual script
###

### evaluate CGI request
form = cgi.FieldStorage()

### "key" is a hidden form element with an
### action command such as "process"
try:
    key = form["key"].value
except:
    key = None
if key == "process":
    ProcessForm(form)
else:
    DisplayForm()


Thanks in advance for all your responses




From fiona at sitegnome.com  Wed Dec  6 08:00:13 2000
From: fiona at sitegnome.com (Fiona Czuczman)
Date: 6 Dec 2000 13:00:13 -0000
Subject: [FAQTS] Python Knowledge Base Update -- December 6th, 2000
Message-ID: <20001206130013.31329.qmail@synop.com>

Hi All,

The latest entries into http://python.faqts.com

regards,

Fiona Czuczman


## Unanswered Questions ########################################


-------------------------------------------------------------
Idle and Pythonwin correct setup
http://www.faqts.com/knowledge-base/view.phtml/aid/6842
-------------------------------------------------------------
Sven Hohage



-------------------------------------------------------------
How can I setup Idle(starts with an ugly dos-box) and Pythonwin(doesn't start at all)-it is Python 2.0 on Win98
http://www.faqts.com/knowledge-base/view.phtml/aid/6843
-------------------------------------------------------------
Sven Hohage



-------------------------------------------------------------
How to use Dictionary look-up
http://www.faqts.com/knowledge-base/view.phtml/aid/6867
-------------------------------------------------------------
hasan kacmaz



## Edited Entries ##############################################


-------------------------------------------------------------
Are there any bug tracking systems written in Python?
http://www.faqts.com/knowledge-base/view.phtml/aid/1508
-------------------------------------------------------------
Nathan Wallace, Bernhard Reiter, Fiona Czuczman
Alessandro Bottoni

There is a nice project-management, task-tracking (problem-tracking)
system based on ZOPE (and hence on Python) out there:

    http://bits.netizen.com.au/Xen/

XEN seems to be a quite interesting tool, really... 
but at time of writing (Dec 2000) the project seems to be dead.
Check http://sourceforge.net/projects/midnightxen/ for a possible
revival. XEN's license is not GPL compatible, because based on the NPL.


You might also like to check out:

    http://software-carpentry.codesourcery.com/

There are two others, based on Zope (and thus python): 
Open Ticket: http://openticket.point-one.net/
Tracker: http://www.zope.org/Members/klm/TrackerWiki/FrontPage


-------------------------------------------------------------
What IDEs, debuggers, and other development tools are available for Python?
http://www.faqts.com/knowledge-base/view.phtml/aid/6433
-------------------------------------------------------------
Stephan Deibel, Bernhard Reiter


There are a number of options for those that want to go beyond the basic
debugging tools that come with Python itself, or want more than the
commonly used text editors like emacs or vi: 

Commercial
 
1) http://www.wingide.com/wingide -- Wing IDE provides a graphical
debugger with remote debug capability, source module and class browser,
source editor with autocompletion, syntax hilighting and an emacs mode,
and a project manager.

2) http://www.pythonworks.com -- PythonWorks is a rapid application
development environment with editor, debugger, TkInter-based user
interface layout tools, and deployment manager.

3) http://www.activestate.com/Products/VisualPython.html -- VisualPython
is a plug in for Microsoft's development environment, Visual Studio 7.

Free Software (sometimes called Open Source)

5) http://www.python.org/windows/pythonwin/ -- PythonWin is a Python
distribution specifically for Microsoft Windows that includes support
for MFC development, and a graphical environment with source editor,
command shell, and a number of other features.

6) http://www.python.org/idle/ -- IDLE is a TkInter-based integrated
development environment for Python, currently in early release form.

7) http://boa-constructor.sourceforge.net/ -- Boa Constructor, a rapid
application development environment with WXPython-based GUI development
capabilities, currently in early release form.


8) http://www.gnu.org/software/ddd/ --- ddd The Data Display Debugger
can be hooked onto the python debugger and allows nice visualisation
of source code and items on the stack. Note that this is not an
IDE as entries 1-7).






From dsh8290 at rit.edu  Fri Dec 22 17:05:01 2000
From: dsh8290 at rit.edu (D-Man)
Date: Fri, 22 Dec 2000 17:05:01 -0500
Subject: coding convention conversion
In-Reply-To: <20001222122015.A7987@llnl.gov>; from trotts@llnl.gov on Fri, Dec 22, 2000 at 12:20:15PM -0800
References: <20001222122015.A7987@llnl.gov>
Message-ID: <20001222170501.A26150@westsidecnc.rh.rit.edu>

On a second thought,  my previous post will only change the
definitions, not the calls.  You could probably use the same technique
for the calls as well.

Alternatively you could make an explicit regex for each function and
use vim or sed to change it.  Ex:  g/fooBarBaz/s//foo_bar_baz/g   (for
vim)


-D



From MarkH at ActiveState.com  Wed Dec 13 17:22:03 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Wed, 13 Dec 2000 22:22:03 GMT
Subject: Serial IO for Python 2.0???
References: <9166jh$5hh$1@nnrp1.deja.com>
Message-ID: <3A37F54D.7010303@ActiveState.com>

harrc at my-deja.com wrote:

> Does anyone know of a Win32 serial port I/O solution for Python 2.0?
> Both win32comm and the sio module are compiled for Python 1.5.
> 
> 
> Sent via Deja.com
> http://www.deja.com/

There is no source-code compatible solution that I know of.

But in Python 2.0, win32file has full serial port support.  See 
demos\win32comport_demo.py

It shouldn't be too hard to move to this, it is probably a better "long 
term" solution than requiring the closed source 3rd party DLL.

Mark.



From vsbabu at erols.com  Fri Dec 15 07:00:33 2000
From: vsbabu at erols.com (Satheesh Babu)
Date: Fri, 15 Dec 2000 07:00:33 -0500
Subject: stewpid newbie tricks
References: <3A39D4F9.648D8DBA@mediaone.net>
Message-ID: <91d14b$1t4$1@bob.news.rcn.net>

> at 16:00 /Every:m,t,w,th,f,s,su "d:program
> files\python\mine\spider\runSpider.bat"

Perhaps it should be

at 16:00 /Every:m,t,w,th,f,s,su "d:\program
files\python\mine\spider\runSpider.bat"

?

:-)




From urner at alumni.princeton.edu  Thu Dec  7 20:45:25 2000
From: urner at alumni.princeton.edu (Kirby Urner)
Date: Thu, 07 Dec 2000 17:45:25 -0800
Subject: Programming Habits in Python
References: <8visq6$p3o$1@news.nuri.net>
Message-ID: <uhe03tkm41d9l010vn2kjcnlqt7ebqarg0@4ax.com>


I'm one of those who thinks Python is a good first
language.  I wouldn't say it's the only option though.

Although it features dynamic typing, the concept of
types is still there, and that's what a beginning
programmer needs to learn.

Every programming langauge teachers reflexes and 
habits of mind which aren't entirely transferrable.

I like Python because it incorporates a lot of 
ideas from many backgrounds, is cosmopolitan, so
when you move to another language, like Haskell,
or Java, or Scheme, it's not like starting over.
You'll draw on your Python experience and find 
that it helps you.

Kirby

See:  http://www.inetarena.com/~pdx4d/ocn/cp4e.html
for more intro Python, used as a teaching langauge



From leonrom at hypermath.org  Tue Dec  5 20:44:40 2000
From: leonrom at hypermath.org (leonrom)
Date: Tue, 05 Dec 2000 17:44:40 -0800
Subject: update
Message-ID: <3A2D9A08.43F8FBE8@hypermath.org>

I run v.1.5.2, do I have to uninstall it before installing v.2.0


From jerry_spicklemire at my-deja.com  Tue Dec  5 09:17:45 2000
From: jerry_spicklemire at my-deja.com (jerry_spicklemire at my-deja.com)
Date: Tue, 05 Dec 2000 14:17:45 GMT
Subject: easy question on input files in Python
References: <908jvp$dd1$1@nnrp1.deja.com> <908mmu02npb@news2.newsguy.com> <ewHRZOAK8CK6EwgW@ntlworld.com> <90c7jn$3bm$1@nnrp1.deja.com> <90d05f01tub@news1.newsguy.com>
Message-ID: <90ite3$2j0$1@nnrp1.deja.com>

Alex Martelli  wrote:

> What Ian wants is a 'overwritable buffer', so the _same_
> memory area is reused at each read.  But Python file
> objects don't support that.  Nor I believe do array
> objects (their fromFile method appends to the array,
> does not overwrite it).  I think it need to be done as a
> C extension (maybe Numeric already offers it...?).

Good point! Object arrays, not limited to numeric objects,
are one of the coolest features of Numeric.
I've no clue about over-writable-ness though.

Thanks for the reply!
Jerry S.


Sent via Deja.com http://www.deja.com/
Before you buy.


From nospam at no.spam  Tue Dec 19 11:38:53 2000
From: nospam at no.spam (Carsten Geckeler)
Date: Tue, 19 Dec 2000 17:38:53 +0100 (CET)
Subject: while true: !!!
In-Reply-To: <91noj60178k@news2.newsguy.com>
Message-ID: <Pine.LNX.4.21.0012191726320.30860-100000@justitia.jura.uni-tuebingen.de>

On Tue, 19 Dec 2000, Alex Martelli wrote:

> "Carsten Geckeler" <nospam at no.spam> wrote in message
[snip]
> > Not fair!  C is a lowlevel language (relatively to interpreter languages
> > as Python etc.).  In C you can say
> 
> C is a low-level language, Java a bit less so, Icon even less, and
> Haskell could hardly be accused of being very low-level -- they all
> distinguish a character from a stringth of length 1 as fundamentally
> different datatypes, as do many, indeed most, other languages.
> 
> My claim is that Python is simpler, clearer, and MORE effective,
> for NOT trying to draw this level of distinction.

Right, it's more effective with regard to programmers' work, but
definitely less effective according to speed.

> > s[3] = char
> > Do you know anything faster than this?  It just writes the byte char into
> > the memory space pointer str + 3.  In most cases strings do consist of
> > chains of bytes in memory.
> 
> Not sure what you mean by 'chain' here -- Haskell is the only one

Is array better?

> of the languages I've listed that models strings as _lists_ (aka
> linked lists, aka chains) of characters, most others (excluding
> some Lisp dialects & variants) normally preferring to model them
> as _arrays_ (aka blocks, &c) of characters (what _Python_ calls
> 'lists' are actually flexible arrays, which further confuses the
> terminology).
[snip]
> So it's clearly not a sacrifice that language designers endure
> just because it will enhance their performance... some must
> believe it's a GOOD distinction to draw.  Python (and Perl
> and Tcl, just to name a couple others) disagree -- and I
> think they're right.
> 
> > Please, don't be so mean to C.  It's such a nice programming language!
> > Remember, Python itself is programmed in C.  ;)
> 
> I program in C when I must (though I'd much rather take a
> step up to C++, which I can in most situations but *NOT* when
> I want my Python extensions to be taken seriously by the
> Python community, alas...).  That doesn't mean I have to
> _like_ it:-).

Right.  Take what programming language you like.  But the fact that Python
and kernels are programmed in C shows, that you always use the language
that fits your problem.  I love programming in Python, I also like Perl,
and I also program in C.  And probably I will also program in Fortran
during my thesis (although I HATE Fortran).  But that's it: If you need
pure speed, Fortran is just a better choice than rare Python.  If you need
high-level construct whithout -- almost -- the work of type declaration,
Python is the better choice.

> > But to your point:  If you are sure that your string is always shorter
> > than 79 characters, it'n no faster way than doing it as C.  No overhead
> > due to range checking and reallocating.  But of course, from the everyday
> > user's point, interpreter languages as Python or Perl do a nicer job.  But
> > every language has it's field of appliance.  Just imagine programming a OS
> > kernel in Python... ;)
> 
> I can at least imagine programming it in C++ (BeOS) or even, if need
> be, in Objective C (Next) -- in both cases, with a C-like subset
> available for the very lowest portions of the job, but a semi-decent
> higher-level approach for most of it.

C++ with the STL is very nice. ;)

Cheers, Carsten
-- 
Carsten Geckeler:  carsten dot geckeler at gmx dot de
To get proper email-address replace `dot' and `at' by the corresponding symbols.




From aleaxit at yahoo.com  Mon Dec 11 14:55:19 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 11 Dec 2000 20:55:19 +0100
Subject: Python2 distutils question: how to best "autoconfig"...?
References: <LNBBLJKPBEHFEDALKOLCOEFFIDAA.tim.one@home.com>
Message-ID: <001701c063ac$5163e5c0$4b8d1497@martelli>

"Tim Peters" <tim.one at home.com> gives me excellent advice...:

> > The specific issue is "how many bits of significance
> > does a Python float have".  (Or is there somewhere in
> > the Python headers a #define for this that I missed...?)
>
> You can use DBL_MANT_DIG from the platform's float.h.  Try hard to ignore
> FLT_RADIX <wink>.

*Wak!*  Shame on me.  How *COULD* I forget we can
now use Standard ISO C...!-(

> *Python*, though, no optimizer known to human-kind is smart enough to keep
> your temps in Pentium extended registers across being stored into PyFloat
> objects.  For example, I use this:

Good point (and, another brainstorm on my part!-).  But I think
DBL_MANT_DIG will do me.


> Alternative:  by hook or by crook, cut your double x back to storage
> precision (see above).  Say two successive convergents are i/j and i'/j'.
> Stop, and return i/j, when abs(double(i)/j - x) <= abs(double(i')/j' - x),
> i.e. stop when the convergents stop improving in double arithmetic.

That would rather frustrate the general purpose of the Stern-Brocot
thingy (which is also used to convert to mpq from unlimited-precision
mpf's).  But maybe it could be worth (speedier, &c) having two copies
of the algorithm in the module -- one implemented in, and used for,
Python floats, and a separate one for mpf's.  gmpy's still pre-alpha, so
I'm not pushing on performance all that much, yet, but I'll keep it in
mind for the future.


Alex



__________________________________________________
Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.
http://im.yahoo.com



From nospam at no.spam  Tue Dec 12 08:57:58 2000
From: nospam at no.spam (Carsten Geckeler)
Date: Tue, 12 Dec 2000 14:57:58 +0100 (MET)
Subject: global __dict__
In-Reply-To: <LNBBLJKPBEHFEDALKOLCIEFNIDAA.tim.one@home.com>
Message-ID: <Pine.LNX.4.10.10012121455230.468-100000@erde.kosmos.all>

Thanks,

That's exactly what I was looking for.  The only point that bothers me a
little bit with __main__.__dict__ is, that I have to import __main__
manually.  Seems more logical to me to have it included from the start.

Cheers, Carsten

On Mon, 11 Dec 2000, Tim Peters wrote:

> [posted & mailed]
> 
> [Carsten Geckeler]
> > I was wondering if there is a __dict__ dictionary for the global
> > namespace as for classes and modules.
> 
> Your terminology is a bit off:  you're *always* executing in a module in
> Python.  "global namespace" has no other meaning than the namespace of the
> module you're currently executing in.
> > I was trying the following
> >
> > >>> var = 1
> > >>> __dict__
> >
> > and expected that __dict__ would contain something like {..., 'var': 1},
> > but there is no global __dict__.  Where can i find the global
> > dictionary?
> 
> You need the module's __dict__.  In an interactive shell, you're running in
> a module Python creates for you, named "__main__".  So:
> 
> >>> var = 1
> >>> import __main__
> >>> __main__.__dict__
> {'__main__': <module '__main__' (built-in)>,
>  '__doc__': None,
>  'var': 1, # <------------------------------------- here's your vrbl "var"
>  '__name__': '__main__',
>  '__builtins__': <module '__builtin__' (built-in)>}
> >>>
> 
> Somewhat easier is to use the globals() builtin function (which returns the
> __dict__ for the module you're currently in):
> 
> >>> __main__.__dict__ is globals()
> 1
> >>>
> 
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 

-- 
Carsten Geckeler:  carsten dot geckeler at gmx dot de
To get proper email-address replace `dot' and `at' by the corresponding symbols.




From dsh8290 at rit.edu  Mon Dec 18 08:52:29 2000
From: dsh8290 at rit.edu (D-Man)
Date: Mon, 18 Dec 2000 08:52:29 -0500
Subject: while true: !!!
In-Reply-To: <v04220801b663769e1800@[192.168.11.33]>; from Blair.Lowe@compeng.net on Mon, Dec 18, 2000 at 01:19:51AM -0700
References: <B43D149A9AB2D411971300B0D03D7E8B0FBCBB@natasha.auslabs.avaya.com> <v04220801b663769e1800@[192.168.11.33]>
Message-ID: <20001218085229.A10917@westsidecnc.rh.rit.edu>

> It is cheap to define true and false once,

I think that some people were confused with this statement.  He meant
that 'true' and 'false' would be constants like in C++ and Java.  It
would be illegal to try 

true = false


Thus while true: is equal to while 1: except that a boolean, not an
integer is used for the test.  I, for one, would like to abolish the
concept that an integer can denote true or false.  How many functions
return 0 for success?  (just take a look at the standard C library).

-D



From clarence at netlojix.com  Fri Dec 29 16:40:22 2000
From: clarence at netlojix.com (Clarence Gardner)
Date: Fri, 29 Dec 2000 13:40:22 -0800
Subject: lazy eval()
Message-ID: <978126073.842112312@news.silcom.com>

Thanks to /F and Emile van Sebille for their suggestions.  /F's solution
is not only simple and elegant, it appears to have minimal overhead
relative to an eval with known names.

Emile, I didn't mean to be cryptic, I was just omitting irrelevancies.
Here's a sample of the kind of expressions that are currently in use:
    nDollars['12h', 'R'].nEQ(0)
This retrieves the values from a datastream identified by 'nDollars' that
were stored within the last twelve hours, creates a list of the differences
between successive values, then counts the number of those differences
that are equal to zero.

The problem is, these expressions are entered into a database by end users,
and I currently require them to store also the definition of the datastreams
that the expression uses.  Yucky; hence, I want to do the binding of the
name to the datastream when the expression is evaluated.  That binding has
to be deferred because
  1) I don't know the datastreams that the expression uses, and
  2) I'm generalizing the thing a bit, and the binding will be more
     complicated, possibly involving several SQL queries.

[Sorry, this isn't a followup; the POS newsreader that comes with KDE
crashed the first time I tried to post this, and lost the whole ng for
me :( ]

Clarence Gardner
clarence at netlojix.com



From erno-news at erno.iki.fi  Wed Dec 27 10:14:11 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 27 Dec 2000 17:14:11 +0200
Subject: Python and Napster
References: <92ap0g12b6r@news1.newsguy.com> <sx226.18970$RC1.511898@newsread1.prod.itd.earthlink.net> <20001227164900.DF7E6A84F@darjeeling.zadka.site.co.il> <mailman.977908882.12582.python-list@python.org>
Message-ID: <kuae9hsxq4.fsf@lasipalatsi.fi>

Alex Martelli <aleaxit at yahoo.com> writes:

| HTTP (and the Python tools around it) is great for file transfer (and
| suchlike) but it's not oriented to peer-to-peer at all.

<rant>with the present rate of destruction of the end-to-end internet, http
is rapidly becoming the only protocol that can traverse through the
various nats and firewalls and god knows what.</rant>

besides, it's not that bad for file sharing imho. just add
a couple of methos, or maybe use something like webdav as the basis.

  -- erno


From kragen at dnaco.net  Sun Dec 31 00:00:17 2000
From: kragen at dnaco.net (Kragen Sitaker)
Date: Sun, 31 Dec 2000 05:00:17 GMT
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <05C6FUhLDNUU-pn2-tuTnzOCokxem@bill> <Sgw36.208288$U46.6385684@news1.sttls1.wa.home.com> <92m9pm$d19$1@nnrp1.deja.com>
Message-ID: <B3z36.122$9s1.18804@e420r-atl2.usenetserver.com>

In article <92m9pm$d19$1 at nnrp1.deja.com>,  <henry_crun3583 at my-deja.com> wrote:
>It does seem to me that "do: while" represents a critically different
>algorithm (ie act, then test information that becomes available as a
>consequence of the action) from "while: ..." (test pre-existing
>information for need or possibility of action, then act)

Yes.  See Knuth, "Structured Programming with Go To Statements", and
examine his proposed loop-while-end construct (which has since been
implemented in TeX and, IIRC, Ada).

The only time I wish I wasn't subject to Python's indentation rules is
when I can't outdent my "if finished(): break" statements like I do in
C.
-- 
<kragen at pobox.com>       Kragen Sitaker     <http://www.pobox.com/~kragen/>
Perilous to all of us are the devices of an art deeper than we possess
ourselves.
       -- Gandalf the White [J.R.R. Tolkien, "The Two Towers", Bk 3, Ch. XI]




From tim.one at home.com  Mon Dec  4 19:45:27 2000
From: tim.one at home.com (Tim Peters)
Date: Mon, 4 Dec 2000 19:45:27 -0500
Subject: int multiplication overflow and the future of python
In-Reply-To: <ghaelzoc9lp.fsf@synopsys.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEGIICAA.tim.one@home.com>

[posted & mailed]

[Ram Bhamidipaty]
> On red hat 7.0 with python 1.6 I am running into overflow problems.
>
> From looking at the bug database at sourceforge I can see
> that other people have run into this, and it looks like
> there is a fix....but I have not been able to find it. Is there
> a patch that I should apply to upgrade my 1.6 source to the
> latest version?

It's not a Python problem so there's not a Python patch.  See "LONG_BIT
Error" at

http://www.python.org/cgi-bin/moinmoin/FrequentlyAskedQuestions

for a workaround.  I have heard (but don't know for a fact) that Red Hat is
shipping their own critical patches to 7.0 that fix this problem (among
others -- they shipped a development snapshot of gcc with 7.0, which they
should not have done).

> This leads to a larger question: Which version is being maintained?
> Is active developement taking place for both 1.6 and 2.0?

No, just for 2.0.

> I've been using 1.6 because the features in 2.0 seem like they
> involve bigger changes to the code base and I thought I should
> wait until the initial bugs are worked out.

Well, that's ironic:  the bugs in new 1.6 features were worked out in 2.0
<wink>.




From nas at arctrix.com  Fri Dec  8 06:07:45 2000
From: nas at arctrix.com (Neil Schemenauer)
Date: Fri, 8 Dec 2000 03:07:45 -0800
Subject: Tuple - List - Whats the difference?
In-Reply-To: <200012080813.eB88DWu01392@ZOOstation.cc>; from rob@ZOOstation.cc on Thu, Dec 07, 2000 at 08:13:31PM -1200
References: <Pine.LNX.4.21.0012071349420.22360-100000@relax.dk> <200012080813.eB88DWu01392@ZOOstation.cc>
Message-ID: <20001208030745.B1931@glacier.fnational.com>

On Thu, Dec 07, 2000 at 08:13:31PM -1200, Rob Brown-Bayliss wrote:
> Hi, brand new to python and I am trying to understand why there are
> tuples and lists.  
> 
> They both look to serve the same purpose.  Can some one enlighten me?

Lists are mutable (you can change them after creating them) while
tuples are not.

  Neil



From erno-news at erno.iki.fi  Thu Dec 21 20:23:32 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 22 Dec 2000 03:23:32 +0200
Subject: PyOpenGL and Nvidia GL?
References: <tmi%5.60958$df5.2143169@news1.crdva1.bc.home.com> <87k88xkmik.fsf@optonline.net> <91tga9$gav$1@nnrp1.deja.com>
Message-ID: <ku1yv1l04r.fsf@lasipalatsi.fi>

[followups set]

|| you will see why i will never buy another video card from NVidia so
|| long as their drivers are closed source.

 | Unfortunately with the recent demise of 3dfx you may have little
 | option.  I don't know of any other manufacturer who has a decent
 | open source OpenGL driver.  Can you think of one? 

matrox ?

  -- erno


From DanielK at aracnet.com  Wed Dec 20 13:01:44 2000
From: DanielK at aracnet.com (Daniel Klein)
Date: Wed, 20 Dec 2000 10:01:44 -0800
Subject: sys.stdout.write()
References: <r3M%5.159$LU6.61501@typhoon.aracnet.com> <91o4ac01iqj@news2.newsguy.com>
Message-ID: <9p606.193$LU6.73144@typhoon.aracnet.com>

"Alex Martelli" <aleaxit at yahoo.com> wrote in message
news:91o4ac01iqj at news2.newsguy.com...
> "Daniel Klein" <DanielK at aracnet.com> wrote in message
> > As a side (but related) issue, the current online reference material
does
> > not always indicate all of the exceptions thrown by a method or
function.
> > What is the best way to determine this when writing code?
>
> Now THAT is something I think you can't get by reflection -- the
> docs WOULD have to provide it, I think.

Sorry, I guess I didn't ask that question correctly. :^( I'd really like to
know if there _is_ any reference material for this. Or can I find this in
the Python source code (I'm not a C programmer so if many of the exceptions
are thrown from C libraries, I'll be sorta lost)?

Daniel Klein




From jwbnews at scandaroon.com  Mon Dec 25 13:53:26 2000
From: jwbnews at scandaroon.com (John W. Baxter)
Date: Mon, 25 Dec 2000 10:53:26 -0800
Subject: os.exec?
References: <LXw16.8561$KY1.23035@news1.rivrw1.nsw.optushome.com.au> <mailman.977727428.19738.python-list@python.org>
Message-ID: <jwbnews-82EF1A.10532625122000@news.olympus.net>

In article <mailman.977727428.19738.python-list at python.org>, Moshe 
Zadka <moshez at zadka.site.co.il> wrote:

> On Mon, 25 Dec 2000, "deadmeat" <root@[127.0.0.1]> wrote:
> 
> > How to I execute an external program without replacing the Python
> > interpreter? I want control to return to Python when the external 
> > program is
> > completed. I still want stdio to go to the console not to a pipe if
> > possible.
> 
> Quite possibly, you'd want os.system, and os.popen.
> Or, you might want to work on a lower level with os.fork/os.exec/os.open.

Or, moving to a higher level, there is the commands module, which wraps 
those things mentioned above.  I tend to be lazy and use that.

   --John

-- 
John W. Baxter   Port Ludlow, WA USA  jwbnews at scandaroon.com


From paulp at ActiveState.com  Wed Dec 13 10:29:33 2000
From: paulp at ActiveState.com (Paul Prescod)
Date: Wed, 13 Dec 2000 07:29:33 -0800
Subject: Self-documenting Python??
References: <x881yvcfodl.fsf@adglinux1.hns.com>
Message-ID: <3A3795DD.B709886@ActiveState.com>

nbecker at fred.net wrote:
> 
> Has there been any proposals to integrate a documentation system with
> python?  I feel this is one feature that is sorely missing.

Good question. Here's the answer:

http://python.sourceforge.net/peps/pep-0233.html
http://python.sourceforge.net/peps/pep-0216.html
http://python.sourceforge.net/peps/pep-0224.html

 Paul Prescod



From dealfaro at eecs.berkeley.edu  Sat Dec 16 20:56:26 2000
From: dealfaro at eecs.berkeley.edu (Luca de Alfaro)
Date: Sat, 16 Dec 2000 17:56:26 -0800
Subject: Alarms in threads - how to get/mimic?
In-Reply-To: <ltbsucbbhb.fsf@asfast.com>
Message-ID: <Pine.LNX.4.21.0012161749380.7913-100000@traal.eecs.berkeley.edu>

Dear Lloyd, 

[in CC to newsgroup, in case this is of general interest]
thank you very much for your reply!  
I got the module, and read its documentation. 
The only point that is not clear to me is the following: 
In the documentation, Timothy states that one can just use
timeoutsocket as a replacement for socket, and all other modules, such
as httplib, will then have timeout capability. 

What is not clear to me is the following.  When there is a timeout,
the exception Timeout is not caught by httplib.  Hence, if the thread
that called httplib just continues its job giving up on the
connection, httplib will not have a chance to do any clean-up, such as
for example releasing/closing the socket. 

Hence, it seems to me that in order to abort a web page load cleanly
upon timeout, I also have to modify the httplib module to handle the
exception Timeout, and do clean-up work if that happens. 

Is my understanding correct? 

Luca

On 16 Dec 2000, Lloyd Zusman wrote:

> Perhaps Timothy O'Malley's `timeoutsocket' class will help you.  It
> allows you to set timeouts for selected socket operations, such as the
> httplib connect calls you're referring to.  You can get it here ...
> 
>   http://www.timo-tasi.org/python/
[...]
> 
> Using this module, you can cause each httplib connection statement
> which times out to throw an exception, and each thread can then
> intelligently respond to its own exception.
> 
> I've been using this module for a similar purpose, and I'm quite
> happy with it.




From sill at localhost.kitenet.net  Wed Dec 13 21:52:28 2000
From: sill at localhost.kitenet.net (Oldayz)
Date: Thu, 14 Dec 2000 02:52:28 GMT
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <915jcf0c6o@news2.newsguy.com> <9162p3$7t2$1@animus.fel.iae.nl>
Message-ID: <MBWZ5.334$h67.37748@newsread1.prod.itd.earthlink.net>

On 12 Dec 2000 21:45:23 +0100, Carel Fellinger <cfelling at iae.nl> wrote:
>Alex Martelli <aleaxit at yahoo.com> wrote:
>
>> Firmitas, Utilitas, Venustas -- in *THIS* order, as Vitruvius
>
>No no, Beauty is the one and true guiding principle, Beauty and Simplicity.
>The trouble being that what one perceives as beauty has ever so much to do
>with ones upbringing. And elas, there are very few with an eye for Beauty.
>
>A lot of us were raised in a time were it was frowned upon to use goto's
>or multi-exit loops.  We were made believe that for-loops, while's,
>repeat-untils were fundamentally different beasts, whilest they were
>but minor variants of that (B|b)eatiful basic loop construct that was
>attributed earlier on in this thread to Knuth himself. 
>
>> "Infinite" loops are hardly ever appropriate -- loops normally
>> do terminate (although it may be appropriate to model, with
>> an "infinite" loop, one that is intended to run until the
>> machine itself is turned off, for example).  However, loops
>> that are meant to "cycle until told othewise" (are terminated
>> by an exception, a signal, or a condition best computed inside
>> the body of the loop itself) are extremely common.  If your
>> aesthetic sense interferes with their effective and idiomatic
>> usage, then it is most likely leading you astray.
>
>He is just hindered by his upbringing:), and has yet to appreciate
>the simplicity you get using that raw and basic loop construct in
>those rare cases where it is really needed.
>
>On the other hand, I still detest the way this construct is spelled in
>Python, would have prefered (I think) something like:
>
>loop: suite
>break <condition>: suite
>...
>
>but then, this wouldn't be general enough. Thus the only thing I would
>change would be to introduce "while:" as a shortcut for while 1:", and
>I would be thrilled, aesthetic wise that is.

I thought about the same thing, but then I thought something like
forever: would be even more lucid, although it sounds a little
silly.. 


-- 

	Andrei


From greg at cosc.canterbury.ac.nz  Wed Dec 13 20:25:58 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Thu, 14 Dec 2000 14:25:58 +1300
Subject: [Q] Are Exceptions used that much in practice?
References: <3A35D546.A4731B5B@enteract.com> <3A36596A.89EEA210@alcyone.com> <3A36C6E3.26181E3F@enteract.com>
Message-ID: <3A3821A6.8107C543@cosc.canterbury.ac.nz>

Jerome Mrozak wrote:
> 
> An example of compulsion is java.lang.Integer.parseInt(), which throws a
> NumberFormatException you *must* deal with or the program won't
> compile.

Ah, I see what you mean. No, there's never a case where you 
really *have* to catch an exception. You always have the 
option of ignoring exceptions; an exception which isn't 
caught anywhere will simply cause the program to abort 
with a traceback.

Sometimes you will want to catch exceptions, but those cases
are very rare, in my experience. Most of the time you can
write code without thinking about exceptions at all.

> I'm aware of examples of Python such as (faking it here):
> 
> while 1:
>   try:
>     input = inputFileObject.readline()
>   catch EOF:
>     break
>   else:
>     process(input)

Python doesn't report end-of-file by means of an exception;
it returns an empty string.

It does report I/O errors such as failing to open a file
with an exception, which is one of the few cases that you
will probably want to catch. You don't necessarily have
to put a try-except around every file open, though, and
that's probably not the best thing to do. A single
try-except somewhere high up in your program's logic
to catch any failures and report them to the user is
usually sufficient.

Hope that helps,

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


From skip at mojam.com  Wed Dec 27 14:02:26 2000
From: skip at mojam.com (Skip Montanaro)
Date: Wed, 27 Dec 2000 13:02:26 -0600 (CST)
Subject: ANNOUNCE: soundex.py
In-Reply-To: <slrn94793q.1qk.amk@207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com>
References: <mailman.977338109.32467.python-list@python.org>
	<3A4119C3.2E5E54E1@uswest.net>
	<mailman.977501169.22862.python-list@python.org>
	<slrn94787k.1q7.amk@207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com>
	<3A43A006.F2CCC36@uswest.net>
	<slrn94793q.1qk.amk@207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com>
Message-ID: <14922.15554.198089.988769@beluga.mojam.com>

    AMK> On Fri, 22 Dec 2000 11:40:06 -0700, Bjorn Pettersen <pbjorn at uswest.net> wrote:
    >> If we could get the soundex.py as part of the standard library, this
    >> would definitely get my vote.

    AMK> That's not what I'm suggesting; it was there once and was removed as
    AMK> being of marginal usefulness to most people. 

Which is why there's now a soundex.py.  I don't believe anyone expected it
to wind up in the standard distribution (I certainly didn't).

    AMK> Modules/soundex.c is still around as a relic of this, though it's
    AMK> not compiled by default.  I'm suggesting it's now safe to delete
    AMK> soundex.c.

I think there's a PEP that discusses deprecating/deleting content from the
distribution.  Deleting soundex.c should follow those guidelines in my
opinion.

Skip











From claird at starbase.neosoft.com  Wed Dec 27 06:00:19 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 27 Dec 2000 05:00:19 -0600
Subject: Chronology of the founding fathers (was: Python vs. PHP (& Java?))
References: <LOBBJMHOIFOLFIPFANHJMEMHCFAA.beno@TheWebsons.com> <mailman.977912052.14706.python-list@python.org>
Message-ID: <54A5DC0883A6A83D.617801CCE49B11D9.FA1AC6E3617A8BBC@lp.airnews.net>

In article <mailman.977912052.14706.python-list at python.org>,
Thomas Wouters  <thomas at xs4all.net> wrote:
			.
			.
			.
>As for why Python isn't PHP, well, because it isn't. Python is a fair bit
>older than PHP, and simply wasn't intended as an 'html-embedded-language'.
>It was intended as a proper, easy to learn, easy to read, efficient (from
>the programmers view) language, and not as a quick-and-dirty-hack to allow
>code inbetween HTML :) The WWW and HTML did exist, but were only just
>starting, when Python emerged.
			.
			.
			.
1989:  Tim Berners-Lee writes hypertext proposal that
       will become the WWW.  "Universal Document Ident-
       ifier" is one keyword.
December 1989:  Guido van Rossum designs Python.
1990:  First Python instance executes (on Macintosh!).
       Guido distributes early versions.
1990:  TBL writes first version of "WorlDwidEweb" point-
       and-click hypertext editor.
1991:  TBL releases Web server and WorlDwidEweb to hyper-
       text community.  He publishes specs for HTTP,
       HTML, and (slightly disguised) URIs.
-- 

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


From aleaxit at yahoo.com  Mon Dec  4 10:58:18 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 4 Dec 2000 16:58:18 +0100
Subject: "inherited" keyword in Python?
References: <3A2B9EFA.3A64C13B@schlund.de>
Message-ID: <90gevs02uu1@news2.newsguy.com>

"Carsten Gaebler" <cg at schlund.de> wrote in message
news:3A2B9EFA.3A64C13B at schlund.de...
> Hi there!
>
> Does Python know something like the "inherited" keyword from Borland
> Pascal? In Pascal it is used to call a method of a class's superclass,
> i.e. instead of saying "superclass.method()" you say "inherited
> method()". Can be quite helpful if you change the name of the superclass
> or just forgot it. :-)

Python has multiple inheritance, so there is (in general) no single
'superclass' that you could use that way.


When this was last discussed in this forum (regarding a different
language for comparison -- Java or Smalltalk, I believe -- anyway,
some other single-inheritance language which uses 'super' as the
keyword), the solution that eventually emerged was (quite similar
to what Stroustrup proposes in C++ in "Design and Evolution"):

class SomeClass(ABaseClass,AnotherOne,YetAnother):
    __super = AnotherOne
    # etc, etc
    def somemethod(self, foo, bar):
        print "before"
        __super.somemethod(self, foo, bar)
        print "after"

The key idea is that a class must explicitly choose one of its
superclasses as 'the' super (presumably, in a derived-from-
single-inheritance style, the others are just mixins, rather
than having any deep semantic significance), and bind its
classobject to the __super class-variable (the two leading
underlines, without _trailing_ underlines, are important: they
tell Python that this is an identifier specific to this one
class, and avoid clashes with other classes in the inheritance
graph that may be using the same idea).


The choice is clearly simplest for a class which has but one
superclass, but it still needs to be explicit:

class SomeOtherClass(AnotherOne):
    __super = AnotherOne
    # etc, etc
    def somemethod(self, foo, bar):
        print "before"
        self.__super.somemethod(self, foo, bar)
        print "after"

Metaprogramming magick was deemed inappropriate here (as I
read the group consensus).  One *might* have a global function
called super(), which tried to gauge the appropriate class
object to use (in single-inheritance cases, only, of course),
but it would have to be very black magic indeed (there seems
to be no 'normal' metaprogramming way to discover what class
the calling-method belongs to) as well as liable to break
under all sorts of circumstances (remember that there is no
guarantee that a method is being used within the class in
which it was defined, nor indeed that it was at all defined
as a method rather than originally defined as a free function
and later also assigned as a method), not to speak of multiple
inheritance.

The solve-it-by-convention-if-you-feel-this-need approach was
thus considered preferable in Python (basically by the same
reasoning through which Stroustrup and the standard committee
settled on it in the context of C++).


Alex





From kirschh at lionbioscience.com  Wed Dec 13 03:31:52 2000
From: kirschh at lionbioscience.com (Harald Kirsch)
Date: 13 Dec 2000 09:31:52 +0100
Subject: running a pipe of commands from python
References: <yv24s08pz0c.fsf@lionsp093.lion-ag.de> <8ZFZ5.3701$Kd1.429028@newsb.telia.net>
Message-ID: <yv2puiwoh6v.fsf@lionsp093.lion-ag.de>

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

> Harald Kirsch wrote:
> > In Tcl I just write
> >
> >    exec doThis | doThat |toMore | finishUp >output <input 2>errors
> >
> > In the Python reference manual I find os.exec(), os.fork(), popen2()
> > etc, but none of these is as easy to use as Tcl's exec. What is
> > Python's equivalent (and I am sure there must be somehting)?
> 
> you don't say what output, input and errors are, so it's
> not obvious what you're looking after.
> 
> if they're filenames, use:
> 
>     os.system("doThis | doThat >output ...")

Buuuh. I don't like the overhead of the call to the shell. In addition
the shell has to parse the given command line which means that I
really have to take care of escaping special characters and the like.

> 
> if they're file handles, use os.popen3:
> 
>     input, output, errors = os.popen3("doThis | doThat")

This also seems to call a shell in between.

> for a more general (but unix-only) solution, see:
> http://www.python.org/doc/current/lib/module-pipes.html

This also uses /bin/sh. And all this although python has all the
necessary ingredients like fork/exec/wait/signal to do the plumbing
itself. Strange.

I think I'll stick to my own hack, the doc of which currently reads
like this:

def execpipe(cmds):
  """
  runs the commands given in the list cmds in a pipe with their stdout
  and stdin plumbed together. Every element of cmds must itself be a
  list giving the name of the command as the first element and the
  rest of the list being its arguments.

  If any of the elements of cmds is a pair, then the first must be the
  command to run and the 2nd must be a file descriptor for output
  which is then connected to the commands stderr. By default, the
  stderr is the same as that of the calling script.
  """

The implementation seems to work for most cases. Not all signals which
cause the parent to die are distributed properly to the childs.

Any comments?

  Harald Kirsch

-- 
----------------+------------------------------------------------------
Harald Kirsch   | kirschh at lionbioscience.com | "How old is the epsilon?"
LION Bioscience | +49 6221 4038 172          |        -- Paul Erd?s
       *** Please do not send me copies of your posts. ***


From katz at Glue.umd.edu  Wed Dec 13 13:16:24 2000
From: katz at Glue.umd.edu (Roy Katz)
Date: Wed, 13 Dec 2000 13:16:24 -0500
Subject: [OT] Re: Dijkstra's Shortest Path algorithm
In-Reply-To: <3A37B48F.FE8F112@uswest.net>
References: <Pine.GSO.4.21.0012090042370.15110-100000@y.glue.umd.edu> <dvrd3t40u9l7atgdro59m5ttg7dsfv7c4i@4ax.com> <Pine.GSO.4.21.0012122249480.19682-100000@y.glue.umd.edu> <3A37B48F.FE8F112@uswest.net>
Message-ID: <Pine.GSO.4.21.0012131314040.761-100000@y.glue.umd.edu>

It was mandated in the assignment that we use cin/cout :)
I considered rewriting them, as they were pretty slow. 
What proved to be the biggest time saver was caching everything: objects,
array indeces, etc.  Given the amount of time I spent optimizing this
project, I have a very profound respect for C now. 


Roey
who would have used only getc(stdin) if it was allowed



On Wed, 13 Dec 2000, Bjorn Pettersen wrote:

> Yes, unfortunately, if you want fast IO you shouldn't use the iostreams
> library (the new standards compliant one is even slower than the old
> iostreams.h version since it has to go through locale *sigh*).
> 
> fopen/fread etc. are still available when you need speed (and usually much
> easier on the eye also <wink>).
> 
> -- bjorn
> 
> 



From jnix at rioverde.cafeconleche.de  Wed Dec  6 12:05:44 2000
From: jnix at rioverde.cafeconleche.de (Johannes Nix)
Date: 06 Dec 2000 18:05:44 +0100
Subject: Q: How to install my own library on linux ?
References: <Pine.GSO.4.10.10011271140450.18120-100000@ux12.cso.uiuc.edu>
Message-ID: <90lr9i$h54@news.Informatik.Uni-Oldenburg.DE>

hwan-jo yu <hwanjoyu at students.uiuc.edu> writes:

> Hi,
> 
> When we do "import something", it checks the files in sys.path list.
> I would like to add my own library path in the sys.path list, but I don't
> want to do sys.path.append() every source.
> Is there any way to modify data of the original sys.path of python2.0 on
> linux, so a source doesn't have to do sys.path append() to import my
> library file ?


For the simple cases it should be sufficient to set the PYTHONPATH
environment variable. For example you could do

export PYTHONPATH=/usr/local/lib/python/site-packages:$HOME/lib/python

in your .profile file.

Of course this can be made processor-specific as well. Clearly, this
does not work if you have conflicting names for modules.


Johannes Nix



From aahz at panix.com  Thu Dec  7 02:05:36 2000
From: aahz at panix.com (Aahz Maruch)
Date: 6 Dec 2000 23:05:36 -0800
Subject: OT: _Fifth Business_ (was Re: Python Origins)
References: <90k3pq$50p$1@nnrp1.deja.com> <slrn92tu7t.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com> <90n1b0$4jr$1@panix3.panix.com> <5hGX5.68$E3.1197@uchinews>
Message-ID: <90ncs0$mt5$1@panix2.panix.com>

In article <5hGX5.68$E3.1197 at uchinews>,
Ben Wolfson <wolfson at midway.uchicago.edu> wrote:
>In article <90n1b0$4jr$1 at panix3.panix.com>, Aahz Maruch <aahz at panix.com> wrote:
>>
>>What's _Fifth Business_?  (I did a quick search on Google and Amazon and
>>didn't find anything related to you.)
>
>_Fifth Business_ is the first book in the Deptford trilogy by Robertson
>Davies, a Canadian and a man with a very impressive beard (the other
>books being _Manticore_ and _World of Wonders_).

<confused look>  But what does that have to do with Andrew?
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

What if there were no rhetorical questions?  --Aahz


From peter at engcorp.com  Thu Dec  7 11:56:56 2000
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 07 Dec 2000 11:56:56 -0500
Subject: parallel port access?
References: <3A268378.4DA1B831@sympatico.ca> <3A272881.6F444750@engcorp.com> <3A2D6761.814ABE8E@bioeng.ucsd.edu> <slrn92su1q.12g.ssthapa@localhost.localdomain> <udhf4gk079.fsf@box.home.de>
Message-ID: <3A2FC158.B0AD2CD2@engcorp.com>

Jochen K?pper wrote:
> 
> Well, on many unices there is the common idea of a printer beeing a
> file:-) This means you have a driver sitting behind it, and simply
> writing to that file sends your data to the port.
> Direct access to the ports in general requires kernel-level priorities
> or root-access and going through ioperm or such...
> 
> So I would say you'll need a generic wrapper function for sending data
> to the OS, that's it. On Unix I'd say all the problem reduces to find
> out wether there is a printer-file and what it is called.

I'm guessing that many of the people interested in this kind of thing
(direct control of parallel port) want to use it to directly twiddle the
bits, in order to gain quick-n-dirty control of some external
equipment.  I doubt treating the parallel port as a file would provide
the necessary control other than maybe leaving the data bits set to the
last bit pattern written.  The handshake lines would be hopeless, and
bidirectional control (reading individual bits) would likely be lost as
well.


From fredrik at effbot.org  Fri Dec 29 18:48:12 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 29 Dec 2000 23:48:12 GMT
Subject: Is 100,000 entries big for a dictionary?
References: <92iug4$vqb$1@nnrp1.deja.com> <92j7d4$6qd$1@panix2.panix.com>
Message-ID: <0p936.2528$Qb7.277261@newsb.telia.net>

Aahz Maruch wrote:
> >I'm in the process of architecting an application with medium sized data
> >sets. Python dictionaries are tempting as part of the representation
> >mechanism. How do they perform when they have 100,000 entries? Can you
> >offer any other experience with this size data set in Python?
>
> I asked this question a couple of years ago.  The Timbot says it
> shouldn't be a problem -- provided you have enough RAM.

The implementation limit is currently one billion objects.

If you plan to store more objects, you need to add more
entries to the "polys" table in Objects/dictobject.c.  See
http://mail.python.org/pipermail/python-dev/2000-December/010913.html
for more info on how to calculate such entries.

if the above link made your brain hurt, try this one instead:
http://mail.python.org/pipermail/python-dev/2000-December/010943.html

</F>




From cobrien at Radix.Net  Wed Dec 27 14:36:44 2000
From: cobrien at Radix.Net (Cary O'Brien)
Date: 27 Dec 2000 14:36:44 -0500
Subject: Controlling the Parallel Port
References: <mvuj4tkmn22gbth1a6t8fag39902qeeu9l@4ax.com>
Message-ID: <92dgcc$5kk$1@saltmine.radix.net>

In article <mvuj4tkmn22gbth1a6t8fag39902qeeu9l at 4ax.com>,
Simon Faulkner  <News at Titanic.co.uk> wrote:
>Is it easy to control (read/write) the Parallel Port in linux using
>Python?
>
>I am hoping to use it to drive some simple electronics
>

You need to do two things to access the parallel port from
a user program.

1) call ioperm() to get permissions to access the ports.  man ioperm
   should provide the gory details.  Requires root access.

2) use the inb() macro to read, and the outb() macro to write.  See
   man outb for the gory details.  Pay attension to the note about using
   -O or -O2 on the compile command line.  They matter.

Now to access this from python, you can either 

a) write C extensions for ioperm() and inb/outb (and run as root)

b) write C extensions for ioperm() and inb/outb and hack the kernel
   to let user programs execute ioperm()

c) write a C program that does the ioperm and low level read/write and access it
   from python, like say using popen2.  If you make the C program suid-root, you 
   don't have to run python as root.

d) Write a device driver for your device.

Much as I like b), I use c).  Well, to be honest I access the C program
from TCL, but it is the same idea.  Works great.

Is this a microwire device? I.E. a serial eeprom?  If so, drop me a line,
I have a program that may do the trick.

-- cary



From gtalvola at NameConnector.com  Fri Dec 29 12:23:02 2000
From: gtalvola at NameConnector.com (gtalvola at NameConnector.com)
Date: Fri, 29 Dec 2000 17:23:02 GMT
Subject: fcgi.py
References: <3a4a865c.1325844015@localhost> <3a4a90ab.1328483109@localhost> <92f20t$re4$4@newsserver.rrzn.uni-hannover.de> <3a4c1e99.60463812@newnews.prodigy.net>
Message-ID: <92ih9f$kgt$1@nnrp1.deja.com>

In article <3a4c1e99.60463812 at newnews.prodigy.net>,
  victor at prodigy.net wrote:
> Well, my most important need was not to have to open database
connections every
> time CGI executed. Currently I don't need multiple instances of the
script
> running simultaneously. If and when I do, I plan to add
multithreading to my
> FastCGI library (I actually believe it had already been done).

Here's another suggestion: skip FastCGI -- I'm sure it's great on
Unixes, but I was never able to get it to work on NT (it sounds like
you had better luck than I did).  Instead, use
the WebKit application server, and serve it up with Apache and the
WebKit mod_python adapter.  This works on NT as well as Unixes, it's
easy to set up, and it's fast because no new processes need to
be spawned for each request.  WebKit is part of WebWare for Python,
and it has lots of other nice features besides being a fast
application server.  Check it out...

http://Webware.sourceforge.net/
http://www.modpython.org/


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


From shindich at my-deja.com  Fri Dec  8 20:56:06 2000
From: shindich at my-deja.com (shindich at my-deja.com)
Date: Sat, 09 Dec 2000 01:56:06 GMT
Subject: Python wrapper for C++ STL?
References: <3A31760E.D157EB54@erols.com> <C0gY5.126062$U46.4232859@news1.sttls1.wa.home.com>
Message-ID: <90s3fk$j8p$1@nnrp1.deja.com>

In article <C0gY5.126062$U46.4232859 at news1.sttls1.wa.home.com>,
  "Greg Jorgensen" <gregj at pobox.com> wrote:
> "Edward C. Jones" <edcjones at erols.com> wrote:
>
> > Has anyone written a Python wrapper for the C++ Standard Template
Library?
>
> C++ templates are compile-time constructs, not run-time features.
Templates
> are analogous to macros in C. They are not library functions that can
be
> wrapped up and called from Python.
>
> --
> Greg Jorgensen
> Deschooling Society
> Portland, Oregon, USA
> gregj at pobox.com
>
>
With all due respect, what would be wrong with a definition like
std::map<PyObject*, PyObject*>? Obviously there is no type safety here.
But Python is not type safe.

Regards,

Alex Shindich
visit http://www.shindich.com/


Sent via Deja.com http://www.deja.com/
Before you buy.


From china at usenet.com  Wed Dec  6 21:41:36 2000
From: china at usenet.com (Albert Chin-A-Young)
Date: 7 Dec 2000 02:41:36 GMT
Subject: urllib2, proxy, basic authentication
References: <90hk8k$fqh$1@bob.news.rcn.net> <3A2D61C1.9182431D@motorola.com> <90k859$adl$1@bob.news.rcn.net> <90lclk$38p$1@nnrp1.deja.com>
Message-ID: <90mtd0$33n$1@bob.news.rcn.net>

matthewrussell at my-deja.com wrote:
> A way to pass authentication details would be to specify them in the
> url as follows

> import urllib2

> urllib2.urlopen("http://username:password at www.mydomain.com")

Tried that and it does not work. I get:
ValueError: invalid literal for int(): [password]@[domain]

> In article <90k859$adl$1 at bob.news.rcn.net>,
>   Albert Chin-A-Young <china at the.web.com> wrote:
>> Jeremy Lamplough <jeremy.lamplough at motorola.com> wrote:
>> > try setting the proxy host prior to importing the other modules.
>>
>> > import os, sys
>> > os.environ['http_proxy'] = 'http://[PROXY_HOST]:5865'
>> > import urllib2
>> > from urllib2 import urlopen
>>
>> > The way i'm doing basic stuff, with NO authentication is (to
> download a
>> > stock quote)
>>
>> We don't have a problem with http_proxy and the remote host not
>> requiring authentication. However, if the remote host does require
>> authentication, then we have our problem.
>>
>> --
>> albert chin (china @at@ thewrittenword .dot. com)

-- 
albert chin (china @at@ thewrittenword .dot. com)


From pbjorn at uswest.net  Tue Dec 26 13:35:14 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Tue, 26 Dec 2000 11:35:14 -0700
Subject: Formatting for hh:mm:ss
References: <Pine.GSO.4.21.0012261309370.13920-100000@z.glue.umd.edu>
Message-ID: <3A48E4E2.81F7BC18@uswest.net>

This should work (although it might be easier doing it directly...):

 def hms(secs):
  import time
  time_tuple = (0,0,0,0,0,secs,0,0,0)
  time_val = time.mktime(time_tuple)
  time_local = time.localtime(time_val)
  return time.strftime('%X', time_local)

-- bjorn


Roy Katz wrote:

> Hello,
>
> I've searched the 'time' and 'locale' modules  for any function which
> would take a difference of two time.time()'s and format the result as
> hour:minutes:seconds, but I can't find any.  All I am doing is this:
>
>   t1 = time.time()
>   # some processing here
>   t2 = time.time()
>
>   total = t2 - t1
>
>   print_time( total )
>
> where print_time() would format the time as above. Does anyone know if
> this function already exists?
>
> Thanks,
> Roey



From sholden at holdenweb.com  Thu Dec 14 10:23:42 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 14 Dec 2000 10:23:42 -0500
Subject: NewBie question: cannot import TKinter and os
References: <90m7fc$r6m$1@nnrp1.deja.com> <p6qitouk6ow.fsf@informatik.hu-berlin.de> <3A368893.FDC9CAC7@boi.hp.com>
Message-ID: <IE5_5.7591$4y4.195350@e420r-atl1.usenetserver.com>

Bill Anderson <anderson at boi.hp.com> wrote in message
news:3A368893.FDC9CAC7 at boi.hp.com...
> Martin von Loewis wrote:
> >
> > etsang at my-deja.com writes:
> >
> > > 1. at the very first beginning, it says import site error. I know this
> > > module site is no longer required in Python 2.0 after dereading some
> > > doc in pytho.org. how can I make thie error go away??
> >
> > That is incorrect. Where did you read that? site.py is still needed.
>
> Perhaps from the site.py included in 2.0?
> """
> Append module search paths for third-party packages to sys.path.
>
> ****************************************************************
> * This module is automatically imported during initialization. *
> ****************************************************************
>
> In earlier versions of Python (up to 1.5a3), scripts or modules that
> needed to use site-specific modules would place ``import site''
> somewhere near the top of their code.  Because of the automatic
> import, this is no longer necessary (but code that does it still
>         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> works).
>
> """
> (emphasis mine)
>
> I, too am having problems importing site.py and os, at least Zope is
> anyway. I can do it just fine from the interpeter... it seems that not
> allcode that does it still works. :(
>
What that means is that, since Python 1.5 final, the site module is imported
without the programmer needing to explicitly request it.  The code is still
an important part of the startup sequence -- that's why the import was made
automatic!

Among other things, site.py looks for sitecustomize.py, and imports that if
it exists, allowing you to provide customization to your environment.

regards
 Steve






From cnelson9 at gte.net  Tue Dec 26 18:28:35 2000
From: cnelson9 at gte.net (Clifford J. Nelson)
Date: Tue, 26 Dec 2000 23:28:35 GMT
Subject: Python in "math through programming" curriculum
References: <2g2t3t06r2agrrjmfnr41eu4lc2u39a8qi@4ax.com> <3A3EFF47.67DE69CF@gte.net> <513v3toslkjkggh9oumlf20qc9bi1gdtif@4ax.com> <3A3FC677.FDB19937@gte.net> <CeP%5.521$AH6.45315@newsc.telia.net> <3A401C2E.1FF789A8@gte.net> <n1t14ts1somvhn7cmqfvu5q8ie1dvkrque@4ax.com>
Message-ID: <3A4930D2.B45CF0F@gte.net>


Kirby Urner wrote:
> 
[snip]

> 
> In your case, I'm not aware of any alternative product,
> just the negative comments, plus some Mathematica
> worksheets which are not even attempting to interpret
> that whole philosophy.  So again, rather than indulge
> in negative comments, how about you advertise your
> own positive contributions, i.e. your own curriculum
> writing making use of these other computer languages
> you've been posting about (the DrScheme team has
> certainly set a high standard, with their high quality
> materials).
[snip]

Half of the DrScheme demo programs have not even been
debugged completely.

Mathematica electronic notebooks are the very best way to
present any mathematical ideas. You can use traditional
mathematical notation with type setting if you want to, and
all statements can be actually executed. Graphics, sound,
and starting with Mma Version 3, interaction.

My Mathematica electronic notebooks get to the heart of the
matter in Synergetics: the "synergetics triangulated
coordinate system" with which Bucky wrote, "you know", as
opposed to the rectilinear coordinate system, with which
"you guess".

They can be viewed with MathReader (free) if you don't have Mathematica.

At: http://forum.swarthmore.edu/epigone/geometry-research/brydilyum

It is a good small exercise to translate the notebooks to
other languages.

  Cliff Nelson


From milhad at ptt.yu  Fri Dec 15 14:55:39 2000
From: milhad at ptt.yu (Milos Hadzic)
Date: Fri, 15 Dec 2000 20:55:39 +0100
Subject: new with python..
Message-ID: <91dt4e$d3u$1@barni-1.tehnicom.net>

i am new with this proggraming tool and i would appriciate if you would tell
me where can i get a good tutorial.




From dennis at codegrinder.org  Fri Dec 15 16:17:53 2000
From: dennis at codegrinder.org (Dennis Conrad)
Date: Fri, 15 Dec 2000 22:17:53 +0100
Subject: Accessing serial port on windows
References: <91dds1$b0i$05$1@news.t-online.com>
Message-ID: <91e20b$pqi$01$1@news.t-online.com>

Thank you both very much!

Dennis




From noahspurrier at my-deja.com  Thu Dec 21 23:59:14 2000
From: noahspurrier at my-deja.com (noahspurrier at my-deja.com)
Date: Fri, 22 Dec 2000 04:59:14 GMT
Subject: Wanted: Python White Papers
References: <91oj8v$i0r$1@nnrp1.deja.com>
Message-ID: <91un30$gdo$1@nnrp1.deja.com>

I found that Inktomi uses Python in their search engine product.
This is the Inktomi copyright acknowledgment page:
http://software-search.infoseek.com/help/copyright.html
It's on the infoseek domain, but I downloaded the
Inktomi search engine from their main site and it's Python...

Yours,
Noah

In article <91oj8v$i0r$1 at nnrp1.deja.com>,
  noahspurrier at my-deja.com wrote:
> Is anyone aware of any white papers to explain the use of Python
> at any big, high-profile companies?
> I'm trying to use a "look-how-these-cool-companies-use-python"
> argument to defend the use of Python for a project. That won't be my
> only argument, of course :-) But currently Python is not winning many
> management converts here based solely on its technical merits.
> Many still say, "What's Python? Never heard of it...". It helps
> a lot when I say "Google uses it!", but it would help more if I
> knew HOW it is used and WHY.
>
> Actually, the only high-profile company that uses Python that I
> can think of is Google. Is anyone familiar with their architecture?
> How does Python fit into their design?
>
> Is there anyone out there from Google that can represent Google's use
> of Python? What are Google's plans for the future of Python in their
> architecture? I heard that they might be switching to Java.
>
> Yours,
> Noah Spurrier
>
> Sent via Deja.com
> http://www.deja.com/
>


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


From rey at typhoon.sdsu.edu  Tue Dec 19 11:43:43 2000
From: rey at typhoon.sdsu.edu (Serge Rey)
Date: Tue, 19 Dec 2000 08:43:43 -0800
Subject: Digest Number 1394
In-Reply-To: <977234466.61955@egroups.com>
References: <977234466.61955@egroups.com>
Message-ID: <20001219084343.V3916@typhoon.sdsu.edu>

-snip- 
> >
> > The problem is that you have to click on the *frame* not the (empty)
> > space in the frame.
> >
> > Try this:
> >
> > frame = Frame(root, bg='red', width=100, height=100)
> >
> >
> > John Grayson
> 
> Well, it doesn't work, either. Probably it will work on linux but not on
> win32.

it does work fine under linux, as well as for win32 (well at least win95). i
think john's point can be illustrated if you maximize the toplevel and then
compare what happens if you click inside the frame (i.e., with the red
background) vs. outside that frame (i.e., grey background of the toplevel
widget).

HTH 


-- 
Sergio J. Rey
Editor, International Regional Science Review
Associate Professor
Department of Geography, San Diego State University
San Diego, CA 92182
Phone 619.594.8029  Fax 619.594.4938
http://typhoon.sdsu.edu/rey.html

This class is hard, I mean this math stuff is not logical.
	- a former student



From root at [127.0.0.1]  Fri Dec 15 01:14:14 2000
From: root at [127.0.0.1] (deadmeat)
Date: Fri, 15 Dec 2000 06:14:14 GMT
Subject: Python advocacy
References: <91bln3$lmr$1@nnrp1.deja.com> <3A39AC3E.FA75C4FF@engcorp.com>
Message-ID: <WEi_5.7880$KY1.21268@news1.rivrw1.nsw.optushome.com.au>

> There is no reason to focus on the fact that people appear to take this
> sort of thing personally.  That's probably just human nature, and likely
> not even a Bad Thing.

the reason people get annoyed, and the reason it's a Bad Thing, is because
most people are incapable of promoting their widget on it's own merits, but
instead resort to bashing a competing widget. IMO, unless your widget can
stand on it's own two feet and be counted, it's worthless.





From jcm at bigskytel.com  Tue Dec 19 08:15:46 2000
From: jcm at bigskytel.com (David Porter)
Date: Tue, 19 Dec 2000 06:15:46 -0700
Subject: while true: !!!
In-Reply-To: <91nc7h0s0p@news2.newsguy.com>; from aleaxit@yahoo.com on Tue, Dec 19, 2000 at 11:10:30AM +0100
References: <mailman.977180341.10206.python-list@python.org> <slrn93t7bv.le.grey@teleute.rpglink.com> <KTz%5.48644$x6.22165257@news2.rdc2.tx.home.com> <91nc7h0s0p@news2.newsguy.com>
Message-ID: <20001219061546.A9939@bigskytel.com>

* Alex Martelli <aleaxit at yahoo.com>:
> "Rainer Deyke" <root at rainerdeyke.com> wrote in message
> news:KTz%5.48644$x6.22165257 at news2.rdc2.tx.home.com...
> >
> > 0 is false, None is false, [] is false, () is false, {} is false, and x is
> > false where x is a class instance and x.__nonzero__() returns false.
> > Everything else is true.  I think.
> 
> Another couple of 'false's: a class instance which doesn't define
> __nonzero__ but does define __length__ when the latter returns 0;
> and, C-implemented user objects, similarly to class instances.
> 
> I don't _think_ there are any more...

An empty string is also false.


 David
 



From fredrik at effbot.org  Tue Dec  5 23:36:48 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 06 Dec 2000 04:36:48 GMT
Subject: Rounding error in Python 2.0
References: <90jscs$uno$1@nnrp1.deja.com>
Message-ID: <AnjX5.2006$O%.346064@newsc.telia.net>

sragsdale at my-deja.com write:
> Can anyone explain to me why python2 can't read a float from a string?
> Python 1.5.2 didn't have this problem.

it sure had the same problem.  it just didn't tell you.

    http://www.python.org/cgi-bin/moinmoin/FrequentlyAskedQuestions

    Q. Floating point is broken!

    A. Binary floating point cannot represent decimal fractions
    exactly, so some rounding always occurs (even in Python
    1.5.2).

    What changed is that Python 2.0 shows more precision than
    before in certain circumstances (repr() and the interactive
    prompt).

for more info, see:

    http://www.python.org/cgi-bin/moinmoin/RepresentationError

</F>





From manyong at gnu.org  Fri Dec  8 03:29:54 2000
From: manyong at gnu.org (Man-Yong Lee)
Date: Fri, 08 Dec 2000 08:29:54 GMT
Subject: SIGCHLD handle & socket
References: <90q16m$tm2$1@nnrp1.deja.com>
Message-ID: <3A309924.3DACBB0C@gnu.org>

> Traceback (most recent call last):
>   File "rjcall.py", line 19, in ?
>     (conn, addr) = s.accept()
> socket.error: (4, 'Interrupted system call')

I think you can guess the error output.

The server process is always(almost) in the middle of
accpet()ing. And when one of childs die, SIGCHLD is
sent to the parent and accept() is interrupted.

socket.error is raised.

So...

redjade at my-deja.com wrote:
> my code is below
> ----------------
> signal.signal(signal.SIGCHLD, handle)
> 
> while 1:

          try:
              (conn, addr) = s.accept()
          except socket.error:
              continue

I think you should protect the accept routine with
try: ... except: ...

>         print 'Connected by', addr
>         pid = os.fork()
>         if pid == 0:
>                 s.close()
>                 conn.send('\n\nConnection Established\n\n\n')
>                 time.sleep(3)
>                 conn.close()
>                 os._exit(0)
>         else:
>                 conn.close()

Maybe you might protect fork() with try: ... except: ...
'cause SIGCHLD can be sent in any time.

At least it works for me.

Good luck!
-- 
Happy Python!
             Man-Yong (Bryan) Lee  <yong at linuxkorea.co.kr>


From fredrik at effbot.org  Thu Dec 21 02:10:16 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Thu, 21 Dec 2000 07:10:16 GMT
Subject: re different between US and German versions of NT
References: <mailman.977362144.17768.python-list@python.org>
Message-ID: <s1i06.771$Qb7.67169@newsb.telia.net>

Darrell Gallion wrote:
> Sorry, left the example at work. Also don't have access to the German
> machine any longer.

I have access to non-US machines (SRE was developed on one).
if you can find the pattern, please post it.

</F>




From joseamartin at venmex.com  Mon Dec 18 10:12:57 2000
From: joseamartin at venmex.com (Jose Martin)
Date: Mon, 18 Dec 2000 11:12:57 -0400
Subject: Super Database Tool in Python for Ever !!!
Message-ID: <t3satn4tam43f9@corp.supernews.com>

Hi, sorry this tool does not exist today !.

But i have a couple ideas to produce the an incredible tool
any voluntary ???.

It is a Convergence of Various tools like:
>From :Gupta's Quest:
    1. You can Create Tables, Browse Tables in Grids and modify data and
save data.
    2. You can copy rows of grids and paste into other grids.
    3. You can Make SQL sentences and browse them in grids.
>From :Oracle SQLPlus :
    1. You can call a Text Editor and save SQL Scripts.
    2. You can add command and procedures and call them from a promt.
    3. Variuos command like DESC,EXEC ,   etc... that make your life esay
and happy.
From: Linux Shell:
    1. Tab Completition. utilfull .

>From :Python IDLE.
1. Auto-Complete Atributes and Call Tips.
2. Extensible Sintax HighLighting.


>From Python Languaje:
A complete applicattion  and a module that  you can import from a python
shell and work with them...


I have a lot of work but i can do some code, any voluntary for Make this
tool reality ?

It is cleary, this tool will go to be the Best tool for Database related
work.

Thanks....
By.

Jose Antonio Martin
joseamartin at venmex.com








From mal at lemburg.com  Mon Dec  4 19:22:39 2000
From: mal at lemburg.com (M.-A. Lemburg)
Date: Tue, 05 Dec 2000 01:22:39 +0100
Subject: More Unicode Trouble
References: <ujnd2tsotru85f9akd5anr2ji16duc18hs@4ax.com> <9080rp01ier@news1.newsguy.com>
Message-ID: <3A2C354F.F5A27785@lemburg.com>

Alex Martelli wrote:
> 
> > and I can't find any documentation on how to handle Unicode.
> 
> I think (and I *dearly hope* I'm wrong) that the best doc
> is still Lemburg's "proposal", at
> http://www.lemburg.com/files/python/unicode-proposal.txt
> Just read it 'imagining' it's not a proposal but an actual
> description of how things work, mentalling translating all
> the 'should do so-and-so' into 'does so-and-so'.  Not sure
> why this rewording hasn't been actually performed (making
> this document, e.g., a chapter or appendix of the Python
> library reference, or whatever) -- as I say, I hope I'm wrong
> here and the current 'real official' docs already have all
> of this wealth of information, but I don't think they do.

The above file was indeed the proposal that was used as
basis for the Unicode implementation in Python. 

Many of the higher-level interfaces are already documented in the
standard documentation, but if you care about the internals
and all the gory details, then it still is the number one
reference.

Note that most of the trouble users have with Unicode comes from
not understanding the difference between 8-bit strings (without
any encoding information) and 2-byte (single encoding) Unicode --
these are really two different worlds and bringing them together
is a hard piece of work. 

Other languages which seem to make
Unicode easy typically don't make this distinction at all: they
simply use Unicode all the way. But this is something we can't
do just yet in Python since it would break too much code.

-- 
Marc-Andre Lemburg
______________________________________________________________________
Company:                                        http://www.egenix.com/
Consulting:                                    http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/


From jcaldero at ic.sunysb.edu  Tue Dec 19 02:32:12 2000
From: jcaldero at ic.sunysb.edu (Jp Calderone)
Date: Tue, 19 Dec 2000 02:32:12 -0500
Subject: embedding help
Message-ID: <3A3F0EFC.D2F2BAE9@ic.sunysb.edu>


  I'm trying to embed python in a much larger project (a MUD)
as a scripting language, but I'm stuck, essentially right at
the begining.  The docs at python.org on this section seem,
ahh, incomplete (unless I'm missing something).  What I'm
trying to do ATM is something along the lines of let python
code access and modify the values of C++ objects outside of
the interpreter.  I believe the correct way to go about this
is to create a new python object type that understands how
to find the C++ objects in the C++ portion of the program,
providing it with get and set attr methods.  However, I'm
at a loss as to how to actually do this.  I found the 
PyTypeObject struct, and I see it has a lot of fields for
functions that do things, but no actual data storage, so
I don't see any way to keep track of which python object
instances points to which C++ object instances.  Can anyone
point me to any examples of how to do this, or explain it?
Or am I barking up the wrong tree entirely?

 Thanks for any and all help,
 Jp Calderone


From scarblac at pino.selwerd.nl  Sun Dec 31 06:13:01 2000
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 31 Dec 2000 11:13:01 GMT
Subject: What happened in this class
References: <92mgca$7kktg$1@ID-11957.news.dfncis.de>
Message-ID: <slrn94u4vt.m8h.scarblac@pino.selwerd.nl>

Emile van Sebille <emile at fenx.com> wrote in comp.lang.python:
> I saw Pearu's lazy extension and built and installed it on
> my linux system(recent cvs).  I thought I'd see how it
> played in a class vs getattr and got this result.  Thinking
> something broke along the way, I then switched over to the
> win95 system(2.0 #8), and got the same thing.  This doesn't
> happen under 1.52.
> 
> 
> class test:
>     def __getattr__(self, name):
>         return "egs"
> 
> t = test()
> print dir(t)[1:5]
> 
> ['e', 'g', 'g', 's']
> 
> 
> Now I know this is contrived, but I still didn't expect
> these results.  Should I have?

Real weird. dir(t) returns ['e', 'e', 'g', 'g', 's', 's']. It's pretty
simple to find out what it actually calls by adding another print
statement:

class test:
   def __getattr__(self, name):
      print name
      return "egs"
      
t=test()
print dir(t)

prints:
__members__
__methods__
['e', 'e', 'g', 'g', 's', 's']

So here we see that apparently dir() in 2.0 works by calling the __members__
and __methods__ special functions, and they both return "egs", and since
dir() was expecting a list it reads that as ['e','g','s'], adds them
together, and sorts them.

These special functions are *not* documented in chapter 3 of the 2.0
language reference and I suppose that is wrong. They don't look like the
sort of thing you would want to change though. Or maybe they can be useful,
if you define a data member by means of __getattr__ but you still want it
to show up in dir(). Needs docs.

In conclusion, making a __getattr__ that returns something completely
regardless of the name you put in is not a good idea :-)

-- 
Remco Gerlich


From SBrunning at trisystems.co.uk  Mon Dec 11 04:35:41 2000
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Mon, 11 Dec 2000 09:35:41 -0000
Subject: creating an odbc link with python
Message-ID: <31575A892FF6D1118F5800600846864D5B15A5@intrepid>

> From:	Niki Spahiev [SMTP:spahievi at vega.bg]
>     from odbc_installer import config_data_source, ODBC_ADD_SYS_DSN
 
What's this odbc_installer thing? I don't seem to have it, and nor can I
find it - a couple of references to *using* it on deja, but that's it.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk





-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From nielsenjf at my-deja.com  Mon Dec  4 09:30:56 2000
From: nielsenjf at my-deja.com (John Nielsen)
Date: Mon, 04 Dec 2000 14:30:56 GMT
Subject: python/erlang interface and free books
References: <903241$rr6$1@nnrp1.deja.com> <903ch1021gl@news1.newsguy.com>
Message-ID: <90g9r0$urc$1@nnrp1.deja.com>

It turns out there is a python/erlang interface!

http://csmctmto.interpoint.net/didx/python_erlang.html

> aleaxit at yahoo.com --
> Probably, but, on the other hand, I know no book on Erlang that
> makes a good introduction to functional programming.  (Not saying
> there *aren't*, mind you -- just that *I* don't know them!).

I found 2 books available for free on the web! Maybe people are starting
to learn about what Bruce Eckel does.

Part one of  Concurrent Programming in Erlang (the 1st 9 chapters) is
available for free on the web.

http://www.erlang.org/download/erlang-book-part1.pdf

In the amazon description it says:

"Requires previous programming experience, but no knowledge of
functional or logic languages."

The 2nd book is:
Erlang in Real Time
http://www.serc.rmit.edu.au/~maurice/erlbk/pdf/main.pdf

For now, I'll give erlang a shot, and also look at python's functional
programming modules.

Thanks for your help,

john




Sent via Deja.com http://www.deja.com/
Before you buy.


From aahz at panix.com  Fri Dec 15 15:12:16 2000
From: aahz at panix.com (Aahz Maruch)
Date: 15 Dec 2000 12:12:16 -0800
Subject: no current thread warning
References: <t3ks6hg7c7rgae@corp.supernews.com>
Message-ID: <91dtv0$g93$1@panix6.panix.com>

In article <t3ks6hg7c7rgae at corp.supernews.com>,
Charles Crain <ccrain at austin.rr.com> wrote:
>
>Near as I can tell, this relates to the statement in the documentation for
>the threading.currentThread() function, which states something to the effect
>of "for threads not created by the threading module, a dummy thread with
>limited capabilities will be created."
>
>A have a couple questions regarding this:
>1.  Exactly how are the capabilities limited?  Are there any gotchas I
>should be on the lookout for?

If you look in threading.py, you'll see this:

# Dummy thread class to represent threads not started here.
# These aren't garbage collected when they die,
# nor can they be waited for.
# Their purpose is to return *something* from currentThread().
# They are marked as daemon threads so we won't wait for them
# when we exit (conform previous semantics).

Essentially, they're kept track of so Python *doesn't* try to do
anything with them.

>2.  Why is the warning printed out unconditionally?  IMHO, it should only be
>printed out if the error is potentially fatal, and/or amounts to an asserion
>failure within the interpreter.  If I interpret the dox correctly, this
>message could at best be considered an informational, and should therefore
>be able to be turned off.  It is only cosmetic, but in considering whether
>to commercially release products we have written in Python, it is an issue.

Again, if you look in threading.py, you'll see that there's an __debug__
test that fails if you do "python -O".  In fact, there are several such
locations within threading.py.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

"It's 106 miles to Chicago.  We have a full tank of gas, a half-pack of
cigarettes, it's dark, and we're wearing sunglasses."  "Hit it."


From fredrik at effbot.org  Mon Dec 18 05:05:30 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Mon, 18 Dec 2000 10:05:30 GMT
Subject: how to do in-memory transformations ?
References: <mailman.977127541.4315.python-list@python.org>
Message-ID: <Kjl%5.1$AH6.46@newsc.telia.net>

Andreas Jung:

> Due to the poor PIL documentation I have not found a suitable
> solution yet. Any ideas ?

how about being a little more polite?

if you cannot figure out how to open, load, and resize an image
from the documentation, I doubt you've read it at all.

(hint: to open, load, and resize, use the methods "open", "load", and
"resize".  load is optional.  to figure out the format, use the attribute
named "format").

to figure out how to save an image to a string (instead of a temporary
file), you might have to read the documentation a little bit more care-
fully:

here's the relevant part:

    "save(outfile, options), save(outfile, format, options).

    Saves the image under the given filename

    ...

    You can use a file object instead of a filename. In this case, you
    must always specify the format. The file object must implement
    the seek, tell, and write methods, and be opened in binary mode."

if you don't know how to create an in-memory file object, look
up StringIO in the Python library reference.

</F>




From kragen at dnaco.net  Sat Dec 30 20:48:44 2000
From: kragen at dnaco.net (Kragen Sitaker)
Date: Sun, 31 Dec 2000 01:48:44 GMT
Subject: HTML Parser
References: <mailman.978224102.28797.python-list@python.org>
Message-ID: <0gw36.45$9s1.9126@e420r-atl2.usenetserver.com>

In article <mailman.978224102.28797.python-list at python.org>,
Voitenko, Denis <dvoitenko at qode.com> wrote:
>HTMLtags=re.compile('<.*>')

In a string like "x<a>b<c>d", this will match "<a>b<c>", because the .*
matches "a>b<c".  This explains your problem.

Fixing it is harder.
-- 
<kragen at pobox.com>       Kragen Sitaker     <http://www.pobox.com/~kragen/>
Perilous to all of us are the devices of an art deeper than we possess
ourselves.
       -- Gandalf the White [J.R.R. Tolkien, "The Two Towers", Bk 3, Ch. XI]




From grey at despair.rpglink.com  Sat Dec 30 00:37:19 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Sat, 30 Dec 2000 05:37:19 -0000
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <3A4D6A05.EEC97FA6@uswest.net> <slrn94qqvg.mgg.grey@teleute.rpglink.com> <ije36.96813$x6.43604873@news2.rdc2.tx.home.com>
Message-ID: <slrn94qt4f.mgg.grey@teleute.rpglink.com>

On Sat, 30 Dec 2000 05:23:26 GMT, Rainer Deyke <root at rainerdeyke.com> wrote:
>At that point it's better to use 'while 1', which is at least idiomatic and
>explicit in its inelegance.

    I'm sorry, but I don't agree.  Even though I have no problem with while 1:
initing variables before the loop is more preferable to a break.  Esp. when
one can put the condition in the while; unlike reading lines from a file where
one cannot.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From root at rainerdeyke.com  Fri Dec  8 19:54:43 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Sat, 09 Dec 2000 00:54:43 GMT
Subject: Python wrapper for C++ STL?
References: <3A31760E.D157EB54@erols.com>
Message-ID: <npfY5.1850$x6.2577114@news2.rdc2.tx.home.com>

"Edward C. Jones" <edcjones at erols.com> wrote in message
news:3A31760E.D157EB54 at erols.com...
> Has anyone written a Python wrapper for the C++ Standard Template
> Library?

Why would anybody want to?  Python provides most STL fnctionality natively.
Such a wrapper would need to constantly translate C++ exceptions to Python
exception and vice versa, making the wrapper be more trouble than it's
worth.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From urner at alumni.princeton.edu  Mon Dec 18 17:07:12 2000
From: urner at alumni.princeton.edu (Kirby Urner)
Date: Mon, 18 Dec 2000 14:07:12 -0800
Subject: Python in "math through programming" curriculum
Message-ID: <2g2t3t06r2agrrjmfnr41eu4lc2u39a8qi@4ax.com>

Repost from edu-sig at python.org elist -- one typo/URL fixed.

=============================

List-Archive: <http://www.python.org/pipermail/edu-sig/>
Date: Mon, 18 Dec 2000 01:08:01 -0800


I've upgraded http://www.inetarena.com/~pdx4d/ocn/clubhouse.html 
to include some RSA stuff.  I think that page is finally 
close to done and I plan to give it a rest.[1]

In exploring the web for relevant input, it became clear 
to me that I'm venturing in territory typically covered 
in college computer science courses.  

That doesn't need to change, but I remain persuaded that 
by adopting a "computer at your elbow" approach to math 
teaching, we could have a lot more segues between K-12 
and what's now considered college-level material.

Awhile back, Tim Peters recommended 'Concrete Mathematics' 
(Knuth a co-author) as a good example of a "math through
programming" text book.  I subsequently bought it, and 
had to agree.  So how do we "lower a ladder" to middle 
and high schoolers starting from there?  I think Python
is part of the answer.

In sum, none of the content I'm covering in my curriculum
writing is really new.  All that I'm doing is wiring up 
the topics in a somewhat different way, changing the mix,
reorganizing the hyperlinks. This is a fairly conservative
approach, in other words.[2]  

However, I think the result could be a much higher level 
of math appreciation for younger students (or for adults 
returning for more), plus better segues into computer 
science, for those wishing to go that route.

>"We have to reinvent the wheel every once in a while,
>not because we need a lot of wheels; but because
>we need a lot of inventors."  - Bruce Joyce
>
>I've always loved that quote (and the book in which I found it,
>for that matter ;-)  so I greatly appreciated these comments
>by Kirby.
>
>jeff

That's a great quote jeff, hadn't seen it before.  Thanks
for posting it.

Kirby

Notes:

[1]

The Miller-Rabin stuff was confusing because I think some 
web authors don't communicate it properly.  Fortunately, 
Knuth includes it in 'The Art of Programming' vol 2 (took
me awhile to realize that) and his Algorithm P on pg. 395
seems pretty clear -- helped me streamline the gist of the
test:

Given candidate prime n, random base b < n, and m,s such 
that m*(2**s) = n-1:

  def algP(m,s,b,n):
      """
      based on Algorithm P in Donald Knuth's 'Art of
      Computer Programming' v.2 pg. 395 
      """
      result = 0
      y = pow(b,m,n)
      for j in range(s):
         if (y==1 and j==0) or (y==n-1):
            result = 1
            break
         y = pow(y,2,n)
      return result

If this spits back a 1, your candidate n passes the test 
for base b.  The idea is to test a number of random bases 
(except I go with low primes in sequence, as per Paul 
Emerson's thesis, an OK alternative, yes?) -- the more 
bases you test, the higher chance a number still passing 
is prime; with the chance of being wrong equal to 
(1./4)**(number of tests) -- gets small quickly.

Anyway, the above source is quite a bit more succinct 
than before.

[2]  

Probably the one thing I include that's controversial is 
mention of Bucky Fuller's concentric hierarchy of polyhedra
-- a way of scaling/organizing common polys which brings 
out some whole-number relationships and morph-relationships
which geometers haven't traditionally focussed on (even 
though they've always been there).  I think Fuller's 
approach is a pedagogic advance and makes polys more 
friendly to younger people (a thesis I've empirically 
tested in the field, with good results).  Plus it phases
in sphere packing in tandem with the polys, and that's
highly useful as well.  (Note:  Alexander Graham Bell was
also obsessed with the tetrahedron, not just Bucky).

With more polyhedra earlier in the curriculum, we have 
more opportunities to use them later (say in high school)
as objects subjected to transformations (scaling, translation,
rotation), using object-oriented programming techniques. 
Lots of intro-to-OOP books use polygons as example objects,
(e.g. triangle as a subclass of polygon), but I'd go with 
full-fledged polyhedra instead (e.g. tetrahedron as a 
subclass of polyhedron).  With polyhedra, your Python can 
be about manipulating 3-tuple vectors (x,y,z), not just 
planar (x,y) vectors, and the graphics/animations will 
be far more interesting.  Plus you might want to explore 
with quadrays -- another wrinkle/feature in my curriculum 
writing: see http://www.teleport.com/~pdx4d/quadrays.html 
for more info.






From l_hill at mindspring.com  Fri Dec  8 21:29:45 2000
From: l_hill at mindspring.com (Laura Hill)
Date: Fri, 08 Dec 2000 18:29:45 -0800
Subject: OOPSLA 2001 CFP
Message-ID: <3A319919.473E1A61@mindspring.com>

OOPSLA 2001 Call for Participation

OOPSLA 2001
October 14-18, 2001
Tampa Bay, Florida USA

The ACM OOPSLA Conference (Object-Oriented Programming,
Systems, Languages, and Applications) is the world's
premier forum for object-oriented technology.
If you're involved in object technology, OOPSLA 2001
is the place to be.

Submissions are solicited in the following categories:

Technical Papers,  Tutorials,  Practitioner Reports,
Panels, Demos,  Educators' Symposium,  Workshops,  Posters,
Doctoral Symposium,  DesignFest,  Vendor Exhibits

For more info, check out our Web site:
    http://oopsla.acm.org

Important Dates
March 23, 2001:  Due date for Technical Papers,
    Practitioner Reports, Educators' Symposium,
    Tutorials, Workshops, and Panel proposals,
    and DesignFest problems
July 20, 2001:  Due date for Posters, Demonstrations,
    Doctoral Symposium, and Student Volunteers





From hat at se-46.wpa.wtb.tue.nl  Fri Dec 15 03:56:39 2000
From: hat at se-46.wpa.wtb.tue.nl (Albert Hofkamp)
Date: 15 Dec 2000 08:56:39 GMT
Subject: launching a python program from GNU make (Win2K)
References: <918lp0$6q4$1@nnrp1.deja.com>
Message-ID: <slrn93jn66.p9r.hat@se-46.wpa.wtb.tue.nl>

On Wed, 13 Dec 2000 20:22:02 GMT, jschmitt at vmlabs.com <jschmitt at vmlabs.com> wrote:
>From my experience, GNU make has the most features that I want to use.
>However, I find it so hard to use under Windows.  For example, I wrote

No kidding :-)

>a small python program that I can launch from the command line and
>execute without problems.  I can't seem to launch it from GNU make.  I

We had the same kind of problems, ever since we switched from Novell
to NT servers.
Until the arrival of the latter servers, we had a number of command-line
utils that worked fine within the limits of 'fine' as we know it on W*.

When the Novell server was replaced by an NT server, long filenames
suddenly stopped working, both on the network and locally.
Also, it appeared that 1 command-line program (e.g. GNU-make), could not
start another program (e.g. a compiler), without giving an error of any
kind.
The core of the problem was tracked to the NT server that does not handle
long filenames (even in the 21st century), although they are listed.

I don't know about W2K, but given the fact that it is an NT-based version,
I would not be surprised to see that the situation did not improve.
In fact, I consider command-line oriented programs DEAD on W* systems.

>What could the reason be and how could I fix it?  Does it have

I think the reason is W2K. The only fix I know is complain to B.G.
If that would fail for some reason, there are only 2 options left imho.
1) Switch to visual programming (aka do it the MS way),
2) Drop the platform.

FYI: We did the 2nd option.

>something to do with GNU make going out of it's way to launch sh.exe if
>it's on your hard drive, rather than launching command.exe to shell out

Nope, it is simply not possible to launch a command-line program from
another command-line program.

>PS Choosing a different make is not an option here.

I didn't suggest that :-)



Albert
-- 
Signature is being repaired, sorry for the inconvenience


From nospam at aol.com  Mon Dec  4 04:32:58 2000
From: nospam at aol.com (anand)
Date: Mon, 04 Dec 2000 09:32:58 GMT
Subject: hash() algorithm?
References: <nospam-36B66D.13423003122000@news> <ikzW5.1434$Kd1.226017@newsb.telia.net> <nospam-79EF46.15110903122000@news> <90fitt$q3h$1@news.udel.edu> <87bsus8sx1.fsf@psyche.dnsalias.org>
Message-ID: <nospam-634BA6.02345204122000@news>

<c.evans at clear.net.nz> wrote:
> .../Objects/stringobject.c, function string_hash().

cool!
Although I'm having trouble finding that file....

Is the path diffent for the mac distribution?  Because I don't have a 
Objects directory, and stringobject.c doesn't exist anywhere.


From thomas at xs4all.net  Tue Dec 26 09:38:58 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Tue, 26 Dec 2000 15:38:58 +0100
Subject: One Python 2.1 idea
In-Reply-To: <roy-9332FF.08555126122000@news.panix.com>; from roy@panix.com on Tue, Dec 26, 2000 at 08:55:51AM -0500
References: <mailman.977771949.17911.python-list@python.org> <9292tn$jka$1@nnrp1.deja.com> <929n9c013m3@news1.newsguy.com> <mailman.977835669.28993.python-list@python.org> <roy-9332FF.08555126122000@news.panix.com>
Message-ID: <20001226153858.D25820@xs4all.nl>

On Tue, Dec 26, 2000 at 08:55:51AM -0500, Roy Smith wrote:

> x.port = x.page.frame.server.connection.input.socket.port

> but I do it because I think it makes the code easier to read (and 
> certainly easier to type!), not because I'm particularly worried about 
> how efficient it is.

And that's exactly the right reason, too :) Even though I'm always
interested in optimization, it's purely for the art of it, not for any
actual need for speed. I just like playing with such things :-)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From jon at rne-s.com  Tue Dec  5 10:18:27 2000
From: jon at rne-s.com (jon at rne-s.com)
Date: Tue, 05 Dec 2000 15:18:27 GMT
Subject: chunks in recordsets
Message-ID: <90j0vs$5ot$1@nnrp1.deja.com>

I've written a snipplet whitch i supposed to get a binary field from a
database. Bur it don't work... here it is...

(adoRS, success) = adoConn.Execute('select binary_field from multimedia
where multimedia_id=41616')

anyone who have a hint?

is there any documentaton in ADO in python anywhere?


Sent via Deja.com http://www.deja.com/
Before you buy.


From donn at oz.net  Sun Dec 10 01:11:58 2000
From: donn at oz.net (Donn Cave)
Date: 10 Dec 2000 06:11:58 GMT
Subject: Handling signals and performance issues.
References: <meRU5.211$Mx.12164@sjc-read.news.verio.net> <m9dgmAAe+tM6EwnY@gol.com> <roy-FD4A85.20535009122000@news.panix.com>
Message-ID: <90v6re$99n$0@216.39.151.169>

Quoth Roy Smith <roy at panix.com>:
...
| As for watching the end of a log file without burning CPU, as somebody 
| else already mentioned, sleeping for a second between reads would be one 
| good way.  Another thing you might try is using a select() call to tell 
| you when new data is readable on the file.
|
| You might want to look at the source to the unix "tail" command (you 
| should be able to find it on any of the many GNU/Linux distribution 
| sites) and see how it implements the "-f" option, which is essentially 
| what you are trying to emulate.

The sleep sounds like a winner.  I wouldn't expect too much from
select() in this application.  select() wakes up when the indicated
I/O operation wouldn't block;  for a read operation, because there's
either data or an end of file.  At the end of a disk file, the second
condition obtains.  (I wouldn't think select() is normally even
supported at all for disk files, but I'm too lazy to check.)  The
bottom line is, there's no way to be notified by the system when
new data is appended to the end of a file, you just have to keep
checking.

	Donn Cave, donn at oz.net


From aleaxit at yahoo.com  Mon Dec 11 05:03:12 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 11 Dec 2000 11:03:12 +0100
Subject: Parsing text in python
References: <tYIY5.47$17.2755@m2newsread.uni2.es>
Message-ID: <9128pd02v8p@news2.newsguy.com>

"Arturo P?rez Mulas" <arturo_perez at wanadoo.es> wrote in message
news:tYIY5.47$17.2755 at m2newsread.uni2.es...
Hello,

I was wondering if there is a parsing tool for python similar to yacc/lex. I
have only come across Plex, but
seems very limited in comparison with yacc (lacks iterations).


Berkeley YACC, at least, can generate Python (as well as other
languages).  http://sourceforge.net/projects/byacc/.

Also see
http://www.foretec.com/python/workshops/1998-11/proceedings/papers/aycock-li
ttle/aycock-little.html,
which also includes pointers to other tools/approaches; and
kwParsing, at http://www.pythonpros.com/arw/kwParsing/kwParsing.html
(aka kjParsing it seems).  I'm sure there are many others yet...


Alex






From dnew at san.rr.com  Thu Dec  7 14:54:45 2000
From: dnew at san.rr.com (Darren New)
Date: Thu, 07 Dec 2000 19:54:45 GMT
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <3A2DD2CA.14A5D7E2@engcorp.com> <90l3j9$4bn$1@kapsel.intranet> <udzoi9cedt.fsf@box.home.de> <KHOX5.3213$TU6.94362@ptah.visi.com>
Message-ID: <3A2FEB05.5E9531C4@san.rr.com>

Grant Edwards wrote:
> If a person who speaks two languages is "bilingual" and a
> person who speaks three languages is "trilingual", then what do
> you call a person who only speaks one langage?
> 
> American!

Yah. We just export our culture to y'all. ;-)

Actually, we don't all speak the same language. We can just all understand
each other.

The traveling salesman saw "grits" on the menu. He asked the waitress "What
are grits?"  She said "They're extra."  "But what *are* they?" he asked
again.  "They're fifty cents" she answered.  "Yes, I'll have the grits,
please."

-- 
Darren New / Senior MTS & Free Radical / Invisible Worlds Inc.
San Diego, CA, USA (PST).  Cryptokeys on demand.
Personal malapropism generator free with purchase!
Steganography: The manual is hidden in the source code.


From judi_ahivodji at yahoo.fr  Fri Dec 29 05:18:18 2000
From: judi_ahivodji at yahoo.fr (judi_ahivodji at yahoo.fr)
Date: Fri, 29 Dec 2000 10:18:18 GMT
Subject: Newbie question : What is a thread?
Message-ID: <92hod8$298$1@nnrp1.deja.com>

Hi !
I'm a beginner in Python an I would like someone to tell me what is a
thread
and/or where I can find other deep explanations about it.

Thanks


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


From root at rainerdeyke.com  Sat Dec 23 12:36:03 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Sat, 23 Dec 2000 17:36:03 GMT
Subject: Some Python 2.1 ideas
References: <mailman.977590932.19335.python-list@python.org>
Message-ID: <7o516.66566$x6.31049352@news2.rdc2.tx.home.com>

"Bob Alexander" <bobalex at home.com> wrote in message
news:mailman.977590932.19335.python-list at python.org...
> New operator: embedded assignment
>
> As C and Java programmers know, it is sometimes more expressive to embed
> assignments inside other expressions. It's my guess that one reason Python
> doesn't support this is to make "=" illegal inside expressions, thus
> avoiding the "=" vs. "==" errors common in C. A response to that objection
> is to use a different token for embedded assignment -- I suggest ":="
since
> it's already familiar to many folks as an assignment operator.

<snip>

> Is there a downside? The only one I can think of is that assignment
targets
> would not always appear at the left end of program lines!

The downside is that it makes Python code less conistent by creating two
ways to do the same thing ('=' and ':=').  Nonetheless, I am very much in
favor.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From henry_crun3583 at my-deja.com  Fri Dec 29 22:00:05 2000
From: henry_crun3583 at my-deja.com (henry_crun3583 at my-deja.com)
Date: Sat, 30 Dec 2000 03:00:05 GMT
Subject: why no "do : until"?
Message-ID: <92jj3j$ff7$1@nnrp1.deja.com>

Can someone point me at an explanation of the logic for not having
do:
until

or explain the elegant way to achieve this in Python.

eg
do :
    R=Themistor.ReadResistance()
    T=R_to_Temperature(R)
  until (T >=25)


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


From dale at out-think.NOSPAMco.uk  Mon Dec  4 11:44:48 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Mon, 04 Dec 2000 16:44:48 +0000
Subject: PythonWin
References: <3A2BAF44.178AC60@connection.com>
Message-ID: <6cin2tctn0kopvl45i5tmhdl7o18b1up5s@4ax.com>

"Colin J. Williams" <cjw at connection.com> wrote:

><!doctype html public "-//w3c//dtd html 4.0 transitional//en">
><html>
>Often, if an execution fails and re-execution is tried, one gets the message:
><p>&nbsp; File "g:\program files\python20\wxPython\wx.py", line 1646, in
>__init__
><br>TypeError: Only 1 wxApp per process!
><p>One can work around this by exiting PythonWin and returning.
><p>Is there a simpler way?
><p>Thanks
><p>Colin W.
><br>&nbsp;</html>
>
>

The problem is that PythonWin doesn't clean-up after running a program. This, presumably,  is so you
can inspect objects and variables for debugging. However, it's a bit troublesome in such situations.

The only solution I know is to run programs outside of PythonWin unless you are using the debugger.

--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants




From jadestar at idiom.com  Wed Dec 27 15:57:34 2000
From: jadestar at idiom.com (James T. Dennis)
Date: 27 Dec 2000 20:57:34 GMT
Subject: Kanji
Message-ID: <92dl3u$to0$1@news.idiom.com>

 I'd like to write an improved Kanji drill (kdrill) in 
 Python.  kdrill is a Linux program which displays Kanji
 (and Katakana/Hirigana) and their definitions/pronuciations 
 (from edict, the free "English Dictionary" from Japan).

 I have the necessary fonts installed.  But how can I access 
 them from Python?




From pehr at pehr.net  Wed Dec  6 00:52:07 2000
From: pehr at pehr.net (pehr anderson)
Date: Wed, 06 Dec 2000 05:52:07 GMT
Subject: setsockopt and urllib
References: <slrn92moue.hc.frank.neuber@mars.private.de>
Message-ID: <3A2DD408.D4E0B8E3@pehr.net>

Dear Frank, 

Timo has written an modification to sockets that 
adds user-controlled timeouts. This should meet
your needs.  You just import the module and, well
read the documentation in timeoutsocket.py

Timo is usually happy to provide help and feedback
on the use of his code. 
His email address is also int he source file.

	-pehr

http://www.timo-tasi.org/python/



Frank Neuber wrote:
> 
> Hi group,
> I have to set special timeouts on receive and send calls in conjunction with
> urllib-calls.
> Is there a simple solution to do this without modifying the urllib itself.
> For Example:
> I wish to set the timeout to 60 s for the following command:
> (tmpfile, hdr) = urllib.urlretrive(urlname)
> It is possible to to this with the socks module call
> setsockopt(SOL_SOCKET,SO_RCVTIMEO, ttime) over the urllib module.
> 
> thanks in advance
> Frank
> 
> --
>      _/_/_/_/ _//   _/ Frank Neuber
>     _/       _/_/  _/  frank.neuber at gmx.de (private)
>    _/_/_/   _/ _/ _/
>   _/       _/  _/_/    neuber at opensource-systemberatung.de
>  _/       _/    // http://www.opensource-systemberatung.de


From aleaxit at yahoo.com  Tue Dec 19 12:01:36 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 19 Dec 2000 18:01:36 +0100
Subject: sys.stdout.write()
References: <r3M%5.159$LU6.61501@typhoon.aracnet.com>
Message-ID: <91o4ac01iqj@news2.newsguy.com>

"Daniel Klein" <DanielK at aracnet.com> wrote in message
news:r3M%5.159$LU6.61501 at typhoon.aracnet.com...
> In 'Learning Python' by Mark Lutz and David Ascher, on page 77 it makes
> mention of the 'sys.stdout.write' method. However, when I check this out,
I
> don't see the 'write' method in the list...
>
> >>> import sys
> >>> dir(sys.stdout)
> ['shell', 'softspace', 'tags']
>
> I must be doing something fundamentally wrong. :^(

This is what I see when I run Python in a commandbox:

D:\PySym>python
Python 2.0 (#8, Dec 18 2000, 13:10:05) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
>>> import sys
>>> dir(sys.stdout)
['close', 'closed', 'fileno', 'flush', 'isatty', 'mode', 'name', 'read',
'readin
to', 'readline', 'readlines', 'seek', 'softspace', 'tell', 'truncate',
'write',
'writelines']
>>>

You may be using some other environment, where sys.stdout refers
to an object that's not a builtin file object; it will surely
have a write method, but it need not define it itself -- it can
delegate.  dir() does not consider such delegation, only the
dictionary of the object itself.

For example, your result is what I also get if I run this in
IDLE.


> I guess this brings up the bigger question of self help. Is there an easy
or
> prescribed way to find out what module (or class) a particular method is
> implemented in?

The Index and Search tabs of the (HTML Help version of) Python 2
docs are pretty good that way.  But, if you mean 'by reflection'...:

>>> w=getattr(sys.stdout,'write')
>>> w
<built-in method write of file object at 007C22F0>

in the Python at the command line; in IDLE, though:

>>> w=getattr(sys.stdout,'write')
>>> w
<method PseudoFile.write of PseudoFile instance at 00F47CEC>

Exploring further in IDLE we see:

>>> type(sys.stdout)
<type 'instance'>
>>> sys.stdout.__class__
<class PyShell.PseudoFile at 00F0F03C>
>>> dir(sys.stdout.__class__)
['__doc__', '__init__', '__module__', 'flush', 'isatty', 'write',
'writelines']
>>> sys.stdout.__class__.__bases__
()

which is about it -- it's pretty easy to automate this
'browsing', of course, but doing it by hand can be quite
educational.


> As a side (but related) issue, the current online reference material does
> not always indicate all of the exceptions thrown by a method or function.
> What is the best way to determine this when writing code?

Now THAT is something I think you can't get by reflection -- the
docs WOULD have to provide it, I think.


Alex





From hs at paradise.nirvananet  Mon Dec 25 19:19:19 2000
From: hs at paradise.nirvananet (Hartmann Schaffer)
Date: 25 Dec 2000 19:19:19 -0500
Subject: variable scope?
References: <c5S16.8616$KY1.23318@news1.rivrw1.nsw.optushome.com.au>
Message-ID: <928o67$9qj$1@paradise.nirvananet>

In article <c5S16.8616$KY1.23318 at news1.rivrw1.nsw.optushome.com.au>,
deadmeat <root@[127.0.0.1]> wrote:
>This doesn't work:

it's all in the manual

>---
>import time, string
>
>month = "2"
>year = "2000"
>
>def foo():

    global year
    # unless you insert this declaration, an assignment to year introduces
    # a new variable named "year" in function "foo".  when you try to read
    # it after having skipped the assignment you'll get the message below

>   if string.atoi(month) <= 3:
>      year = `string.atoi(year) - 1`
>      print year

    # i suspect in your actual code the print statement wasn't indentes
    # as shown hear

>
>foo()
>---
>
>reporting:
>
>Traceback (most recent call last):
>  File "pfft.py", line 11, in ?
>    foo()
>  File "pfft.py", line 8, in foo
>    year = `string.atoi(year) - 1`
>UnboundLocalError: Local variable 'year' referenced before assignment
>
>but this does
>
>---
>import time, string
>
>month = "2"
>year = "2000"
>
>def foo():
>   year = "2000"
>   if string.atoi(month) <= 3:
>      year = `string.atoi(year) - 1`
>      print year
>
>foo()
>---
>
>how do I force foo() to use the global 'year' variable?

hs


From fpm at u.washington.edu  Fri Dec 29 10:52:15 2000
From: fpm at u.washington.edu (Frank Miles)
Date: 29 Dec 2000 15:52:15 GMT
Subject: identifying user from Win9x?
Message-ID: <92ibvf$if2$1@nntp6.u.washington.edu>

I have a set of python routines that are used to coordinate and
account for various activities in my group.  I would like to be
able to tailor how these routines work, and what options are first
presented to each user.

There are ways to determine the user ID in Unix-based systems (os.getuid).
I haven't yet seen anything comparable for Windows.  Possible source
of information: it is SOP for each user to login to the (Novell-based)
network before doing anything else.  Is there some way to extract this
information from Windows, or failing that, a Novell network/server?

TIA...

	-frank

-- 


From joconnor at cybermesa.com  Tue Dec 12 12:21:52 2000
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Tue, 12 Dec 2000 17:21:52 +0000
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
Message-ID: <3A365EB0.C3E02AA8@cybermesa.com>

> > The real problem is that many Python tutorials fail to explain
> > how things work (or hide it somewhere in the exercises), but
> > that's not really Guido's fault...
>
> Tough for the new Python programmer, then.  It's nt fair to them.

Actually, it's not really that hard for the new programmer, but maybe a
little odd for a programmer coming from certain languages.  I've been
programming Python seriously for about 3 months and had no difficulty
with the concept. But I've been programming  Smalltalk for a long time,
which uses a similar approach (as does Java)

The requirement to know about such things in C and Perl really annoy me,
especially Perl which is new enough to have been designed without such
artifacts 

> While I'm brainstorming with Python grammar changes, what about a real &
> operator.  No more of this hokey is-it-modifiable-in-place-or-not
> business.  Explicit control is important.

Not really.  It just confuses the issue by adding one more piece of
syntax for he developer to remember and one more piece of un-needed
detail for the new programmer to learn.



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

"God himself plays the bass strings first when He tunes the soul"



From mwh21 at cam.ac.uk  Sun Dec 10 13:20:10 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 10 Dec 2000 18:20:10 +0000
Subject: Related to os module
References: <m3k898yxaj.fsf@boss.dreamsoft.com>
Message-ID: <m3y9xo2l5x.fsf@atrus.jesus.cam.ac.uk>

Venkatesh Prasad Ranganath <rvprasad at cis.ksu.edu> writes:

> Hi,
> 
> Following is the code in question
> 
> >>>import os
> >>>os.execv("/bin/ls", ["-l"])

Try os.execv("/bin/ls", ["ls","-l"]), or alternatively reading section
6.1.5 of the library reference....

Cheers,
M.

-- 
  I'm okay with intellegent buildings, I'm okay with non-sentient
  buildings. I have serious reservations about stupid buildings.
     -- Dan Sheppard, ucam.chat (from Owen Dunn's summary of the year)


From root at rainerdeyke.com  Thu Dec 14 14:42:26 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Thu, 14 Dec 2000 19:42:26 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <r0pzoi11lhz.fsf@black.cs.washington.edu> <5jvZ5.3588$Kd1.418770@newsb.telia.net> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com> <kuu28971kp.fsf@lasipalatsi.fi> <3A380E4E.8537A0FB@san.rr.com> <91a1vr02b02@news2.newsguy.com>
Message-ID: <Co9_5.25465$x6.13756579@news2.rdc2.tx.home.com>

"Alex Martelli" <aleaxit at yahoo.com> wrote in message
news:91a1vr02b02 at news2.newsguy.com...
> There is no value *IN* x.
>
> A variable *REFERS TO* 'a value' (aka 'an object').

If somebody really wants mutable integer objects, they're not that hard to
create.

class MutableInt(UserInt.UserInt):
  def set(self, value):
    self.data = value

x = MutableInt(2) + MutableInt(2)
print x
x.set(5)
print x


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From joconnor at cybermesa.com  Sat Dec 23 18:53:30 2000
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Sat, 23 Dec 2000 23:53:30 +0000
Subject: Some Python 2.1 ideas
Message-ID: <3A453AFA.8F422671@cybermesa.com>

>It's my guess that one reason Python
>doesn't support this is to make "=" illegal inside expressions, thus
>avoiding the "=" vs. "==" errors common in C. A response to that objection
>is to use a different token for embedded assignment 

Another response to that objection would be that languages that support
a proper Boolean type do not exhibit errors based on "=" versus "=="


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

"God himself plays the bass strings first when He tunes the soul"



From urner at alumni.princeton.edu  Tue Dec 19 16:06:05 2000
From: urner at alumni.princeton.edu (Kirby Urner)
Date: Tue, 19 Dec 2000 13:06:05 -0800
Subject: Python in "math through programming" curriculum
References: <2g2t3t06r2agrrjmfnr41eu4lc2u39a8qi@4ax.com> <3A3EFF47.67DE69CF@gte.net> <513v3toslkjkggh9oumlf20qc9bi1gdtif@4ax.com> <3A3FC677.FDB19937@gte.net>
Message-ID: <rviv3tglqs07o0ni5c8e64gelarcc33qvt@4ax.com>

"Clifford J. Nelson" <cnelson9 at gte.net> wrote:

>I'm sure I don't attach ANY weight to any of your opinions.

Ooooo, touchy touchy...

>I've seen your web sites about Synergetics and programming!
>
>  Cliff Nelson

For the curious, that's:
http://www.inetarena.com/~pdx4d/ocn/numeracy0.html and
http://www.teleport.com/~pdx4d/synhome.html

Anyhow, enough people do (attach weight) that I can easily 
afford to discount your negative assessment -- it's a 
selling point in fact, that this Cliff Nelson character 
(whoever he is) doesn't like my green eggs and ham.

Kirby




From SBrunning at trisystems.co.uk  Wed Dec 13 06:13:57 2000
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Wed, 13 Dec 2000 11:13:57 -0000
Subject: Python and databases
Message-ID: <31575A892FF6D1118F5800600846864D5B15DC@intrepid>

> From:	Kevin B [SMTP:kbass1 at nc-rr.com]
> Are there any resources that have information about Python and database
> access?
 
"Python Programming on Win32" (Hammond and Robinson, O'Reilly,
1-56592-621-8) has a chapter on DB access, much of which isn't Win32
specific.

Also, you could take a look at <http://www.python.org/sigs/db-sig/>.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk





-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From rwklee at home.com  Thu Dec 14 15:54:10 2000
From: rwklee at home.com (Rick Lee)
Date: Thu, 14 Dec 2000 20:54:10 GMT
Subject: Tkinter prob in Python 2.0, Macintosh
References: <muvol.remove-this-ABD013.16350709122000@news.knology.net> <9130bq$cnsu$1@nntp6.u.washington.edu> <muvol.remove-this-F79F6B.22235713122000@news.knology.net>
Message-ID: <3A393372.30DF6A1B@home.com>

I don't program with Tkinter directly, but I have been using other Python
programs that use Tkinter, such as redemo.py.  The Tkinter window comes up OK,
and everything seems to work properly.  There are always 2 windows that  come
up: a Tkinter window (depending on the application, I guess there can be
multiple), and a Python interpreter window.  To terminate Python, first I have
to quit in the Tkinter window; which sends a terminate signal to the Python
Interpreter window; I then have to quit the Python window.  Occasionally, not
knowing how to reproduce this problem, I have the entire Apple menu duplicated
again, to the right of the original menu; if this happens, it is the new menu
that works, the old one seems to have no effect.  I never bothered to track this
down.

- Rick Lee

Mike wrote:

> "Russell E. Owen" <owen at astroNOJNK.washington.edu.invalid> wrote:
> > In article <muvol.remove-this-ABD013.16350709122000 at news.knology.net>,
> > Mike <muvol.remove-this at knology.net> wrote:
> >
> > >...The problem occurs when I press the Quit key.  The Tk window goes away,
> > >but
> > >I do not get the >>> prompt back in the interpreter.  Even though I can
> > >still select the menus, choosing Quit from the file menu has no effect.
> > >Also, Command-Q does not work.  At this point, the About box has changed
> > >from "About Python..." to "About Tcl & Tk"
> > >
> > >I have to force quit the Python Interpreter to regain control.
> >
> > Try de-selecting the startup option "Enter interactive mode after
> > script". This problem was present in 1.5.2c and based on your
> > description it may well still be present in 2.0. (For a one-time test,
> > run the interpreter with the option key down. For a longer-term
> > solution, run EditPythonPrefs.)
> >
> > Warning: Tk is not fully functional in MacPython 2.0 -- file events are
> > broken. Stick to MacPython 1.5.2 for now if you need file events.
> >
> > -- Russell
>
> Thank you for the suggestions.  I tried them, but my problem persists.
>
> Maybe I am asking the wrong question.  Is anyone else using Tkinter, Python
> 2.0 on the Mac?  If so, how do you get it to work, and what is the best way
> to develop code?  Or is there a better way to develop cross platform Python
> GUI apps on the Mac?
>
> I would appreciate any suggestions.
>
> Thanks,
> Mike
> ---



From ullrich at math.okstate.edu  Sun Dec 31 11:28:30 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Sun, 31 Dec 2000 16:28:30 GMT
Subject: how do i check if a file exists with python?
References: <3a4c95dc.88792787@nntp.sprynet.com>, <fBN26.99045$yR4.2640173@news1.rdc1.tx.home.com>, <RlM26.98766$yR4.2639444@news1.rdc1.tx.home.com> <92g5v8$r5k$1@nnrp1.deja.com> <mailman.978078157.26485.python-list@python.org> <mailman.978264789.28740.python-list@python.org> <3a4f2bd4.45424534@nntp.sprynet.com>
Message-ID: <3a4f5c76.57876580@nntp.sprynet.com>

On Sun, 31 Dec 2000 13:10:30 GMT, ullrich at math.okstate.edu (David C.
Ullrich) wrote:

>On Sun, 31 Dec 2000 17:39:21 +0200 (IST), Moshe Zadka
><moshez at zadka.site.co.il> wrote:
>
>>On Fri, 29 Dec 2000 14:02:25 GMT, ullrich at math.okstate.edu (David C. Ullrich) wrote:
>[...]
>>
>>> I sometimes make cumulative backups of files containing
>>> important stuff. I sort of need to determine that file.001,
>>> file.002 and file.003 already exist and file.004 does not
>>> before deciding what to name the current backup file.
>>
>>That's *really* evil. What if there is another version of your
>>process running???  Damn!, just lost that backup file. In that
>>case, I'd go for a Maildir-like algorithm, which promises no colisions.
>>Actually, in that case, usually what you want is CVS, but that's
>>a matter for another day.
>
>[...] Is it possible
>to give a hint how a "Mailder-like" algorithm
>promises no collisions?

Oh. I thought that Maildir was like the name of the guy who invented
the algorithm. It's Maildir, like a dir where incoming mail goes.
Google.Bernstein says in part

"Why should I use maildir?

Two words: no locks. An MUA can read and delete messages while new
mail is being delivered: each message is stored in
a separate file with a unique name"

I'd considered the idea that each piece of input could become a
separate file but I thought there was probably a cleverer way to
do it. But if it's good enough for maildir it's probably good enough.

>DU



From aleaxit at yahoo.com  Mon Dec  4 18:06:33 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 5 Dec 2000 00:06:33 +0100
Subject: A better python interface for buffer objects
References: <3A2C1188.D384E47C@stsci.edu>
Message-ID: <90h8320n4r@news2.newsguy.com>

"Paul Barrett" <barrett at stsci.edu> wrote in message
news:3A2C1188.D384E47C at stsci.edu...
    [snip]
> confusing?)  Since I can't use a string object as a writeable buffer, is
> there some other object that I can use as a writeable buffer, not

Maybe an array.array object?  Haven't tested... it should, I guess.


Alex





From wware at world.std.com  Sat Dec  9 00:56:32 2000
From: wware at world.std.com (Will Ware)
Date: Sat, 9 Dec 2000 05:56:32 GMT
Subject: making a DLL
Message-ID: <G5ADuH.GC8@world.std.com>

This is admittedly a pretty dumb question, please bear with me, I'm
not a Windows programmer. I've often made shared objects in Unix to
speed up Python with the addition of code written in C. Now I want to
do the equivalent thing in Windows. So I've written a dumb example
C++ file, and put together an apparently adequate .def file, and told
MS VC++ 5.0 to build a DLL. I even figured out how to set up the
include directory (C:\Python20\include) so that it doesn't have any
trouble finding Python.h. So far so good. Alas, when I try to link, I
get this:

Linking...
LINK : fatal error LNK1104: cannot open file "python20_d.lib"
Error executing link.exe.

So I searched the directory tree starting at c:\Python20 for *.lib
files, and found c:\Python20\libs\python20.lib, but no python20_d.lib
anywhere. So I thought to try forcing the thing to use python20.lib
instead (Project->Settings->Link(input)->Object/library modules) and
gave it the library path. Now I get a different error:

Linking...
c:\Python20\libs\python20.lib : fatal error LNK1106: invalid file or \
               disk full: cannot seek to 0x39ed0908
Error executing link.exe.

I know the disk isn't full, so I conclude that python20.lib is somehow
deemed invalid. Was the Windows version of Python 2.0 built in some way
that Microsoft Visual C++ would consider invalid? (E.g., with Borland's
compiler?) Is there a fix for all this mess? Any hints appreciated, thanks.

-- 
# - - - - - - - - - - - - - - - - - - - - - - - -
# Resistance is futile. Capacitance is efficacious.
# Will Ware	email:    wware @ world.std.com


From chris at hddesignNOSPAM.com  Fri Dec  8 15:52:25 2000
From: chris at hddesignNOSPAM.com (Chris Meyers)
Date: Fri, 8 Dec 2000 14:52:25 -0600
Subject: xml.dom
References: <90q292$k9k$1@merlin.rembrandtstr.bocholt.de> <3A3100FD.D6C09C1D@ogbuji.net> <90raqe$2f6$1@merlin.rembrandtstr.bocholt.de>
Message-ID: <90rhkp$35l$1@news.chorus.net>

> Are there any good books to recommend in this area?

I am just starting to think about working with XML, so I don't know much
yet, but my boss has loaned me a book called "XML Processing with Python" by
Sean McGrath published by Prentice Hall (ISBN 0-13-021119-2). It seems
pretty good so far (I've only hit the first couple chapters)

Good luck to you and me both.
Chris

"I don't want the World... I just Want your half" TMBG




From alan.gauld at gssec.bt.co.uk  Wed Dec  6 09:10:45 2000
From: alan.gauld at gssec.bt.co.uk (Alan Gauld)
Date: Wed, 06 Dec 2000 14:10:45 +0000
Subject: "inherited" keyword in Python?
References: <3A2B9EFA.3A64C13B@schlund.de>
Message-ID: <3A2E48E5.7D293DE1@gssec.bt.co.uk>

Carsten Gaebler wrote:
> i.e. instead of saying "superclass.method()" you say "inherited
> method()". 

Mostly they are equivalent but they are not identical.
ISTR inherited can only be used on virtual methods. 
Static methods of the superclass must be called using 
the first style. 

Also inheritred can be used for message handling methods
(where the superclass's handler may not have the same name 
as the method in the subclass...)

The OP language reference says of 'inherited':
-------------------------
Within a message-handler method, you can call the message-handler
inherited 
from the object's ancestor type. Because the name and parameter
of the 
method might vary, you can call the inherited method just by
using inherited; 
you do not have to include the ancestor method's name. 

For example, the WMPaint method described above might be
implemented 
as follows:

procedure TMyControl.WMPaint(var Message: TWMPaint);
begin
  with Message do begin
    ...
    inherited;
    ...
  end;
end;

The inherited method called will be the one with the same message
index 
(in this case, WM_PAINT). For example, if the method declared
above, TMyOtherControl.PaintIt, included an inherited statement,
the method 
called would be TMyControl.WMPaint.

Default message handler

It is always safe to call inherited within a message-handler
method. 
If the ancestor type does not declare a specific message handler
for 
a particular message index, inherited calls the TObject method 
DefaultHandler.
---------------------------------


> Can be quite helpful if you change the name of the superclass

It can be useful for that but that's not really its purpose.
You can fake this behaviour in Python by defining a local 
'hidden' variable (__inherited, say) and calling that:

class foo(Bar):
   __inherited = Bar  # the '__' hides it
   def spam(self):
      __inherited.spam(self)



Alan G.

-- 
=================================================
This post represents the views of the author 
and does not necessarily accurately represent 
the views of BT.


From gregj at pobox.com  Wed Dec 20 01:18:40 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Wed, 20 Dec 2000 06:18:40 GMT
Subject: while true: !!!
References: <mailman.977268243.15476.python-list@python.org> <NnS%5.482$Qb7.42792@newsb.telia.net>
Message-ID: <4bY%5.172261$U46.5444054@news1.sttls1.wa.home.com>

"Fredrik Lundh" <fredrik at effbot.org> wrote in message
news:NnS%5.482$Qb7.42792 at newsb.telia.net...
> "Delaney, Timothy" wrote:

> so you prefer slow and fragile code over fast and robust
> code.  fine.  now please shut up, and let this thread die.

This thread is a great example of what newsreader filters are good for.
Obviously quite a few of us are getting something out of it.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From syver at NOSPAMcyberwatcher.com  Wed Dec 13 16:29:15 2000
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Wed, 13 Dec 2000 22:29:15 +0100
Subject: class win32com.client.genpy.DispatchItem collection method generation
Message-ID: <918ps7$u47$1@troll.powertech.no>

I am currently knee-deep in class DispatchItem working on getting it to
generate __getitem__ and __setitem__ methods for makepy generated classes..

Generated __getitem__ seems to work fine, (passes the testmakepy) test. I
have more problems with __setitem__ because I haven't found any code
describing how it should look except in dynamic.py.

Any hint on how to generate the code for __setitem__ much appreciated.




From thomas at xs4all.net  Sat Dec 23 12:43:57 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Sat, 23 Dec 2000 18:43:57 +0100
Subject: Why Won't This Work?
In-Reply-To: <LOBBJMHOIFOLFIPFANHJKEFGCFAA.lists@TheWebsons.com>; from lists@TheWebsons.com on Sat, Dec 23, 2000 at 06:31:38AM -0800
References: <LOBBJMHOIFOLFIPFANHJKEFGCFAA.lists@TheWebsons.com>
Message-ID: <20001223184357.G5515@xs4all.nl>

On Sat, Dec 23, 2000 at 06:31:38AM -0800, Ben Ocean wrote:

> Because I'm building Zope, I'm trying to install Python-1.5.2. I've
> successfully run:
> ./configure --with-thread
> make
> and everything appears to be in order. But when I type
> python
> at a command prompt, I get a bash error. What have I not done?

That's because the current directory ('.') isn't in your PATH. Try typing
'./python' instead. Note that the best way to test your python binary is
'make test' -- it'll catch practically everything you can try to catch
yourself, and a lot more to boot.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From mikael at isy.liu.se  Wed Dec 20 03:06:55 2000
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Wed, 20 Dec 2000 09:06:55 +0100 (MET)
Subject: Wanted: Python White Papers
In-Reply-To: <91oj8v$i0r$1@nnrp1.deja.com>
Message-ID: <XFMail.001220090655.mikael@isy.liu.se>

On 19-Dec-00 noahspurrier at my-deja.com wrote:
 >  Actually, the only high-profile company that uses Python that I
 >  can think of is Google. Is anyone familiar with their architecture?
 >  How does Python fit into their design?

To my knowledge, both Yahoo and eGroups use Python.

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    20-Dec-00
Time:    09:02:37

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

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


From owen at astroNOJNK.washington.edu.invalid  Fri Dec 15 13:03:21 2000
From: owen at astroNOJNK.washington.edu.invalid (Russell E. Owen)
Date: Fri, 15 Dec 2000 10:03:21 -0800
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <3a36c85e.4409136@News.CIS.DFN.DE> <owen_nospam-1312001131590001@cs204-90.fsmodem.washington.edu> <3A37DD31.C9A5F18C@pacbell.net> <91bhlh$c64u$1@nntp6.u.washington.edu> <91d7270niu@news1.newsguy.com>
Message-ID: <91dmda$89ei$1@nntp6.u.washington.edu>

In article <91d7270niu at news1.newsguy.com>, "Alex Martelli" 
<aleaxit at yahoo.com> wrote:

>> Smalltalk is my favorite language. However, I am the only one at the
>> observatory who knows it and feels it is worthwhile. As I fight Python's
>> disgusting GUI I keep wavering, and I may yet give up and switch to
>
>Which one of Python's many available GUIs do you find disgusting?  Me,
>I like wxPython, and HTML-based interfaces, but I realize Tkinter's
>more widespread -- and there are just SO many others that, if I was
>GUI-oriented at all, I'd be in a tizzy about trying them all...

Tkinter. Lack of built-in cut/copy/paste is a problem (trying to 
implement it well is a real bear), similarly drag-and-drop and read-only 
text displays that allow updating and copying. Appearance is OK. Widget 
set is reasonable. Menus don't seem to work reliably on the Mac, and in 
MacPython 2.0 file events are broken.

I need something that works, or will work, on Mac and unix for sure, 
Windows would also be good. The only Python GUI library that does that 
now is Tkinter, but wxPython is likely to do so the future. Are there 
any others?

I am trying to get wxPython installed on a Solaris box and will try it 
out if I'm ever successful.

(The Mac issue is also why IBM's smalltalk is out, unless they're 
planning a port to MacOS X. And whoever said Cincom's GUI is slow on a 
Mac, I agree, and unattractive, too. I didn't realize it was fast on 
other platforms. That's very good to know.)

-- Russell


From fluxent at yahoo.com  Wed Dec  6 13:23:16 2000
From: fluxent at yahoo.com (Bill Seitz)
Date: Wed, 06 Dec 2000 18:23:16 GMT
Subject: CGI File Upload
References: <p8ce0tcgtt2bksb3adrf9cfr2i8egr4ff5@4ax.com> <8u7fn7$jpf$1@la-mail4.digilink.net> <90j10o$5p5$1@nnrp1.deja.com> <ulmttsihs.fsf@cs.uu.nl>
Message-ID: <90m06a$kg1$1@nnrp1.deja.com>

In article <ulmttsihs.fsf at cs.uu.nl>,
  piet at cs.uu.nl wrote:
> >>>>> Bill Seitz <fluxent at yahoo.com> (BS) writes:
>
> BS> Hmm, this works for me with text data (though the file length is
> BS> slightly different: maybe line-end changes from Win to *nix?).
With
> BS> binary it's totally off. Uploading a JPEG I consistently only get
109
> BS> bytes written to the file system. With GIFs I consistently get
~500
> BS> bytes (but not always exactly the same number). With an MsWord
file I
> BS> get a bigger file written, but still not the whole thing (60kb
out of
> BS> 75kb).
>
> BS> What do I do next?
>
> Is your server on Windows? Then it probably is a matter of the
standard
> input being in text mode rather than binary mode.
> So you have to set it to binary mode before anything is read from it.
>
> See msvcrt.setmode()

Yes, this turned out to be the issue. I had it set for stdout, but not
stdin.

msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
msvcrt.setmode(sys.stdin.fileno(), os.O_BINARY)


Sent via Deja.com http://www.deja.com/
Before you buy.


From rgeorgesprint at earthlink.net  Mon Dec  4 08:34:50 2000
From: rgeorgesprint at earthlink.net (robert george)
Date: Mon, 04 Dec 2000 13:34:50 GMT
Subject: Beginers question
Message-ID: <_3NW5.44729$nh5.3218436@newsread1.prod.itd.earthlink.net>

I started the beginers tutorial. My operating system is Winme.
The tutorial says I should get a" C:\WINDOWS> " prompt. But I get
a"C:\WINDOWS\DESKTOP>" prompt.
Why is that? Will Python work with Winme?
Thanks for your help
An extreme novice.
Rob




From rapto at my-deja.com  Mon Dec 18 14:09:56 2000
From: rapto at my-deja.com (rapto at my-deja.com)
Date: Mon, 18 Dec 2000 19:09:56 GMT
Subject: Stored Procedures (ODBC) from Python?
References: <91l6mq$lge$1@nnrp1.deja.com> <91l8va$qa$1@reader1.imaginet.fr>
Message-ID: <91lne0$4q3$1@nnrp1.deja.com>

The standard ODBC idiom for calling stored procedures (or saved
parameter queries in Access) is
{call MyStoredProc(arg1, arg2...)}

For example:

{call getOrderId('9/312')}

You may look for documentation for ODBC escape sequences. There are some
for dates, outer joins, concatenating strings, data type conversion,
parameter return, etc.

Using these you may get a bit more vendor-independent.


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


From prestonlanders at my-deja.com  Thu Dec 21 12:57:34 2000
From: prestonlanders at my-deja.com (Preston Landers)
Date: Thu, 21 Dec 2000 17:57:34 GMT
Subject: PyOpenGL and Nvidia GL?
References: <tmi%5.60958$df5.2143169@news1.crdva1.bc.home.com> <87k88xkmik.fsf@optonline.net>
Message-ID: <91tga9$gav$1@nnrp1.deja.com>

In article <87k88xkmik.fsf at optonline.net>,
  Les Schaffer <schaffer at optonline.net> wrote:

> you will see why i
> will never buy another video card from NVidia so long as their drivers
> are closed source.

Unfortunately with the recent demise of 3dfx you may have little option.
I don't know of any other manufacturer who has a decent open source
OpenGL driver.  Can you think of one?  I sure would like to know because
my Voodoo 3000 is looking older and less supported by the day.

--
|| Preston Landers <prestonlanders at my-deja.com> ||


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


From pbjorn at uswest.net  Tue Dec 19 09:39:50 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Tue, 19 Dec 2000 07:39:50 -0700
Subject: while true: !!!
References: <mailman.977127721.4398.python-list@python.org> <slrn93rsl2.qb2.grey@teleute.rpglink.com> <3A3E217C.5F2F4996@uswest.net> <slrn93sn65.tvg.grey@teleute.rpglink.com>
Message-ID: <3A3F7336.36F2330C@uswest.net>

Steve Lamb wrote:

> On Mon, 18 Dec 2000 07:38:52 -0700, Bjorn Pettersen <pbjorn at uswest.net> wrote:
> [snip[
> >   for line in f.readlines():
> >      ...
>     Which...
> >has the potential of crashing your machine with an out-of-memory error if you
> >don't know the sizes of files you're operating on.
>     ...wastes memory.

Well, if you have it, and use it, you're hardly wasting it <wink>.

> >So, the status quo is that because the straight-forward way of doing it is
> >too slow, we're all using an optimized idiom that while easy to understand if
> >coming from a CS background, doesn't model what you're doing (at least not
> >until you've programmed enough Python that reading a file seems like an
> >endless loop <wink>).
>
>     Or doing it in the 2-3 other languages I know of which use while in the
> exact same manner.  I'm sure if we took a pole of the computer languages known
> here you'd find that it is pretty much a standard across many languages, not
> just Python.

Thank you for restating what I said: coming from a CS background/other languages,
"while 1:" is quite understandable. I don't think this is necessarily an area
worth mimicking other languages though (especially since one of the main arguments
for adding += to the language was because it closer modelled the intended semantic
of the expression).

>     Hell, if you want to grouse about readability in Python, why not attack
> some of the more obvious targets?
> [snip examples of else clauses on for and while loops]

So you're arguing that since Python is so unreadable in it's use of else clauses,
"while 1:" hardly matters? <wink>

-- bjorn

ps: the only questions I have seen about "else on while" have indicated that
beginners expect the else clause to be executed if the condition fails the first
time (I could probably come up with a coherent argument for why that is the
"negative" of a loop <wink>)
pps: I did enjoy the "while 1: ... else: ..." construct ;-)




From fluxent at yahoo.com  Tue Dec  5 12:28:10 2000
From: fluxent at yahoo.com (Bill Seitz)
Date: Tue, 05 Dec 2000 17:28:10 GMT
Subject: CGI POST binary file upload?
Message-ID: <90j8j0$cmt$1@nnrp1.deja.com>

I've tweaked some code which someone else posted a while back, to allow
a CGI (behind Apache on Win2K) accept a POST with a file upload, and
write it out to the file tree.

It works fine for text/html files (though the byte count is slightly
off: line endings?). But for binary files it's a total mess. For GIFs I
generally only get ~100 bytes, for JPEGs ~500bytes. MsWord and PDF
files also don't work, though I usually get a bigger end result.

I don't think the problem is the file writing, it's the
receiving/reading. I say this because I also output in my response page
len(data['image'].value)
and it agrees with the size of the output file.

I tried adding
msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
up at the top of my code, but (a) it doesn't seem to matter, and (b)
I'm not sure whether I'm doing it right.

thx


Sent via Deja.com http://www.deja.com/
Before you buy.


From harisri at my-deja.com  Wed Dec  6 22:30:22 2000
From: harisri at my-deja.com (harisri at my-deja.com)
Date: Thu, 07 Dec 2000 03:30:22 GMT
Subject: Newbie question on 'cmpfiles':
References: <90i074$d73$1@nnrp1.deja.com> <JmeX5.39553$II2.3513820@newsread2.prod.itd.earthlink.net> <sZnX5.153284$e5.99945@newsfeeds.bigpond.com> <53AX5.51696$nh5.3969094@newsread1.prod.itd.earthlink.net>
Message-ID: <90n08e$gf9$1@nnrp1.deja.com>

In article <53AX5.51696$nh5.3969094 at newsread1.prod.itd.earthlink.net>,
  ak at silmarill.org wrote:
> On Wed, 6 Dec 2000 21:05:31 +1100, Srihari Vijayaraghavan
<harisri at bigpond.com> wrote:
> >Oldayz wrote:
> >
> >> On Tue, 05 Dec 2000 05:59:02 GMT, harisri at my-deja.com
> >> <harisri at my-deja.com> wrote:
>
> >Hello Andrei,
> >
> >Thanks. However I am having some problem while using "dircmp" class.
> >
> >I have the following 2 (identical) directories under my home
directory:
> >/home/hari/python1 and /home/hari/python2.
> >
> >I am getting few error messages, only while using the following
attributes:
> >diff_files, report(), report_partial_closure(), and
report_full_closure().
> >
> >I am enclosing the error messages for your reference. Am I doing
something
> >wrong?
> >
> >Python 2.0 (#1, Oct 16 2000, 18:10:03)
> >[GCC 2.95.2 19991024 (release)] on linux2
> >Type "copyright", "credits" or "license" for more information.
> >>>> from filecmp import *
> >>>> x=dircmp("/home/hari/python1", "/home/hari/python2")
> >>>> x.left_list
> >['about.html', 'acks.html', 'api', 'dist', 'doc', 'ext', 'icons',
> >'index.html',
> >'inst', 'lib', 'mac', 'modindex.html', 'ref', 'tut']
> >>>> x.right_list
> >['about.html', 'acks.html', 'api', 'dist', 'doc', 'ext', 'icons',
> >'index.html',
> >'inst', 'lib', 'mac', 'modindex.html', 'ref', 'tut']
> >>>> x.common
> >['ref', 'ext', 'acks.html', 'inst', 'api', 'mac', 'about.html',
> >'modindex.html', 'tut', 'doc', 'icons', 'lib', 'dist', 'index.html']
> >>>> x.left_only
> >[]
> >>>> x.right_only
> >[]
> >>>> x.common_dirs
> >['ref', 'ext', 'inst', 'api', 'mac', 'tut', 'doc', 'icons', 'lib',
'dist']
> >>>> x.common_files
> >['acks.html', 'about.html', 'modindex.html', 'index.html']
> >>>> x.same_files
> >Traceback (most recent call last):
> >  File "<stdin>", line 1, in ?
> >  File "/usr/local/lib/python2.0/filecmp.py", line 147, in
__getattr__
> >    self.phase3()
> >  File "/usr/local/lib/python2.0/filecmp.py", line 214, in phase3
> >    xx = cmpfiles(self.left, self.right, self.common_files)
> >  File "/usr/local/lib/python2.0/filecmp.py", line 288, in cmpfiles
> >    res[_cmp(ax, bx, shallow, use_statcache)].append(x)
> >TypeError: too many arguments; expected 2, got 4
> >>>> x.report()
> >diff /home/hari/python1 /home/hari/python2
> >Traceback (most recent call last):
> >  File "<stdin>", line 1, in ?
> >  File "/usr/local/lib/python2.0/filecmp.py", line 241, in report
> >    if self.same_files:
> >  File "/usr/local/lib/python2.0/filecmp.py", line 147, in
__getattr__
> >    self.phase3()
> >  File "/usr/local/lib/python2.0/filecmp.py", line 214, in phase3
> >    xx = cmpfiles(self.left, self.right, self.common_files)
> >  File "/usr/local/lib/python2.0/filecmp.py", line 288, in cmpfiles
> >    res[_cmp(ax, bx, shallow, use_statcache)].append(x)
> >TypeError: too many arguments; expected 2, got 4
> > >>> x.report_partial_closure()
> >diff /home/hari/python1 /home/hari/python2
> >Traceback (most recent call last):
> >  File "<stdin>", line 1, in ?
> >  File "/usr/local/lib/python2.0/filecmp.py", line 258, in
> >report_partial_closure
> >    self.report()
> >  File "/usr/local/lib/python2.0/filecmp.py", line 241, in report
> >    if self.same_files:
> >  File "/usr/local/lib/python2.0/filecmp.py", line 147, in
__getattr__
> >    self.phase3()
> >  File "/usr/local/lib/python2.0/filecmp.py", line 214, in phase3
> >    xx = cmpfiles(self.left, self.right, self.common_files)
> >  File "/usr/local/lib/python2.0/filecmp.py", line 288, in cmpfiles
> >    res[_cmp(ax, bx, shallow, use_statcache)].append(x)
> >TypeError: too many arguments; expected 2, got 4
> >>>> x.report_full_closure()
> >diff /home/hari/python1 /home/hari/python2
> >Traceback (most recent call last):
> >  File "<stdin>", line 1, in ?
> >  File "/usr/local/lib/python2.0/filecmp.py", line 264, in
> >report_full_closure
> >    self.report()
> >  File "/usr/local/lib/python2.0/filecmp.py", line 241, in report
> >    if self.same_files:
> >  File "/usr/local/lib/python2.0/filecmp.py", line 147, in
__getattr__
> >    self.phase3()
> >  File "/usr/local/lib/python2.0/filecmp.py", line 214, in phase3
> >    xx = cmpfiles(self.left, self.right, self.common_files)
> >  File "/usr/local/lib/python2.0/filecmp.py", line 288, in cmpfiles
> >    res[_cmp(ax, bx, shallow, use_statcache)].append(x)
> >TypeError: too many arguments; expected 2, got 4
> >
> >Thanks in advance,
> >
> >Hari.
>
> According to
>
http://sourceforge.net/bugs/?group_id=5470&func=detailbug&bug_id=124120
> this is a known bug and it can be fixed by applying the latest patch,
which
> you can get at
>
http://sourceforge.net/patch/?func=detailpatch&patch_id=102628&group_id=
5470
>
> --
> Hope this helps,
>
> 	Andrei
>
Hello,

Thank you Andrei, the patch made it to work as expected.
I highly appreciate your help on solving this issue.

Hari.


Sent via Deja.com http://www.deja.com/
Before you buy.


From emile at fenx.com  Sun Dec 24 13:54:54 2000
From: emile at fenx.com (Emile van Sebille)
Date: Sun, 24 Dec 2000 10:54:54 -0800
Subject: soundex (revisited)
References: <Var16.263$LU6.109277@typhoon.aracnet.com>
Message-ID: <925gqa$606tp$1@ID-11957.news.dfncis.de>

"Daniel Klein" <DanielK at aracnet.com> wrote in message
news:Var16.263$LU6.109277 at typhoon.aracnet.com...
> After seeing the post from several days ago on soundex, I
gave it whirl to
> see if I could come up with something different (and
possibly better),
> following the rules laid down by Knuth:
>
<snip>
> One question: how is it that this understands the
'upper()' function without
> explicitly importing the 'string' module?
>
> Dan:
>
>


In 2.0, strings now have methods:


Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)]
on win32
Type "copyright", "credits" or "license" for more
information.
IDLE 0.6 -- press F1 for help
>>> print dir("")
['capitalize', 'center', 'count', 'encode', 'endswith',
'expandtabs', 'find', 'index', 'isalnum', 'isalpha',
'isdigit', 'islower', 'isspace', 'istitle', 'isupper',
'join', 'ljust', 'lower', 'lstrip', 'replace', 'rfind',
'rindex', 'rjust', 'rstrip', 'split', 'splitlines',
'startswith', 'strip', 'swapcase', 'title', 'translate',
'upper']
>>>



--

Emile van Sebille
emile at fenx.com
-------------------





From cjensen at bioeng.ucsd.edu  Wed Dec 27 14:36:10 2000
From: cjensen at bioeng.ucsd.edu (Curtis Jensen)
Date: Wed, 27 Dec 2000 11:36:10 -0800
Subject: newbie : array initialization, anyone ?
References: <3a4a1352.9818548@news.t-online.de>
Message-ID: <3A4A44AA.D66B0485@bioeng.ucsd.edu>

Gerson Kurz wrote:
> 
> you can initialize a fixed-size array using something like
> 
>         array = [0] * 10
> 
> now, one would think that you can initialize a multidimensional array
> like this:
> 
>         multidim = [ [0] * 10 ] * 3
> 
> but now, all 3 elements in multidim are references to the same single
> array, so that if you type
> 
>         multidim[0][0] = 1
>         print multidim
> 
> you actually get three 1 in all those 0s. Where did my thinking go
> wrong along these lines ?

If you use NumPy (http://sourceforge.net/projects/numpy/) you can do:
  array = zeros([10]) or 
  array = ones([10]) if you want initualized to ones.

-- 
Curtis Jensen
cjensen at bioeng.ucsd.edu
http://www-bioeng.ucsd.edu/~cjensen/
FAX (425) 740-1451


From sdm7g at virginia.edu  Wed Dec  6 14:05:18 2000
From: sdm7g at virginia.edu (Steven D. Majewski)
Date: Wed, 6 Dec 2000 14:05:18 -0500 (EST)
Subject: Creating Python objects out of a memory pool
In-Reply-To: <Pine.LNX.4.30.0012061415500.481-100000@localhost.localdomain>
Message-ID: <Pine.NXT.4.21.0012061402560.230-100000@localhost.virginia.edu>


On Wed, 6 Dec 2000, Damjan wrote:

> Is it posible to create Python objects in a known space in the memory
> (like in a shared memory).
> I want some some objects created in python to be in the shared memory
> pool, while others to be allocated as usual. Is there a way to hook in to
> the Python memory allocation from Python itself or I'll haveto use the C
> API?
> 

You need to write a custom allocator in C and recompile Python. 
Look at Include/pymem.h : 

/*
 * Core memory allocator
 * =====================
 */

/* To make sure the interpreter is user-malloc friendly, all memory
   APIs are implemented on top of this one.

   The PyCore_* macros can be defined to make the interpreter use a
   custom allocator. Note that they are for internal use only. Both
   the core and extension modules should use the PyMem_* API.

   See the comment block at the end of this file for two scenarios
   showing how to use this to use a different allocator. */



-- Steve Majewski <sdm7g at Virginia.EDU>





From Daniel.Kinnaer at AdValvas.be  Tue Dec 26 15:30:41 2000
From: Daniel.Kinnaer at AdValvas.be (Daniel Kinnaer)
Date: Tue, 26 Dec 2000 21:30:41 +0100
Subject: How to question
Message-ID: <LPBBLJLBPKOAHOOJGJIJKELICGAA.Daniel.Kinnaer@AdValvas.be>

How can one 'clear' or 'cls' the interactive window from PythonWin or form
IDLE? That is : I'd like to clear/cls this window so that all previous
output in the interactive window is erased and only the last 'run' is shown
in the interactive window...  Thanks

best regards,
Daniel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20001226/c53eb158/attachment.html>

From greg at cosc.canterbury.ac.nz  Thu Dec  7 19:27:07 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Fri, 08 Dec 2000 13:27:07 +1300
Subject: Programming Habits in Python
References: <cspU5.82459$U46.2617014@news1.sttls1.wa.home.com> <8vvb6p$ct2$1@news.wrc.xerox.com> <XiMU5.87123$U46.2803855@news1.sttls1.wa.home.com> <900cpl02ss1@news2.newsguy.com> <Ue3V5.91231$U46.2956112@news1.sttls1.wa.home.com> <slrn92tv0u.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com>
Message-ID: <3A302ADB.C4B86000@cosc.canterbury.ac.nz>

On Wed, 29 Nov 2000 08:37:08 GMT, Greg Jorgensen <gregj at pobox.com>
wrote:
> it's hard for me to imagine how anyone
> claiming to be a programmer can sleep without Fundamental Algorithms under
> their pillow ;-).

The thickness of the book makes the pillow rather too
high to be comfortable, in my experience. I'm thinking
of cutting a book-shaped depression in my mattress to
accommodate it. Must remember to allow enough room
for the Python Language & Library References, too...

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


From grante at visi.com  Tue Dec 12 10:40:43 2000
From: grante at visi.com (Grant Edwards)
Date: Tue, 12 Dec 2000 15:40:43 GMT
Subject: Invoking Python modules from Scheme
Message-ID: <%FrZ5.4704$TU6.497437@ptah.visi.com>

There seem to be a fair number of Schemers around, and I
thought you might find this interesting.  

Background: STk is a Scheme implimentation that includes
bindings to the Tk widget set, and an object system based on
tiny CLOS.

STk uses direct bindings to the Tk lib w/o Tcl, which has some
advantages and disadvantages. Scheme variables can be attached
directly to widgets without the intermediate IntVar and
StringVar stuff. The main disadvantage is that there's a lot
more work involved when the Tk library changes.

The STk home page is at:

             http://kaolin.unice.fr/STk/

Anyway, Walter Moreira has developed an interface that allows
gives STk access to Python modules/objects:


--------------------[from the STk mailing list]--------------------

From: Walter Moreira <wall at adinet.com.uy>
To: stk at kaolin.unice.fr
Subject: STk interface to Python
User-Agent: Mutt/1.0.1i
Resent-Message-ID: <"j-XDsB.A._3.z5VN6"@kaolin.unice.fr>
Resent-From: stk at kaolin.unice.fr
X-Mailing-List: <stk at kaolin.unice.fr> archive/latest/488
X-Loop: stk at kaolin.unice.fr
Precedence: list
Resent-Sender: stk-request at kaolin.unice.fr
     
Hello.

I wrote a small interface for invoking Python's libraries from
STk. It includes a STk type that represents an arbitrary Python
object and provides a mecanism for handling Python exceptions:
a try-except block.

For example, you can do

    (define http (py-getattr "BaseHTTPServer"))

and you have the complete Python's BaseHTTPServer' module for
writing your own Scheme HTTP server.

It can be downloaded (only 25k) from
	      
  http://www.cmat.edu.uy/~walterm/spy

Please, take a look and let me know if you find it interesting.
Walter

-------------------------------------------------------------------


-- 
Grant Edwards                   grante             Yow!  Okay... I'm going
                                  at               home to write the "I HATE
                               visi.com            RUBIK's CUBE HANDBOOK FOR
                                                   DEAD CAT LOVERS"...


From trentm at ActiveState.com  Mon Dec 11 11:39:38 2000
From: trentm at ActiveState.com (Trent Mick)
Date: Mon, 11 Dec 2000 08:39:38 -0800
Subject: A Newbie Question: How To Call Other Python Programs From A Main Python Program.
In-Reply-To: <912tpb$efl$1@nnrp1.deja.com>; from jbranthoover@my-deja.com on Mon, Dec 11, 2000 at 04:01:52PM +0000
References: <912tpb$efl$1@nnrp1.deja.com>
Message-ID: <20001211083938.B18075@ActiveState.com>

On Mon, Dec 11, 2000 at 04:01:52PM +0000, jbranthoover at my-deja.com wrote:
> Hello All,
> 	I have written several small Python programs.  These programs
> are not just functions but complete programs with their own menuing
> systems.
> 
>         My question is,  how do I (and is it a good idea) call these
> independent Python programs from one main program that selects which
> program to run from a main menu?  I tried to use the import command.
> This seems to work OK except that the programs only run once.  If the
> same program is selected from the menu,  nothing happens.  It seems to
> return immediately without running the program.

That is because an import call in Python will only execute (if that is the
correct term) the top-level Python code in that module the first time. The
module information is effectively cached after that. You *can* use the reload
function to force a reload of your modules, however the best way to set this
up is probably to change your Python programs so that their functionality is
in a function and then have you menu program run that function each time. So:

-------- func1.py ------------
def DoIt():
    print "hello from func1.py"

if __name__ == "__main__":
    # this is a magical way to get DoIt() to run only when func1.py
    # is run from the command line, this block is *not* run when you
    # "import func1" from another Python program. 
    DoIt()
------------------------------

-------- func2.py ------------
def DoIt():
    print "hello from func2.py"

if __name__ == "__main__":
    DoIt()
------------------------------

------------- menu.py ---------------------------------------
import func1, func2
import sys


if len(sys.argv) == 0:
    print "error: must have one argument"
elif sys.argv[0] == "func1":
    func1.DoIt()
elif sys.argv[0] == "func2":
    func2.DoIt()
else:
    print "error: unrecognized function name:", sys.argv[0]
------------------------------------------------------------


Then try running:

   python func1.py
   python menu.py func1
   python menu.py func2


Hope that helps,
Trent


-- 
Trent Mick
TrentM at ActiveState.com



From kelly.kranabetter at gems6.gov.bc.ca  Fri Dec 29 15:10:04 2000
From: kelly.kranabetter at gems6.gov.bc.ca (KellyK)
Date: Fri, 29 Dec 2000 20:10:04 GMT
Subject: 8.3 peculiarity if run w/i console
References: <92g4dn$5ug$1@newsreaderm1.core.theplanet.net> <VjP26.36442$xW4.287502@news-server.bigpond.net.au> <92ih6d$bt0$1@newsreaderm1.core.theplanet.net>
Message-ID: <3a4cee96.4237881050@nntp.gov.bc.ca>

On Fri, 29 Dec 2000 18:18:24 +0100, "Franz GEIGER" <fgeiger at datec.at>
wrote:
>Well, actually that's me, I call it "abbreviated". I know there are those
>two forms of file names, but I dont like the one containing "~1". How can I
>get rid of that?

Check out this web page:

http://ntfaq.com/Articles/Index.cfm?ArticleID=13721



From amb at nobjects.com  Thu Dec  7 15:12:37 2000
From: amb at nobjects.com (Are Magnus Bruaset)
Date: Thu, 07 Dec 2000 21:12:37 +0100
Subject: Problem with freeze.py in ActivePython 2.0
References: <3A2FB744.A942AD46@nobjects.com> <20001207082729.A23558@ActiveState.com>
Message-ID: <3A2FEF35.C001AD15@nobjects.com>

Thanks Mick,

I'm sorry I didn't tell what freeze.py is about. It's a part of Python
that will "freeze" a Python script (hello.py in the example below) to a
binary executable (hello.exe on Win32). From the Readme file in the
Python20\Tools\freeze directory it should work on Windows provided the
-p option is used to set the root of the Python installation. We have
successfully frozen our python scripts on 8 different Unix platforms
(part of a new software distribution from our company), but I have
failed to get it working on Windows.

Any advice on this matter is greatly appreciated. Please copy any
answers to amb at nobjects.com as I'm not a regular reader of the list.

Trent Mick wrote:
> 
> On Thu, Dec 07, 2000 at 05:13:56PM +0100, Are Magnus Bruaset wrote:
> >
> > Hi,
> >
> > I can't get freeze.py to behave under ActivePython 2.0 (Windows NT 4.0).
> > The command
> >
> >    python freeze.py -p E:\Python20 hello.py
> >
> > issued in E:\Python20\Tools\freeze (E:\Python20 is the root my
> > installation) leads to the error message
> >
> > Error: needed directory E:\Python20\lib\python2.0\config not found
> 
> I am not familiar with freeze.py but it looks to me like it is expecting a
> *Linux* installation directory structure. Perhaps freeze.py only works on
> Linux, or perhaps there is a command line option to signify window (though I
> don't know why it would require that).
> 
> The <sys.prefix>/lib/python2.0/config is the path to the directory containing
> the Python 2.0 library (i.e. the Linux equiv of the python20.lib file) on a
> Linux installation.
> 
> Trent
> 
> --
> Trent Mick
> TrentM at ActiveState.com

-- 
----------------------------------------------------------------------
   Are Magnus Bruaset, PhD      Email : amb at nobjects.com
   VP Development               Phone : +47 22 06 76 10
   Numerical Objects AS         Mobile: +47 92 61 00 63
   P.O. Box 124 Blindern        Fax   : +47 22 06 73 50
   N-0314 Oslo, Norway          WWW   : http://www.nobjects.com
----------------------------------------------------------------------



From fredrik at effbot.org  Fri Dec  8 05:25:10 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 08 Dec 2000 10:25:10 GMT
Subject: Changing image in label
References: <90q9k6$468$1@nnrp1.deja.com>
Message-ID: <aG2Y5.2164$Kd1.312886@newsb.telia.net>

2rkil at my-deja.com wrote:
> What happens is that the first picture disappears,
> and just the frame (the sizes) of the new one appears.

you have to keep a reference to the image in your Python
program.  the easiest way is to make it a global variable, or
to add it as an attribute to the widget objekt:

    global gif # keep a reference
    gif = PhotoImage(file=photo)
    photo_widget["image"] = gif

or

    gif = PhotoImage(file=photo)
    photo_widget.image = gif # keep a reference
    photo_widget["image"] = gif

for more info, see:

    http://www.python.org/doc/FAQ.html#4.69

</F>




From aleaxit at yahoo.com  Fri Dec 29 19:22:39 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 30 Dec 2000 01:22:39 +0100
Subject: Is 100,000 entries big for a dictionary?
References: <92iug4$vqb$1@nnrp1.deja.com>
Message-ID: <92ja97064m@news1.newsguy.com>

<rturpin at my-deja.com> wrote in message news:92iug4$vqb$1 at nnrp1.deja.com...
> I'm in the process of architecting an application with medium sized data
> sets. Python dictionaries are tempting as part of the representation
> mechanism. How do they perform when they have 100,000 entries? Can you

Some empirism to the rescue...:

import whrandom as rand
import time

dict = {}

start = time.clock()
for i in range(100000):
    dict[i] = i
stend = time.clock()

print "Populating the dict (100,000 items):",stend-start

start = time.clock()
for i in range(10000):
    x = rand.randrange(100000)
stend = time.clock()
gentime = stend-start


start = time.clock()
for i in range(10000):
    x = rand.randrange(100000)
    y = dict[x]
stend = time.clock()
reatime = (stend-start) - gentime

print "10,000 reads:", reatime


start = time.clock()
for i in range(10000):
    x = rand.randrange(100000)
    dict[x] = -93
stend = time.clock()
reatime = (stend-start) - gentime

print "10,000 writes", reatime


It should not matter that the dictionary is filled with increasing
numeric keys (except they're guaranteed to hash well!).

On my old cranky K5/300MHz running Win98, 96M RAM:


D:\PySym>python provdic.py
Populating the dict (100,000 items): 1.03804371511
10,000 reads: 0.174285522721
10,000 writes 0.21197220872

D:\PySym>python provdic.py
Populating the dict (100,000 items): 1.18972242243
10,000 reads: 0.0994661325198
10,000 writes 0.123712264704

D:\PySym>python provdic.py
Populating the dict (100,000 items): 1.22702609833
10,000 reads: 0.0950619353325
10,000 writes 0.0779203473072

So: about 1.0 to 1.2 seconds to insert 100,000 items;
about 0.1 to 0.2 seconds for 10k reads or writes,
depending on cache effects.  Prudentially, I'd state
this as about 10 to 20 microseconds per operation,
depending on cache effects, if hashing is maximally
fast and effective.

To test using strings as keys, I modified the script a
bit.  I prepended the generation of a list of 100,000
random words of 8 letters each:

print "generating 100,000 words..."
start = time.clock()
words = 100000*[None]
for i in range(100000):
    word = 6*[None]
    for j in range(6):
        word[j] = rand.choice(string.lowercase)
    words[i] = ''.join(word)
    if i&255==0: print '.',
stend = time.clock()
print
print "Words gemerated",stend-start

and I changed each dict[i] to dict[words[i]] in the
later part of the script.

Snipping the dots, I saw:

D:\PySym>python provdic.py
generating 100,000 words...
Words gemerated 48.0594755192
Populating the dict (100,000 items): 1.50119428753
10,000 reads: 0.0199911161769
10,000 writes 0.0385516016024

D:\PySym>python provdic.py
generating 100,000 words...
Words gemerated 54.563155601
Populating the dict (100,000 items): 1.48951792688
10,000 reads: 0.0423557216849
10,000 writes 0.134785195863

The populating-time grows modestly but consistently,
but the read and write times don't -- maybe some anomaly
in my measurements, but still, it seems 15 microseconds
per operation are still a good, rough estimate.


If these very approximate estimates are not decisive
enough to be a red or green light for your project, it
should not be hard to prepare a simple simulation that
mimics your intended use more closely.


> offer any other experience with this size data set in Python?

As long as everything fits comfortably in memory (and
these days you can fit quite a few 100,000-items dicts
in a typical PC's RAM), I never had any surprises in terms
of performance from dictionary use (lists and strings can
give such surprises, if naive use builds strings by many
small appends, or innocently inserts/deletes from the
interiors of big lists, say; not had to finesse once you
know what to look out for; but never did I notice any
such issue with dictionaries).


Alex





From pbjorn at uswest.net  Wed Dec 27 15:28:22 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Wed, 27 Dec 2000 13:28:22 -0700
Subject: Some Python 2.1 ideas
References: <mailman.977590932.19335.python-list@python.org> <Cw616.23054$qN6.3308205@typhoon.we.rr.com> <3A452881.7686E717@uswest.net> <%vb16.187233$U46.5801214@news1.sttls1.wa.home.com> <3A464D5D.B541C04E@uswest.net> <mailman.977688488.25438.python-list@python.org> <20001225150250.6FA06A84F@darjeeling.zadka.site.co.il> <mailman.977760789.10224.python-list@python.org> <roy-EA8C3B.11361325122000@news.panix.com> <3A479C29.2631F27B@uswest.net> <mailman.977876470.23568.python-list@python.org>
Message-ID: <3A4A50E6.11BBE0FD@uswest.net>

D-Man wrote:

> On Mon, Dec 25, 2000 at 12:12:41PM -0700, Bjorn Pettersen wrote:
> > I think this is a good idea by itself and should be implemented for all of
> > strip, lstrip, and rstrip, however it doesn't solve the problem we want to
> > solve with chomp (removing a line terminator, whatever the line terminator
> > happens to be).
>
> Java has a solution to this:
>
> import sys
> sys.line_terminator

Which would be wrong, since the file you're reading doesn't necessarily come
from the platform you're running on... (otherwise os.linesep would work).

-- bjorn



From dragon at hitech.net.au  Sat Dec 16 22:09:21 2000
From: dragon at hitech.net.au (dragon at hitech.net.au)
Date: Sun, 17 Dec 2000 03:09:21 GMT
Subject: Python for REAL Projects - Thanks for your input
References: <3a380d1e.409238@bendigo.news.origin.net.au> <919erq$tsr$1@bunyip.cc.uq.edu.au>
Message-ID: <3a3b131b.1967078@bendigo.news.origin.net.au>

Many thanks for the range of useful inputs I recieved for my posting

On Thu, 14 Dec 2000 13:28:21 +1000, "Chui Tey" <teyc at bigfoot.com>
wrote:

>Delphi is a development environment and you should really compare
>
>Python vs Pascal,
>Delphi vs PythonWorks, WingIDE
>
>
>
>



From deha at poczta.onet.pl  Tue Dec 12 07:36:06 2000
From: deha at poczta.onet.pl (DeHa)
Date: Tue, 12 Dec 2000 13:36:06 +0100
Subject: A question.
References: <9152a0$acb$1@panorama.wcss.wroc.pl> <slrn93c4mc.78m.scarblac@pino.selwerd.nl>
Message-ID: <91561v$cpm$1@panorama.wcss.wroc.pl>

Thanks, and sorry for my english:).

Dawid Hulisz




From jscott at marge.hologic.com  Mon Dec  4 14:45:30 2000
From: jscott at marge.hologic.com (Jeremy Scott)
Date: Mon, 04 Dec 2000 14:45:30 -0500
Subject: SAX2 XML validating?
Message-ID: <3A2BF45A.246B757F@marge.hologic.com>

Does anybody know how to validate an XML file using the SAX2 on
python2.0?



From MarkH at ActiveState.com  Thu Dec 21 21:57:50 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Fri, 22 Dec 2000 02:57:50 GMT
Subject: Events in win32com
References: <Y1g=Op6IfWSoJn1ISFtOyuvqwBb3@4ax.com> <FY0=OlZFLy71DtnK62u5TEPlhOts@4ax.com> <3A40060B.10504@ActiveState.com> <cYxAOmxi9FgNERgBo29zuCl5c2h1@4ax.com> <ik724tc2v3cpcqfbt7i1ookvaona0mk5ql@4ax.com> <3A418D02.2040302@ActiveState.com> <91t2mf02dob@news2.newsguy.com>
Message-ID: <3A42C1F7.7040407@ActiveState.com>

Alex Martelli wrote:

> Yes, but -- how does this play with the bForDemand flag, which
> is really a must when the module is as huge as MSHTML...?

Hrm.  Good question.

> Sure, once I managed to somehow grab the right module it's very
> easy; but I think I've not yet grasped how to do that grabbing
> in the current 'bForDemand' architecture (indispensable for
> really huge type libraries, of course).  For example, consider
> the following attempt:

with bForDemand, there are 2 distinct steps - getting the module and 
getting the class.  You are correct that the code I outlined wont work 
there.

The module is _always_ created at makepy time, and references to all 
subordinate items created at that time.

> But how do I instantiate and call genpy.Generator, or arrange for
> it to be instantiated and called, to ensure the HTMLDocumentEvents2.py
> gets built in the bForDemand setting...?  Is there a canonical/
> recommended way...?

gencache.GetClassForProgID() and GetClassForCLSIDID - however, the 
"EnsureModule" will already need to have been run.  There is no 
single-step generate-module-for-demand-and-get-a-specific-class.

I would be more than happy for us to step back and look at the 
makepy/gencache exposed API - I agree it sucks, as it "just grew" and 
was really never designed for direct use by client code - it just 
happened that way.

I would be happy for discussion either here, or on the Python COM 
developers list at 
http://mailman.pythonpros.com/mailman/listinfo/pycom-dev - that is very 
quiet tho - not much COM work is going on at the moment.

Mark.



From gerson.kurz at t-online.de  Sat Dec 30 12:29:05 2000
From: gerson.kurz at t-online.de (Gerson Kurz)
Date: Sat, 30 Dec 2000 17:29:05 GMT
Subject: Newbie: Manipulating Binary Files
Message-ID: <3a4e1b02.13335555@news.t-online.de>

How do you efficiently manipulate binary files in Python ? Say, you
want to extract a data frame from a MP3 file. Is there some package
that allows manipulation of bytes in the usual [] notation ?


From mdtompkins at home.com  Mon Dec 18 19:52:46 2000
From: mdtompkins at home.com (Mark Tompkins)
Date: Tue, 19 Dec 2000 00:52:46 +0000
Subject: HELP - Threading with SCO Unix
References: <91dbd4$9ot@dispatch.concentric.net>
Message-ID: <3A3EB15E.980C04C4@home.com>

Have you tried contacting SCO directly on this? It would definitely be in
their interest to have the latest and greatest version of Python running on
their platform.

Mark

Jonathan LaCour wrote:

> I am working with a corporation that is beginning development on a Python
> based product using Zope.  In order for our product to function, we need
> to have Python built with thread support.  One platform that we must
> support is SCO OpenServer.  We have been attempting to compile
> python-1.5.2 with thread support for several days without success.  We
> *must* have thread support, or we will be forced to move to another
> solution.
>
> Does Python even support threads in SCO Unix?
>
> Any suggestions?  We have attempted our build using a binary distribution
> of GNU PThreads for SCO OpenServer.  Installing another UNIX is not an
> option at this stage.
>
> If you are running SCO OpenServer and have python with threading support,
> please contact me!  We need your help!
>
> Thanks for you help!  This is absolutely critical for success, and any
> assistence would be appreciated!
>
> Jonathan LaCour
> Developer
> VertiSoft, Corp.



From jh at cddk.dk  Mon Dec  4 09:55:55 2000
From: jh at cddk.dk (Jesper Hertel)
Date: Mon, 4 Dec 2000 15:55:55 +0100
Subject: newbie problem with 'ord' function
References: <3A291456.B7402EC0@sci.fi> <90b53f$31r$1@news.tpi.pl>
Message-ID: <90gbac$ncb$1@news.inet.tele.dk>

Here is another way to do it:


digits = "0123456789"

print "NUMBER", "ASCII Code"

for digit in digits:
    print digit, ord(digit)


This works because a string in Python can also act as a list of characters.
You could expand the 'digits' string to contain other characters than
digits.

/Jesper


"dr. franken" <oniet at poczta.onet.pl> wrote in message
news:90b53f$31r$1 at news.tpi.pl...
> >    numbers = [1, 2, 3, 4, 5, 6, 7, 8, 9, 0]
> >    NUMBER   ASCII Code
> > 1 49
> > 2 50 etc.
> >
> > What would do the trick?
>
> for x in numbers:
>   print x, ord(`x`)  #not 'x' but `x`!
>
> --
> franken
>
>




From estechmann at abivest.com  Mon Dec  4 11:56:39 2000
From: estechmann at abivest.com (Eric Stechmann)
Date: Mon, 04 Dec 2000 10:56:39 -0600
Subject: (no subject)
Message-ID: <5.0.2.1.1.20001204105504.00aa6ec0@mail.abivest.com>

Sorry to waste bandwidth but...

   I'm receiving little or nothing on this newsgroup.  Can anyone
   fill me in on the situation, please?

Many thanks

Eric
--------------------------------------------------------------------------
Eric Stechmann                  Direct:  +1 (651) 234-1217
Software Critter                   Fax:  +1 (651) 490-1484
American Biosystems, Inc.        E-mail:  estechmann at abivest.com
1020 West County Road F            URL:  www.abivest.com
St.Paul  MN  55126
--------------------------------------------------------------------------
The program said "Requires Windows 9* or better" so I bought a Macintosh.




From bernhard at abnoba.intevation.de  Thu Dec 28 04:32:59 2000
From: bernhard at abnoba.intevation.de (Bernhard Reiter)
Date: 28 Dec 2000 09:32:59 GMT
Subject: Problems installing PIL
References: <923cf6$s7q$1@nnrp1.deja.com>
Message-ID: <92f1cb$re4$2@newsserver.rrzn.uni-hannover.de>

In article <923cf6$s7q$1 at nnrp1.deja.com>,
	EdK <e_d_k at my-deja.com> writes:
> I am unsing Python 1.6 on Windows 98. 

Switch to 2.0 or downgrade to 1.5.2.
There ist no reason to use 1.6.

Then try to download PIL from
	http://piddle.sourceforge.net/downloads/

Fredrik also had a place where you could download PIL,
but I do not remember it right now...
	Bernhard
-- 
Professional Service around Free Software                (intevation.net)  
The FreeGIS Project                                         (freegis.org)
Association for a Free Informational Infrastructure            (ffii.org)
FSF Europe                                            (www.fsfeurope.org)


From chicago_george at my-deja.com  Wed Dec 13 11:58:33 2000
From: chicago_george at my-deja.com (chicago_george at my-deja.com)
Date: Wed, 13 Dec 2000 16:58:33 GMT
Subject: Simple question/problem
References: <916b28$9lf$1@nnrp1.deja.com> <3A36B4BF.80323E6F@yale.edu>
Message-ID: <9189rk$r43$1@nnrp1.deja.com>

In article <3A36B4BF.80323E6F at yale.edu>,
  Paul Magwene <paul.magwene at yale.edu> wrote:
> chicago_george at my-deja.com wrote:
> >
> > Hello, I'm new to Python.. When I try to import java classes, I get
an
> > error msg saying 'no module named java'
> > For example;
> >
> > >>> from java import awt
> > ImportError: no module named Java
> >
> > Why is this happening?  Anyone???
> >
>
> I guess I'll ask the obvious - are you using Python (i.e. CPython) or
> JPython/Jython?
>

I'm using JPython
> --
>
> Paul Magwene
> paul.magwene at yale.edu
>


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


From syver at NOSPAMcyberwatcher.com  Sun Dec 10 11:05:58 2000
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Sun, 10 Dec 2000 17:05:58 +0100
Subject: reloading modules question
References: <90scdf$hpe$1@troll.powertech.no> <p6qsnnxamtc.fsf@informatik.hu-berlin.de>
Message-ID: <9109q3$ter$1@troll.powertech.no>

"Martin von Loewis" <loewis at informatik.hu-berlin.de> wrote in message
news:p6qsnnxamtc.fsf at informatik.hu-berlin.de...
> "Syver Enstad" <syver at NOSPAMcyberwatcher.com> writes:
>
> > Is it safe to loop through sys.modules.items()
> > reloading each module, as long as I catch the ImportError?
>
> What do you mean by "safe"? Python will continue to work, if that is
> what you mean.

Yes that's what I mean. I've read statements to the effect that it's unsafe
to reload extension modules.

> > Am I guaranteed to have all python modules reloaded if I do the loop
twice?
>
> No. Reloading a builtin module does not do anything.
That's okay I'm not talking about builtin's or extensions here just python
modules (ie. written in python). Let me rephrase: Am I guaranteed to have
all python source code modules reloaded if I do the loop twice?


> Which of your modules require reloading when they are reused? I
> recommend to redesign the modules so that they act as libraries,
> instead of keeping session state.

I don't understand you here. All modules that are changed require reloading,
or else the changes won't come into play. I don't see what this has to do
with session state.

To clarify: The IIS appears to dynamically link (via COM) to the Python
interpreter and keep it in memory. Therefore, if you don't want to stop the
IIS process (or subprocesses) you'll have to order the interpreter to reload
the Python source or it won't get used.




From kiko at async.com.br  Mon Dec  4 19:59:37 2000
From: kiko at async.com.br (Christian Robottom Reis)
Date: Mon, 4 Dec 2000 22:59:37 -0200 (BRST)
Subject: [pygtk] Python and Orbit Examples
In-Reply-To: <200012040748.eB47mEh22820@ZOOstation.cc>
Message-ID: <Pine.LNX.4.20.0012042258150.21593-100000@blackjesus.async.com.br>

On 3 Dec 2000, Rob Brown-Bayliss wrote:

> Has anyone got some simple, usefull examples for a beginer to Python and
> Corba in how to make two python apps on two different machines say hello
> to each other, preferably with examples that works with pyGnome

Using orbit-python this is quite easy, though you'll have to take care of
the IOR passing between them which is tricky.

http://projects.sault.org/orbit-python/

There is a very simple example included that will work.

Take care,
--
/\/\ Christian Reis, Senior Engineer, Async Open Source, Brazil
~\/~ http://async.com.br/~kiko/ | [+55 16] 274 4311




From entropiamax at jazzfree.com  Wed Dec 27 08:31:54 2000
From: entropiamax at jazzfree.com (entropia)
Date: Wed, 27 Dec 2000 14:31:54 +0100
Subject: Python vs. PHP (& Java?)
References: <LOBBJMHOIFOLFIPFANHJMEMHCFAA.beno@TheWebsons.com> <20001227110037.G25820@xs4all.nl>
Message-ID: <3A49EF49.8042E3F2@jazzfree.com>

As an example of how easy is python compared to java, look at the diferences of
xml-rpc implementations:
The links can be find at: www.xmrpc.org
I think that using python for xmlrpc is more elegant:
In Java I should do something like this:
try
          {
                  Object[] numParam = new Object[1];
                  Long stateNum = new Long(24);
                  numParam[0] = stateNum;
                  RPCClient rpc = new RPCClient();
                  rpc.setServer("betty.userland.com");
                  rpc.setPort(80);
                  rpc.setProcedureName("examples.getStateName");
                  rpc.setParams(numParam);
                  String stateName = (String)rpc.executeRPC();
          }
          catch (RPCException ex)
          {
                  ex.printStackTrace();
          }

In Python:
import xmlrpclib
betty =
                      Server("http://betty.userland.com")
                      print betty.examples.getStateName(41)


Thomas Wouters escribi?:

> On Tue, Dec 26, 2000 at 10:03:05PM -0800, Ben Ocean wrote:
>
> > I?m an intermediate-level programmer that?s new to Python (via Zope) and I
> > like it a lot. However, it appears to me that, while Python is a good
> > replacement for Perl, it cannot replace much of the functionality of PHP,
> > which can be scripted into HTML (whereas Python, it appears, must be
> > called). So, I have to wonder about the speed of Python vs. PHP for
> > Internet-based applications. Python has the advantage of byte-code and
> > compilation, but it?s server-side. PHP might not be as fast on the server,
> > but it doesn?t have to make as many trips back and forth between server
> > and client. Plus, I?m curious as to why Python wasn?t written in such a
> > way that it could be written directly into HTML documents like PHP.
>
> You don't need roundtrips with Python either, you just need to start it as a
> CGI script (which is some additional overhead, granted.) And note that there
> are several extentions (PSP, ASP and if-I-recall-correctly mod_snake for
> apache, and Zope's DTML) that allow embedding Python into HTML.
>
> As for why Python isn't PHP, well, because it isn't. Python is a fair bit
> older than PHP, and simply wasn't intended as an 'html-embedded-language'.
> It was intended as a proper, easy to learn, easy to read, efficient (from
> the programmers view) language, and not as a quick-and-dirty-hack to allow
> code inbetween HTML :) The WWW and HTML did exist, but were only just
> starting, when Python emerged.
>
> And then there's two other points. I do a reasonable amount of PHP, and I
> think it sucks :) I don't see a terribly good reason to mix HTML and script
> code, for one. It looks sexy but there isn't a real advantage, and it only
> serves to clutter the code, IMHO. And besides that I've been bitten by
> irregularities in PHP code way too often, like how you can pass environment
> variables to the html page which end up variables inside the script --
> that's been the cause of several huge security issues for me. So even if you
> really want to embed code into HTML, I'd go for Python (or even Perl!)
> rather than PHP.
>
> > On a similar subject, someone once informed me that Java was by far the
> > best language for scripting communication programs over the Internet (like
> > chat, IRC, etc.). How does Python compare? Your comments are appreciated.
>
> Well, not suprising (since this is a Python newsgroup/list, not a Java one)
> I don't agree. Python is easier to write and easier to read (and
> understand), and way easier to learn. Python is (supposedly :) slower than
> Java in execution, but especially in networked environments, where you spend
> most of your time waiting for the network anyway, you aren't going to notice
> that.
>
> And besides, Python is so easy, you can learn it pretty quickly (in a few
> hours if you're already a somewhat-experienced programmer) and I doubt
> you'll ever regret it. If anything, it'll be a good programming learning
> experience. And it's just a fraction of the time you'll spend learning Java
> :)
>
> --
> Thomas Wouters <thomas at xs4all.net>
>
> Hi! I'm a .signature virus! copy me into your .signature file to help me spread!
>
> --
> http://www.python.org/mailman/listinfo/python-list




From MURBAN at houston.rr.com  Sat Dec 16 12:45:23 2000
From: MURBAN at houston.rr.com (Marcus Urban)
Date: Sat, 16 Dec 2000 17:45:23 GMT
Subject: Compiling PyQT-2.2
Message-ID: <TSN_5.8418$wX2.112805@typhoon.austin.rr.com>

I've been trying to build PyQt for several days now and it continues to
fail during the "make install" process. SIP compiles just fine. No
complaints during ./configure or make/install. PyQt configures fine, too,
with no complaints. The make process succeeds. But "make install" aborts.
And yes, QTDIR is set correctly.


/bin/sh ../mkinstalldirs /usr/local/lib/python2.0/site-packages
 /usr/bin/ginstall -c -m 644 ./qt.py /usr/local/lib/python2.0/site-packages/qt.py
make  install-data-hook
make[3]: Entering directory `/home/urban/py/PyQt-2.2/qt'
(cd /tmp; PYTHONPATH=/usr/local/lib/python2.0/site-packages:/usr/local/lib/python2.0/site-packages python -c "import qt")
Traceback (innermost last):
  File "<string>", line 1, in ?
  File "/usr/local/lib/python2.0/site-packages/qt.py", line 43, in ?
    import libqtc
ImportError: /usr/local/lib/libsip.so.3: undefined symbol: PyObject_Init
make[3]: *** [install-data-hook] Error 1
make[3]: Leaving directory `/home/urban/py/PyQt-2.2/qt'
make[2]: *** [install-data-am] Error 2
make[2]: Leaving directory `/home/urban/py/PyQt-2.2/qt'
make[1]: *** [install-am] Error 2
make[1]: Leaving directory `/home/urban/py/PyQt-2.2/qt'
make: *** [install-recursive] Error 1


From grante at visi.com  Thu Dec 28 11:57:06 2000
From: grante at visi.com (Grant Edwards)
Date: Thu, 28 Dec 2000 16:57:06 GMT
Subject: Python in embedded systems?
References: <_w706.11237$TU6.1069571@ptah.visi.com> <JFa06.749$Qb7.61608@newsb.telia.net>
Message-ID: <ChK26.3661$4c.279560@ruti.visi.com>

In article <JFa06.749$Qb7.61608 at newsb.telia.net>, Fredrik Lundh wrote:

>Grant Edwards wrote:
>> Has anybody heard of a Python interpreter being used in an
>> embedded system to provide user scripting capabilities?
>
>    http://www.abo.fi/~iporres/python/
>
>    "Deeply Embedded Python"
>
>    "With this patch, it is possible to build a python
>    interpreter WITHOUT support for float, complex
>    and long numbers, file objects, parser, compiler
>    and depedencies with any operating system."

Cool! That's pretty much exactly what I'm looking for.

-- 
Grant Edwards                   grante             Yow!  So this is what it
                                  at               feels like to be potato
                               visi.com            salad


From pbjorn at uswest.net  Thu Dec 28 13:37:25 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Thu, 28 Dec 2000 11:37:25 -0700
Subject: Some Python 2.1 ideas
References: <3A452881.7686E717@uswest.net> <%vb16.187233$U46.5801214@news1.sttls1.wa.home.com> <3A464D5D.B541C04E@uswest.net> <mailman.977688488.25438.python-list@python.org> <20001225150250.6FA06A84F@darjeeling.zadka.site.co.il> <mailman.977760789.10224.python-list@python.org> <roy-EA8C3B.11361325122000@news.panix.com> <3A479C29.2631F27B@uswest.net> <mailman.977876470.23568.python-list@python.org> <3A4A50E6.11BBE0FD@uswest.net> <mailman.977962871.16241.python-list@python.org> <3A4AB05C.D41F4022@uswest.net> <92ftif$17fc$1@nntp6.u.washington.edu>
Message-ID: <3A4B8865.86712FA1@uswest.net>

Donn Cave wrote:

> Quoth Bjorn Pettersen <pbjorn at uswest.net>:
> | D-Man wrote:
> |
> | > Yeah, it is a problem if you don't know where the file came from.  How
> | > about creating a standard text file format?  All files could be opened
> | > in binary mode on all platforms, then EOL is handled by each
> | > application (or better yet make a standard module that does the
> | > implementation correctly once).
> |
> | Feeling like fighting windmills today, eh <wink>? I think the current wish is to
> | do it from the other end, i.e. instead of relying on the text file format to be
> | "correct", implement a method on string object that is smart enough to handle all
> | text file format line endings intelligently.
>
> Without the context of the I/O stream, to decide whether '\r' is
> data or separator, isn't that string function going to be the opposite
> of smart?  How does a string handle '\r' intelligently?

By assuming that if it's the last character in the string, it came from a mac file and
is the line terminator. Similarly for '\r\n' (dos) and '\n' (unix) when at the end of
the string.

It might be cute to have a MacTextFile class etc. that ensures that the line endings
are consistent, but personally I think that is overkill (and is likely to be very
slow).

-- bjorn




From glenfant at equod.com.nospam  Thu Dec 21 04:25:34 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Thu, 21 Dec 2000 10:25:34 +0100
Subject: ´º½º±×·ì°¡ÀÔ
References: <mailman.977386685.2743.python-list@python.org>
Message-ID: <91si1m$aml$1@reader1.imaginet.fr>

Is this a new regular expression language ? 8))

"??????" <hjyi83 at nownuri.net> a ?crit dans le message news:
mailman.977386685.2743.python-list at python.org...
>
> ???????? ????
>
>
> --MIME Multi-part separator--
>
>



From s2mdalle at titan.vcu.edu  Tue Dec 19 23:24:23 2000
From: s2mdalle at titan.vcu.edu (David Allen)
Date: Wed, 20 Dec 2000 09:24:23 +0500
Subject: Import problem
Message-ID: <91r1j3$94h$1@bob.news.rcn.net>

I have a module that uses a particular object.  So
at the top of the file, I import that object.  But
for some reason, trying to create instances of that
object always fails with a NameError.  Here's the
error message:

Unhandled exception in thread:
Traceback (innermost last):
  File "./TkGui.py", line 285, in downloadThread
    contentKey = self.createContentFrame(request)
  File "./TkGui.py", line 350, in createContentFrame
    resp = conn.getResource(resource, self.messageBar)
  File "./GopherConnection.py", line 79, in getResource
    resp = GopherResponse()
NameError: GopherResponse

Now, to verify that I'm not smoking crack, here
are the first few lines of GopherConnection.py:

from GopherResponse      import *
from GopherResource      import *
from ResourceInformation import *

And just in case you think I'm crazy and that
GopherResponse isn't defined in GopherResponse.py,
here are the first few lines of that file:


from GopherObject        import *
from GopherResource      import *
from ResourceInformation import *

class GopherResponse(GopherObject):
	... code ...

So here's the issue - I'm importing a module, but
things in its namespace aren't coming in.

As far as I can come up with, the only idea I have
on why this might be happening is because many of
the objects use each other, so you get a situation
where many things import many other things, and you
may be getting a chicken and egg situation with
the importer.  But I thought python's importer was
brigher than that.

What else can cause this?  How is it that you can
from modulename import * and then NOT have access
to the things in that namespace?  Alternatively,
why does this case not work, when, say, I can use
module X in module Y and it works fine?  

Any help on figuring this out would be much appreciated.
-- 
David Allen
http://opop.nols.com/
----------------------------------------
It's 5:50 AM...do you know where your stack pointer is?


From loewis at informatik.hu-berlin.de  Fri Dec  8 14:59:31 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 08 Dec 2000 20:59:31 +0100
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu> <976153091.257544@clam-ext> <3A30A181.4302F102@caltech.edu>
Message-ID: <p6qr93ik7ks.fsf@informatik.hu-berlin.de>

Nathaniel Gray <nospam at caltech.edu> writes:

> Granted, this would have been a better approach.  I'm not saying it
> was the best question ever, but I have no qualms with calling it
> legitimate.

To be honest, I do. Some people have spent quite some time putting
together an FAQ list for Python, so that frequently-asked questions
would not need to be answered over-and-over-again. There was no
indication that the original poster was aware of

http://www.python.org/doc/FAQ.html#1.2
"Why is it called Python" or

http://www.python.org/doc/FAQ.html#1.15
"Why was Python created in the first place"

It may be that the author was aware of it, but I feel a response that
just says "RTFM" is as legitimate as the question (of course, such an
answer should at least supply the location of the FAQ list).

Regards,
Martin


From trotts at llnl.gov  Fri Dec 22 15:20:15 2000
From: trotts at llnl.gov (Issac Trotts)
Date: Fri, 22 Dec 2000 12:20:15 -0800
Subject: coding convention conversion
Message-ID: <20001222122015.A7987@llnl.gov>

Does anyone know of a utility to change between the coding conventions 
of

def fooBarBaz(): pass

and 

def foo_bar_baz(): pass

It is a pain to do this by hand.

Thanks,
Issac




From qdlaty at wielun.dhs.org  Thu Dec 28 11:20:24 2000
From: qdlaty at wielun.dhs.org (QdlatY)
Date: Thu, 28 Dec 2000 17:20:24 +0100
Subject: cgi parameters
References: <3A4B4BD8.6A696387@wielun.dhs.org>
Message-ID: <3A4B6848.71EB6015@wielun.dhs.org>

Hello!

QdlatY wrote:
> 
> Hello!
> 
> I need to pass index parameter to script executed by post method (via
> form)
> 
> I tried to use following code:
> 
> <form action="newpost.cgi?index=1" method="post">
> 
> But this won't work
> (fieldstorage doesn't have "index" property)


Ok. Thanks.
Solution is to create input with "hidden" type and "index name.

Yours
QdlatY


From tgagne at ix.netcom.com  Wed Dec 20 13:21:28 2000
From: tgagne at ix.netcom.com (Thomas Gagne)
Date: Wed, 20 Dec 2000 13:21:28 -0500
Subject: Catching Database Exceptions (Sybase)
References: <3A40EEBF.3262040F@ix.netcom.com>
Message-ID: <3A40F8A8.8CD9A563@ix.netcom.com>

Just to keep checking..

.. I deliberately used a procedure I know didn't exist.  This time it seems as
though it tried to test the exceptions:

Traceback (innermost last):
  File ".//floater.py", line 50, in ?
    except Warning, env:
NameError: Warning

The DB API 2.0 documentation suggested that all database warnings and errors
should be subclasses of Error and Warning.  Is there something I'm supposed to
be importing to make this work, or to define these objects?

--
.tom





From joseamartin at venmex.com  Thu Dec  7 13:28:51 2000
From: joseamartin at venmex.com (Jose Martin)
Date: Thu, 7 Dec 2000 14:28:51 -0400
Subject: DCOracle Need Help
References: <t2v9fu2pht2d69@corp.supernews.com> <slrn92vfkd.ut.gerhard.nospam@lilith.hqd-internal.de>
Message-ID: <t2vm6g9u2qhle6@corp.supernews.com>

no, i dont
can you help me with this ?

Gerhard H?ring <gerhard.nospam at bigfoot.de> escribi? en el mensaje de
noticias slrn92vfkd.ut.gerhard.nospam at lilith.hqd-internal.de...
> The Buffer module isn't a pure Python module. It's an extension module
> written in C. Have you read the installation instructions and compiled the
C
> module?
>
> Gerhard H?ring
>
> On Thu, 7 Dec 2000 10:48:32 -0400, Jose Martin <joseamartin at venmex.com>
wrote:
> >I am trying to use DCOracle on Python 2.0 under Windows or Linux
> >but when i try to make my scritp :
> >
> >The first line is
> >from DCOracle import *
> >
> >when i try to run:
> >Python say me:
> >
> >File "d:\python20\DCOracle\__init__.py", line 52, in ?
> >    import sys, Buffer
> >ImportError: No module named Buffer
> >
> >I was searched my python path for a module called Buffer, but i dont have
it
> >Whats happennig ?
>
> --
> mail:   gerhard <at> bigfoot <dot> de
> web:    http://highqualdev.com




From emile at fenx.com  Fri Dec 29 11:52:56 2000
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 29 Dec 2000 08:52:56 -0800
Subject: identifying user from Win9x?
References: <92ibvf$if2$1@nntp6.u.washington.edu>
Message-ID: <92ifhb$75gql$1@ID-11957.news.dfncis.de>

One way:

>>> import os
>>> info = os.popen('net config').readlines()

HTH

--

Emile van Sebille
emile at fenx.com
-------------------


"Frank Miles" <fpm at u.washington.edu> wrote in message
news:92ibvf$if2$1 at nntp6.u.washington.edu...
> I have a set of python routines that are used to
coordinate and
> account for various activities in my group.  I would like
to be
> able to tailor how these routines work, and what options
are first
> presented to each user.
>
> There are ways to determine the user ID in Unix-based
systems (os.getuid).
> I haven't yet seen anything comparable for Windows.
Possible source
> of information: it is SOP for each user to login to the
(Novell-based)
> network before doing anything else.  Is there some way to
extract this
> information from Windows, or failing that, a Novell
network/server?
>
> TIA...
>
> -frank
>
> --




From pbjorn at uswest.net  Sat Dec 30 13:49:39 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Sat, 30 Dec 2000 11:49:39 -0700
Subject: Newbie: Manipulating Binary Files
References: <3a4e1b02.13335555@news.t-online.de>
Message-ID: <3A4E2E43.C20D860F@uswest.net>

Take a look at the struct module. If you just want to look at the bytes,
opening the file in binary mode:

  binaryfile = open('foo.mp3', 'rb')
  data = binaryfile.read()

will work...

all-depends-on-what-you-want-to-do-with-the-bytes'ly y'rs
-- bjorn

Gerson Kurz wrote:

> How do you efficiently manipulate binary files in Python ? Say, you
> want to extract a data frame from a MP3 file. Is there some package
> that allows manipulation of bytes in the usual [] notation ?



From jorn.verwey at psi.ch  Thu Dec 14 08:33:38 2000
From: jorn.verwey at psi.ch (Jorn Verwey)
Date: Thu, 14 Dec 2000 14:33:38 +0100
Subject: sorting date-strings of form: Wed, 05 Jan 2000 10:07:02
Message-ID: <3A38CC30.8DA734F4@psi.ch>

Hello,

I have a document with a lot of sections which need to be ordered by date. The date is given in the
format given in the subject: Wed, 05 Jan 2000 10:07:02. How would I go about? Being a novice I would
rather not re-invent the wheel (it will end up square). I took notice of two packages which might
help me (mxDateTime and normalDate), but it didn't seem straight forward. I welcome suggestions!
TIA.

Jorn

--
Jorn Verwey
PSI
Life Sciences WMSA/B14
CH 5232 Villigen PSI
Switzerland
tel. +41 56 310 4246 -o)
fax. +41 56 310 3132 /\\
Jorn.Verwey at psi.ch  _\_v





From claird at starbase.neosoft.com  Fri Dec 29 09:20:53 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 29 Dec 2000 08:20:53 -0600
Subject: Python vs. PHP (& Java?)
References: <69BC4C6AE83ED311BC9400805FA7B14AF84D2C@il93exp01.css.mot.com> <mailman.978059571.14515.python-list@python.org>
Message-ID: <397C7FA100EE65A5.0FF0B81DCC646CA3.C3E3EFFAF818CC30@lp.airnews.net>

In article <mailman.978059571.14515.python-list at python.org>,
Moshe Zadka  <moshez at zadka.site.co.il> wrote:
			.
			.
			.
>>   Isn't there an axiom in physics that an experimenter influences the 
>>   outcome of his experiment (physicists, help me out here)?  
>
>Widely overused by laymen.
>It says that in *some* experiments, measuring one quantity *has* to
>interfere with another quantity some specified amount. 
			.
			.
			.
It's such a beautiful result as to be worth 
expression in more suggestive language:  cer-
tain measurements are in a dual relationship,
one manifestation of which is that there is
a precise lower bound on the accuracy of
their *simultaneous* determination.  The
analytic form of this leads immediately to 
the mathematics of, for example, signal com-
pression or image enhancement.

Metaphoric application to macroscopic phenomena
is, as Moshe notes, often overdone.
-- 

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


From joseamartin at venmex.com  Wed Dec 20 07:33:22 2000
From: joseamartin at venmex.com (Jose Martin)
Date: Wed, 20 Dec 2000 08:33:22 -0400
Subject: real life examples
References: <u48q19.cth.ln@adsl.bapnt>
Message-ID: <t41abeqn4uel26@corp.supernews.com>

Show him VPython , do you know VPython.

VPython is a 3D modelator. it is like a library.
You can do animmations and 3d figures using Python, it is very easy !

There are libraries to make 3D Games using python and VPython.

Try it !.


huber <nospam at wanted.com> escribi? en el mensaje de noticias
u48q19.cth.ln at adsl.bapnt...
> i try to teach python to my son (13 years old)!
> but instead of stupid for loops, classes nobody ever needs and so on .....
> are there some good real life examples a child could be interested in ?
>
> thanks
> udo




From alan.gauld at gssec.bt.co.uk  Thu Dec 14 09:15:50 2000
From: alan.gauld at gssec.bt.co.uk (Alan Gauld)
Date: Thu, 14 Dec 2000 14:15:50 +0000
Subject: Python for REAL Projects ?
References: <3a380d1e.409238@bendigo.news.origin.net.au>
Message-ID: <3A38D616.B00927B@gssec.bt.co.uk>

dragon at hitech.net.au wrote:
> although Python SEEMS interesting can it "cut the mustad" on a real
> client project ??

As ever it depends on the project.

Personally I use Python for general scripting and as a
prototyping 
tool for the non GUI bits of projects and Delphi for GUI bits.

The final code is nearly always in Delphi.

Python servers are a possibility but the lack of speed means 
bigger hardware so the Delphi approach usually wins there too.

But if your user base is not huge and you have powerful servers 
available and development times are the critical factor then 
yes Python can produce industry strength apps.

Alan g.

-- 
=================================================
This post represents the views of the author 
and does not necessarily accurately represent 
the views of BT.


From sergecpelletier at yahoo.com  Mon Dec  4 23:52:02 2000
From: sergecpelletier at yahoo.com (Serge Chevarie-Pelletier)
Date: Mon, 4 Dec 2000 23:52:02 -0500
Subject: Changing Users in Python
References: <9TkW5.40369$nh5.2798316@newsread1.prod.itd.earthlink.net> <slrn92lsap.f20.cjc26@synecdoche.sowrong.org>
Message-ID: <Is_W5.2621$0f1.166960@wagner.videotron.net>

Using popen is as bad as having a bad setuid program.



Cliff Crawford wrote:

> You might be able to do it using the getpass module to get the oracle
> user's password, and then use something like
> 
> cmdin, cmdout = os.popen('su - oracle -c "cmd you want to run as oracle"')
> cmdin.write(passwd + "\n")
> 
> to execute the command.  I haven't actually tested this though.
> 
> Another way to do it would be to setuid your script, but this can cause
> a whole lot of security problems.  So don't do it. ;)
> 
> 
> * Doug Farrell <dfarrell01 at earthlink.net> menulis:
> | 
> | I'm fairly new at Python and I'm thinking about writing some programs in
> | for use at work on the Sun web servers we use. But I need to know
> | something, I have a user on the system who delivers data for input into
> | the Oracle database, and I'd like them to run some scripts to do it,
> | rather than me doing all the time. However, to do this they have to 'su
> | - oracle', in otherwords, the script would have to change userid and run
> | some Oracle scripts than exit from that user and return to the delivery
> | user. Can Python do this kind of thing? And if so, how?
> 
> 




From not.this at seebelow.org  Fri Dec  8 17:09:30 2000
From: not.this at seebelow.org (Grant Griffin)
Date: Fri, 08 Dec 2000 22:09:30 +0000
Subject: exec'ing an import
Message-ID: <3A315C1A.90CB8D98@seebelow.org>

Hi Gang,

I had a problem the other day trying to exec an import statement, which
I used in order to collect the doc strings of a set of modules.  I kept
getting an ImportError.

I tried several different things, including using the __import__
function.  One surprising symptom was that my method worked fine when I
started the interpreter and did it interactively, but it didn't work in
a Python script.

Finally, I discovered that that when I executed the script "script.py"
using just the script name (like "script") I got the ImportError, but
when I executed it by calling the Python interpreter (ike "python
script.py") it worked fine.

Does anybody have an explanation?  This thing kindda sounds like a
security feature of some kind, but if so, can somebody give me a
reference to the docs to explain it?  (I'm using Python 2.0 and Windows
NT 4.0.)

import-ant-ly y'rs,

=g2
-- 
_____________________________________________________________________

Grant R. Griffin                                       g2 at dspguru.com
Publisher of dspGuru                           http://www.dspguru.com
Iowegian International Corporation	      http://www.iowegian.com


From joseamartin at venmex.com  Fri Dec 22 07:06:57 2000
From: joseamartin at venmex.com (Jose Martin)
Date: Fri, 22 Dec 2000 08:06:57 -0400
Subject: Help with Oracle Headers files to compile DCOracle
Message-ID: <t46hihbdldm430@corp.supernews.com>

Hi, I need this help for you

Could you make a zip file with the headers files (.h) etc, that i need to
compile with DCOracle

Thanks ....







From entropiamax at jazzfree.com  Mon Dec 18 06:09:52 2000
From: entropiamax at jazzfree.com (entropia)
Date: Mon, 18 Dec 2000 12:09:52 +0100
Subject: Error evaluating numbers starting in zero
References: <3A3DEE53.621D0D94@mitre.org>
Message-ID: <3A3DF07F.5C0080F@jazzfree.com>

The numbers starting with 0 are octal (and with 0x hexadecimal).
08 doesn't exist in octal and this is a error.
8 is 010 in octal:
>>010
8


Mike Brenner escribi?:

> >> I have a python checkbook program that evaluates date strings, such
> >> as "01/31/99". Most dates are fine, but if I have a date such as
> >> "01/08/99", python runs into an error when it tries to eval("08").
>
> > Use int() instead of eval().
>
> So, that may also explain the failure with
>
>         i = 08
>
> and with reading in numbers,
> the python interpretor must be using EVAL instead of INT?
>
> --
> http://www.python.org/mailman/listinfo/python-list




From jasonic at nomadicsltd.com  Mon Dec  4 13:09:21 2000
From: jasonic at nomadicsltd.com (Jason Cunliffe)
Date: Mon, 4 Dec 2000 13:09:21 -0500
Subject: Programming with Python
References: <90fq39$jgq$1@nnrp1.deja.com> <9iOW5.55847$L26.16633720@typhoon.southeast.rr.com>
Message-ID: <t2nnac9qvoh664@corp.supernews.com>

Yes. I may be wrong, but Prima_Tech seem to publish particularly useless
ugly books.
Their main contribution seems to be splatter a cheesy rendering of that damn
bathtubtoypenguin everywhere, including giant as a page background = yikes!
They seem to think this will sell books.. what a tragedy-travesty :-(

The general series page design and layout is confusing, and poorly
conceived - a Prime example of 'fluff' or 'value-subtracted' publishing.

Content and editorial also appear to be inexperienced and rushed to market.
I feel sorry for any auhtor sucked into a contract with them and it is a
shaem there is even a Python title in that series at all.

caveat emptor [buyer beware]

- Jason
___________________________________________________________
Jason CUNLIFFE = NOMADICS['Interactive Art and Technology']

Don Tuttle <tuttledon at hotmail.com> wrote in message
news:9iOW5.55847$L26.16633720 at typhoon.southeast.rr.com...
>
> "Simon B." <sbrunning at bigfoot.com> wrote in message
> news:90fq39$jgq$1 at nnrp1.deja.com...
> > I came across a Python book this weekend that I haven't heard of
> > before - Programming with Python, by Tim Altom
> > (see <http://www.prima-tech.com/book.asp?id=3942>). Anyone have this?
> > Any good?
>
> IMO, this is the only bad Python book on the market.  Don't waste your
> money.
>
> Don





From tim.hochberg at ieee.org  Tue Dec  5 19:36:48 2000
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Wed, 06 Dec 2000 00:36:48 GMT
Subject: an apply question
References: <90jsj0$24o$1@front6.grolier.fr>
Message-ID: <ASfX5.52018$15.11095989@news1.rdc1.az.home.com>

"Alan Marchand" <malan at club-internet.fr> wrote in message
news:90jsj0$24o$1 at front6.grolier.fr...
> In Python 2.0:
>
> What am I doing wrong?

Well a string is not a function, and you are trying to call a string as a
function. So the error you're getting makes sense.

It's not entirely clear what you want to do; I see two possibilities: the
first is that you are trying to look up the name given by a string and call
it. This could be accomplished by modifying the second part of your program
as follows:

for key, name in MyMap.items():
    if key == "Key3":
        func = getattr(globals(), name) # look up the function
        func() # or apply( func, ( ) )


The other possibility is that you want to actually store the functions in
your map. In this case, you could use.

def func1 ( ):
     print "Here's func1"
def func2 ( ):
     print "Here's func2"
def func3 ( ):
     print "Here's func3"
def func4 ( ):
     print "Here's func4"

 MyMap = {
 'Key1' : func1,
 'Key2' : func2,
 'Key3' : func3,
 'Key4' : func4
 }
#...

Note that there are no quotes around the functions in MyMap and that the
functions come before they are referred to.


-tim

> #!/bin/env python
> #############
> MyMap = {
> 'Key1' : 'func1',
> 'Key2' : 'func2',
> 'Key3' : 'func3',
> 'Key4' : 'func4'
> }
>
> def func1 ( ):
>     print "Here's func1"
>
> items = MyMap.items( )
> for x in range( len(items) ):
>     key, func = items[x]
>     if key == "Key3":
>         apply( func, ( ) )
> -------------------------
> c:\devpy\python test.py
>
> Traceback (most recent call last):
>   File "test.py", line 17, in ?
>     apply( func, ( ) )
> TypeError: Call of non-function (type string)
>
> Thanks.
> Alan.
>
>




From robericv at my-deja.com  Wed Dec 27 00:26:52 2000
From: robericv at my-deja.com (birdman)
Date: Wed, 27 Dec 2000 05:26:52 GMT
Subject: Just starting out!
Message-ID: <92buip$kcp$1@nnrp1.deja.com>

I just started teaching myself to write code and wonder if Python is the
best way to go. Also, Is there a better book than "Learning Python" by
Mark Lutz and David Ascher available for someone at my level.


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


From piet at cs.uu.nl  Mon Dec 18 10:41:08 2000
From: piet at cs.uu.nl (piet at cs.uu.nl)
Date: 18 Dec 2000 16:41:08 +0100
Subject: launching a python program from GNU make (Win2K)
References: <918lp0$6q4$1@nnrp1.deja.com> <slrn93jn66.p9r.hat@se-46.wpa.wtb.tue.nl> <ug0jpzfue.fsf@roundpoint.com>
Message-ID: <uae9t68kr.fsf@cs.uu.nl>

I have the following setup on Win2K:
1. make:
    GNU Make version 3.77, by Richard Stallman and Roland McGrath.
    Copyright (C) 1988, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98
            Free Software Foundation, Inc.
    This is free software; see the source for copying conditions.
    There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A
    PARTICULAR PURPOSE.

    Report bugs to <bug-make at gnu.org>.
    I think this is the Mingw version, as I use mingw rather than cygwin.
2. Makefile:
all:
	python test.py
3. test.py:
print "Hello World"

And that works.
Even when I add a #!/bin/python to test.py (which is not the correct
location) and change the makefile line to:
all:
	 test.py

it works.
-- 
Piet van Oostrum <piet at cs.uu.nl>
URL: http://www.cs.uu.nl/~piet [PGP]
Private email: P.van.Oostrum at hccnet.nl


From mjjacob at my-deja.com  Wed Dec 13 11:39:45 2000
From: mjjacob at my-deja.com (mjjacob at my-deja.com)
Date: Wed, 13 Dec 2000 16:39:45 GMT
Subject: Problem with += (augmented assignment)
Message-ID: <9188og$q57$1@nnrp1.deja.com>

Hi folks!

I tried this under Python 2.0 running on Win 2k:

x = 0
while x <= 42:
	x += 1

Python always complains a Syntax Error: Invalid
Syntax on this:

    x += 1
       ^
SyntaxError: invalid syntax

Augmented assignment only seems to work when used
in first indentation level (script level).
Otherwise, I always get that syntax error, it
does not matter whether I am using if, while or
whatever.

Is there something terribly wrong with the +=
operator or is that all my fault?

Help appreciated.


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


From m.hadfield at niwa.cri.nz  Wed Dec  6 16:47:50 2000
From: m.hadfield at niwa.cri.nz (Mark Hadfield)
Date: Thu, 7 Dec 2000 10:47:50 +1300
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <3A2DD2CA.14A5D7E2@engcorp.com> <90l3j9$4bn$1@kapsel.intranet>
Message-ID: <976139270.161152@clam-ext>

"Piotr Kuchta" <piotr at itam.zabrze.pl> wrote in message
news:90l3j9$4bn$1 at kapsel.intranet...
> ...Dutchmen are very talented: common Dutchman speaks well two foreign
> languages, English and German.

Whereas common Australian speaks poorly one foreign language: English.

---
Mark Hadfield
m.hadfield at niwa.cri.nz  http://katipo.niwa.cri.nz/~hadfield/
National Institute for Water and Atmospheric Research
PO Box 14-901, Wellington, New Zealand





From gregj at pobox.com  Wed Dec  6 02:40:58 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Wed, 06 Dec 2000 07:40:58 GMT
Subject: Script vs Program
References: <y39V5.524$OC.159868@typhoon.aracnet.com> <bHnV5.96503$U46.3094705@news1.sttls1.wa.home.com> <jDFV5.102058$U46.3208014@news1.sttls1.wa.home.com> <3A27BAB5.CF6D060D@san.rr.com> <TI0X5.116132$U46.3743208@news1.sttls1.wa.home.com> <slrn92qqjl.tl0.tim@degree.ath.cx>
Message-ID: <e4mX5.117676$U46.3887083@news1.sttls1.wa.home.com>

"Tim Hammerquist" <tim at degree.ath.cx> wrote:

> To my understanding, Netscape included "JavaScript" in its web clients,
> whereas MSIE, ever stubborn, included it's own modified but similar
> "JScript."  JavaScript has been standardized by ECMA.  The standard is
> available http://www.ecma.ch/ under the code ECMA-262.

JavaScript (formerly LiveScript) was the scripting language in Netscape's
LiveWire server, then they implemented it in the browser and client-side
scripting was born. Microsoft implemented JavaScript in IE (along with their
own Visual Basic-derived VBScript) back when they didn't have dominant
browser share and had to stay compatible with Netscape. Microsoft's language
implementation is compatible with Netscape's (and now Microsoft's
implementation complies more or less with the ECMA standard, whereas
Netscape's doesn't). Internet Explorer's Document Object Model (DOM) is not
compatible with Netscape's, and that is the main source of incompatibilities
at the client side. I think MS had to call their implementation JScript for
legal reasons; Netscape owns the name JavaScript (they changed the name from
LiveScript to associate LiveScript with Java, and JavaScript books are still
mixed up in the Java section at most bookstores). In the Microsoft/Active
Server Pages community it's usually understood that "JavaScript" refers to
the client-side language and JScript refers to the server side. They are the
same language, but with different object models. Internet Explorer
understands both <script language="javascript"> and <script
language="jscript">, but Netscape Navigator doesn't recognize the JScript
variant, so you rarely see that designation in a web page.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From erno-news at erno.iki.fi  Tue Dec 12 20:02:59 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 13 Dec 2000 03:02:59 +0200
Subject: Spinning off time intensive procedures to speed CGI
References: <916bpn$a6i$1@nnrp1.deja.com>
Message-ID: <ku4s096sl8.fsf@lasipalatsi.fi>

In article <916bpn$a6i$1 at nnrp1.deja.com>, leejor at my-deja.com writes:

| I want to spin off these proceedures to
| allow the original program to imediately continue.

pid = os.fork()
if pid == 0:
   # child process
   try:
        save_log_file() 
   finally:
        os._exit(0)
else:
   # parent process (pid contains process-id of child process)
   [continue execution as usual]


you can probably also do this with threads, if you need your cgi
program to work on windows and unix. (though then you are excluding
platforms that don't support threads, incl. many unixes)

the trick is to close stdout when you are done with the thread that
talks to the browser, that will close the connection to the remote
ttp client. caveat emptor: i haven't tried this (recently enough
to remember, anyway).

  -- erno


From nickb at fnord.io.com  Tue Dec 19 12:42:00 2000
From: nickb at fnord.io.com (Nick Bensema)
Date: Tue, 19 Dec 2000 17:42:00 GMT
Subject: teekay
References: <x%D%5.106612$IP1.3577747@news1.giganews.com>
Message-ID: <F5N%5.108769$IP1.3655247@news1.giganews.com>

In article <x%D%5.106612$IP1.3577747 at news1.giganews.com>,
Nick Bensema <nickb at fnord.io.com> wrote:
>Tk would make a great baby name.

It just occurred to me that if your baby was named Tk and you wanted
to make him/her giggle, you could Tcl/Tk.

-- 
Nick Bensema <nickb at io.com>      ICQ#2135445 
==== ======= ==============
CREDIT 0     THY GAME IS OVER     NO CARRIER


From MUSAJADAKISS at aol.com  Sat Dec 16 07:17:34 2000
From: MUSAJADAKISS at aol.com (MUSAJADAKISS at aol.com)
Date: Sat, 16 Dec 2000 07:17:34 EST
Subject: (no subject)
Message-ID: <c.ea9ae12.276cb75e@aol.com>

please can you help me in completing this to search for a route using a* 
algorithim 
the file will be in text written and stored in notepad plase i will be glad 
if it is done as soon as you can thank you very much
Amsterdam:
    Berlin,650
    Brussels,197
    Calais,367
    Cologne,256
    Edinburgh,1093
    Hamburg,447
    Paris,510
Athens:
    Sofia,828
Belfast:
    Cork,416
    Dublin,167
    Kilkenny,277
Berlin:
    Edinburgh,1696
    Hamburg,285
    Prague,345
    Rostock,222
    Stuttgart,629
    Warsaw,606
Budapest:
    Burcharest,852
    Prague,537
    Sofia,790
    Vienna,242
Cork:
    Kilkenny,148
Dublin:
    Cork,259
    Kilkenny,119
    Edinburgh,346
Edinburgh:
    Copenhagen,479
Hamburg:
    Rostock,175
    Stuttgart,659
Istanbul:
    Athens,1145
    Burcharest,690
    Sofia,550
London:
    Brussels,333
    Dublin,430
    Cologne,508
    Edinburgh,608
    Paris,399
Madrid:
    Bercelona,617
    Lisbon,651
    Paris,1280
Munich:
    Berlin,594
    Cologne,580
    Frankfurt,398
    Paris,810
    Prague,388
    Vienna,430
Oslo:
    Copenhagen,590
    Hamburg,900
    stocholm,530
Paris:
    Brussels,320
    Cologne,495
Prague:
    Frankfurt,512
    Hamburg,652
    Oslo,1350
    Warsaw,616
    Vienna,295
Rome:
    Athens,1140
    Milan,606
Vienna:
    Berlin,640
    Warsaw,727
Warsaw:
    Moscow,1245
Zurich:
    Milan,292
    Munich,303
    Paris,592
    Vienna,743

this is the pprogramme the hueristic distance is not measured but you can use 
any distance you want
 
import sys, os
import string

##  global vars :
citysign = ':'
NL = '\n'
alphanum = string.letters + string.digits


## Main program
#
def main():
        "TUBEREAD : Main line!"

        global  citysign, NL
        args = sys.argv[:]          ##  gets command-line arguments
        na = len(args)
        print args[0],"[version 1.0]"
        print "command-line args. =",na
        if na < 2:
                print "normal usage:  python tuberead.py <textfile>"
                args.append(raw_input("please give input textfile name: "))
                
        na = len(args)
        
        textname = args[1]
        textfile =  open(textname,"r")
    if textfile == None:
                print "Can't open file:",textname
        
        lc = 0 ; sc = 0
        lines = {}  ##  empty dictionary (station names)
        citylist = []  ##  empty list (line names)
    thisline = "Zonk"
    citydic = {}
    
        while 1:
      textline = textfile.readline()
      if not textline:  break
      name = string.capwords(string.strip(textline))
      sc = sc + 1
      if citysign in name:
          newline = 1
          print "city = ", name
          thiscity = string.replace(name,citysign,'')
          if not citydic.has_key(thiscity):
              citydic[thiscity] = []
          ##  citylist.append(thiscity)
      else:
          x = string.split(name,',')
          ##print x
          c = x[0]
          ##print c
          if not citydic.has_key(c):
              lc = lc + 1
              citydic[c] = []
          d = eval(x[1])
          t1 = (c,d)
          citydic[thiscity].append(t1)
          ##  also append symmetric link :
          t2 = (thiscity,d)
          citydic[c].append(t2)
          
    citylist = citydic.keys()
    citylist.sort()
        print citylist
    print
    print len(citylist),"distinct cities read from textfile."
    print
    ## print citydic
    ##  I/O from user
    while 1:
        sn = raw_input("Enter city name: ")
        sn = string.strip(sn)
        if len(sn) < 1:  break  ##  user hits CRLF to quit loop
        sd = raw_input("Enter destination city name: ")
        if len(sd) < 1:  break  ##  user hits CRLF to quit loop     
        sl = citydic.get(sn,[])
        ra = len(sl)
        st = citydic.get(sd,[])
        mu = len(st)
        print sn,sl,ra
        print sd,st,mu
                
        
    return 0
        
        
            ##  do search

            ##  show result to user


##  Call main and attempt to honour exit status:
try:
    main()
except KeyboardInterrupt:
    sys.exit(1)



From boud at rempt.xs4all.nl  Mon Dec 18 01:24:31 2000
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 18 Dec 2000 06:24:31 GMT
Subject: Unicode and default encodings
References: <91g9e1$4t2$1@news1.xs4all.nl> <NVN_5.199$u17.33869@newsb.telia.net>
Message-ID: <91kaiv$k3l$1@news1.xs4all.nl>

Fredrik Lundh <fredrik at effbot.org> wrote:
> Boudewijn Rempt wrote:
>> I'm currently trying to work with Unicode strings, and
>> I was kind of wondering where sys.setdefaultencoding has
>> gone to - it's still mentioned in Misc/unicode.txt, but
>> it isn't there anymore ;-)

> You can use it in sitecustomize.py, but not in user code.
> See Lib/site.py for details.

Do I understand correctly then that there is no way to set a default
encoding per application? I mean, most of my users will live their
lives in latin1 land, but the application is meant to handle unicode
data exclusively.

Phil Thompson suggest sneakily keeping a reference to setdefaultencoding
from sitecustomize.py, but I keep wondering why I would need to do
something so hackish.


-- 

Boudewijn Rempt  | http://www.valdyas.org


From tim.one at home.com  Wed Dec  6 13:53:01 2000
From: tim.one at home.com (Tim Peters)
Date: Wed, 6 Dec 2000 13:53:01 -0500
Subject: Schwartzian Transform (was Re: Sort)
In-Reply-To: <200012061827.LAA03198@mail.cybermesa.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEKKICAA.tim.one@home.com>

[posted & mailed]

[Alex Martelli]
> The technique is named after its inventor Randall
> Schwartz, who thought it up in the context of Perl
> (but it works even better in Python:-).  The key
> idea is that, in scripting languages such as Perl
> and Python, 'sort' slows down a lot if each of its
> O(N log N) comparison involves interpreting script
> code, rather than just the fast, built-in "lexical
> comparison" that the language provides.

[Jay O'Connor]
> Why is it better in Python than Perl?
>
> This subject came up recently about Perl and list handling in
> another context, so I'm curious

Python's builtin sequence comparison is lexicographic (the sequence elements
are compared one at a time, left to right, until the first pair that "breaks
a tie" is found).  In Perl you need to fake that by hand, usually via
chaining together spaceship (<=> or cmp) operators.

besides-which-*everything's*-better-in-python<wink>-ly y'rs  - tim




From root at [127.0.0.1]  Mon Dec 25 02:21:03 2000
From: root at [127.0.0.1] (deadmeat)
Date: Mon, 25 Dec 2000 07:21:03 GMT
Subject: os.exec?
References: <LXw16.8561$KY1.23035@news1.rivrw1.nsw.optushome.com.au> <mailman.977727428.19738.python-list@python.org>
Message-ID: <zzC16.8593$KY1.23139@news1.rivrw1.nsw.optushome.com.au>

> Quite possibly, you'd want os.system, and os.popen.
> Or, you might want to work on a lower level with os.fork/os.exec/os.open.

I just figured out how to do this:

   child = os.popen(thecommand)
   while 1:
      line = child.readline()
      if not line:
         break
      sys.stdout.write(line)

which seems to work OK, so unless anyone has any better suggestions I'll go
with this..





From red5sf at my-deja.com  Sat Dec  9 18:45:51 2000
From: red5sf at my-deja.com (red5sf at my-deja.com)
Date: Sat, 09 Dec 2000 23:45:51 GMT
Subject: editing python path in pythonwin
References: <90ufli$8f3$1@nnrp1.deja.com>
Message-ID: <90ug7e$8s1$1@nnrp1.deja.com>

just found my problem,
it's the pythonwin path variable that i should be editing, not the
pythonpath one!

oops.
christian

In article <90ufli$8f3$1 at nnrp1.deja.com>,
  red5sf at my-deja.com wrote:
> for some reason, my edits to the pythonpath using the pythonwin
> (release 2.0) "EditPythonPath" feature aren't being made (on both
win98
> and 2000).
>
> Is this a known issue?  The only workaround i've been able to figure
is
> creating a PYTHONPATH environment variable in my autoexec.bat(on 98)
or
> in my user defined varable list (on 2000)
>
> do any of you know how to fix pythonwin to be able to set this?
>
> thanks,
> christian
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From chrisw at nipltd.com  Wed Dec  6 13:00:31 2000
From: chrisw at nipltd.com (Chris Withers)
Date: Wed, 06 Dec 2000 18:00:31 +0000
Subject: [Zope-dev] case insensitive sorts
References: <001301c05fab$feb51320$50f47118@cr582427a>
Message-ID: <3A2E7EBF.47AC8006@nipltd.com>

Andy McKay wrote:
> 
> Anyway Ive patched dtml-in and dtml-tree to do this sort on a ignore_case
> tag. Is this useful to anyone else?

yes! :-)

> And Ive thought of patching my Zope so
> this is the default behaviour.... what does anyone else think. The next
> thing to patch is ZCatalog...

To be honest, why not just mail Guido (not literally! ;-) and suggest a
change to python. IMNSHO, the default sort isn't as useful as it could
be because of this and should be made to behave as you describe rather
than like it is now. Pretty much every sort I do of strings in a python
project uses the sort method below insetad of the default. This is a
pain.

def _default_sort(x,y):
    return cmp(string.lower(x),string.lower(y))

Who thinks the default python sort should stay like it is? Who thinks it
should change? 
(reasons of course would be helpful, particularly if you want it to stay
like it is ;-)

cheers,

Chris

PS: With this being a general python issue, IMO, I've cc'ed to the
python list, hope I'm not out of order...

PPS: Is there such a thing as wishlist at python.org, if you see what I
mean?



From davidw at prosa.it  Wed Dec 27 07:38:54 2000
From: davidw at prosa.it (David N. Welton)
Date: 27 Dec 2000 13:38:54 +0100
Subject: sniffing btw my web browser and a server
References: <92aush$rhc$1@news.nuri.net> <270767A74DE6B86D.ED3C49CF7CD80055.4014344C6ECF85B8@lp.airnews.net> <mailman.977881630.26942.python-list@python.org>
Message-ID: <87ofxyxcm9.fsf@eugene.prosa.it>

Jan Dries <jdries at mail.com> writes:

> >I'm doing some sort of web client programming, and have to observe the
> >messages that my browser and web-servers transmit, being tired of guessing
> >what the messages'd be and looking up the references and the specifications.
> >
> >Anyone has done it in Python or knows how it could be done?
> >
> 
> If all you want is to see the trafic between a browser and server
> [ ... ]

I just use a system call tracer for this sort of thing.  It's kind of
crude, but it's fast, and I can see *exactly* what is happening
between my browser and server without putting anything between them.

-- 
David N. Welton
     Personal:           http://www.efn.org/~davidw/  
Free Software:           http://people.debian.org/~davidw/
   Apache Tcl:           http://tcl.apache.org


From rjh at 3-cities.com  Wed Dec 27 03:41:18 2000
From: rjh at 3-cities.com (Robert J. Harrison)
Date: Wed, 27 Dec 2000 00:41:18 -0800
Subject: gsfish@pcola.gulf.net
Message-ID: <3A49AB2E.C1DB4DBF@3-cities.com>

Guy ... thanks v. much for your hospitality during
our visit.  We both really miss you both and a big
plus for any possible move is to get closer to
you.

Anyway, I was wondering if you had made any progress
with your DAQbooks.

Robert

-- 
Robert J. Harrison (rjh at 3-cities.com)


From glenfant at equod.com.nospam  Fri Dec 29 18:06:26 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Sat, 30 Dec 2000 00:06:26 +0100
Subject: Is 100,000 entries big for a dictionary?
References: <92iug4$vqb$1@nnrp1.deja.com>
Message-ID: <92j54s$t0i$1@reader1.imaginet.fr>

Not for my desktop Sun 4500 with 16 Gb RAM 8))
For your machine:

megadict = {}
for x in xrange(0, 100000):
  megadict[str(x)] = 'depends on the stuff you put in the dict'

So what's the result ?

<rturpin at my-deja.com> a ?crit dans le message news:
92iug4$vqb$1 at nnrp1.deja.com...
> I'm in the process of architecting an application with medium sized data
> sets. Python dictionaries are tempting as part of the representation
> mechanism. How do they perform when they have 100,000 entries? Can you
> offer any other experience with this size data set in Python?
>
>
> Sent via Deja.com
> http://www.deja.com/



From aleaxit at yahoo.com  Tue Dec  5 04:26:34 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 5 Dec 2000 10:26:34 +0100
Subject: SAP R/3 automation with win32com
References: <90flpu$gpu$1@nnrp1.deja.com> <kl5n2tsfk3hud1c37gjo56290sueccuegk@4ax.com> <90i6n6$hg0$1@nnrp1.deja.com>
Message-ID: <90icc6022q9@news2.newsguy.com>

<obou at my-deja.com> wrote in message news:90i6n6$hg0$1 at nnrp1.deja.com...
    [snip]
> I don't know VB at all and don't really know how to translate VB
> functions calls to Python function calls. Especially when using COM.
> There are some strange differences. In my example, there are in VB
> function calls that takes 3 parameters where in Python I have to put
> only 2 parameters in the brackets and get the third as a return value.
> The salesOrd.DimAs function calls are such cases. In VB the return
> value is written as a third parameter and the function has no return
> value.
> Is there a way to identify which parameter have to stay in the brackets
> and which ones will be sent as return value?

Yes.

COM parameters are defined as [in], [out], [out,retval], or [in,out]
(there are also other attributes, but these are the ones that matter
here).  The [out,retval], if present at all, is only the last one.

In VB, a parameter [out,retval] appears as a return value, all others
appear as actual parameters to the function call (and not as returns).

In Python, parameters [in] and [in,out] appear as actual parameters
to the function call; parameters [out], [out,retval], and [in,out]
appear as return values (once you've run makepy on the type library
to let Python know what is what -- otherwise, [out] and [in,out]
can be troublesome).  Note that [in,out] is BOTH an actual parameter
(for the 'in' part) AND a return value (for the 'out' part).

[out] and [in,out] are not all that common in well-designed COM
(Automation) type-libraries, but, they do happen at times.

The VB Object Browser does not clearly display the attributes of
the parameters.  *However*, another precious tool is called "OLE
View" and should be also part of VB's higher editions (and I
think oleview.exe is also one of the tools you can download for
free as part of Microsoft's "Platform SDK").  THAT one, OleView,
lets you display any type-library in terms of IDL, the Interface
Definition Language, which includes the [in], [out], etc,
notations.  It's a very good tool for exploring COM stuff.  Of
course, a component's docs SHOULD include the non-trifling
details of what parameters are input ones, which are output
ones, etc, etc, but, sometimes, such little crucial things
drop by the wayside... and OleView can save your life then!-)

Alternatively, if you'd rather read Python than IDL, look at
the source-file that makepy.py generates for you when you 'hook
up' to a type-library.  It's placed in win32com\genpy, and its
filename is the GUID of the library; in the Python classes it
builds to correspond to each COM interface, it also clearly
separates and distinguishes between input and output parameters
of each method.


Alex





From fpm at u.washington.edu  Fri Dec 29 18:25:33 2000
From: fpm at u.washington.edu (Frank Miles)
Date: 29 Dec 2000 23:25:33 GMT
Subject: Python for prototyping of the user interface
References: <20001228.17165633@mis.configured.host> <92ir54$ftn$1@news1.xs4all.nl>
Message-ID: <92j6hd$k82$1@nntp6.u.washington.edu>

In article <92ir54$ftn$1 at news1.xs4all.nl>,
Boudewijn Rempt <boud at rempt.xs4all.nl> wrote:
>Volkmar Busch <vbusch at gmx.de> wrote:
>> Hi,
>
>> I took my first steps in using Python, and I really like the language. I 
>> read a lot about using Python for prototyping. As a basis for discussion 
>> with our users we use GUI prototypes. Building complex user interfaces in 
>> Python to me seems quite complicated, because
>
>> - I am missing a Gui builder
>> - I am missing controls like notebooks and tables
>
>> So I'd like to asked the experienced Python users how they solve this 
>> problem.
>
>I use PyQt - Qt comes with Qt Designer, a great gui designer that 
>generates xml descriptions of your design; this then is translated
>by pyuic to python. I subclass the result (which means I can 
>regenerate whenever I want), and use it. Quite simple, really. And
>all modern conveniences are available in PyQt, from icon boxes to
>Unicode support. And to my surprise, the large PyQt application I've
>just finished works without changes on Windows...

Python and Glade seem to offer these features, though perhaps without
the same degree of polish.  Qt appears to have a significant cost
if you're going to run on Windows :( [at least that's my reading of
TrollTech's FAQ].

	-frank
-- 


From roy at panix.com  Sat Dec 30 17:23:12 2000
From: roy at panix.com (Roy Smith)
Date: Sat, 30 Dec 2000 17:23:12 -0500
Subject: Great New Book
References: <t4o1vr8hktk531@corp.supernews.com> <92ico9$jde$1@panix2.panix.com> <92ig6k02apk@news2.newsguy.com> <hYe36.204059$U46.6288939@news1.sttls1.wa.home.com> <slrn94soh7.hd.amk@207-172-111-203.s203.tnt1.ann.va.dialup.rcn.com>
Message-ID: <roy-589B2A.17231230122000@news.panix.com>

akuchlin at mems-exchange.org wrote:
> for technical books thickness may indicate that the author can't boil 
> things down to their essence very well.

Absolutely.  I'm horrified at the number of 600, 700, 800, or more page 
programming books out there.  Most of them are just bloated.

I learned C from the original K&R.  The book was about a 1/2 inch thick.  
Probably no more than about 150 pages.  It didn't try to be all things to all 
people, it just tried to teach you C.  And it did a very good job.


From R.Brodie at rl.ac.uk  Mon Dec 11 14:02:30 2000
From: R.Brodie at rl.ac.uk (Richard Brodie)
Date: Mon, 11 Dec 2000 19:02:30 -0000
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A350405.B8F09EC7@gte.net> <9131vd0pnq@news2.newsguy.com>
Message-ID: <9138c4$hrs@newton.cc.rl.ac.uk>

"Alex Martelli" <aleaxit at yahoo.com> wrote in message news:9131vd0pnq at news2.newsguy.com...

> What do you claim is "barbarous", "grammatically faulty",
> etc, about using "while 1:" to express "loop until
> otherwise terminated" in Python?

Speaking for myself (of course), I find it somewhat unaesthetic
to have a loop construct that doesn't document what its terminating
condition is.

The  congruence argument is an interesting one but I'm not sure
entirely to the point. I could say that I prefer:

if (condition):
    do x

rather than a construct like

    do x:
if condition.

The former let's you abstract part of the problem.

It's a valid point on the usability of the language. I'm not sure there
is a good answer to the original one though. If there were a better
pseudocode notation someone would be using it by now. Maybe just
while 1: # Loop condition

I suppose for the specific case we came in on, having an xreadlines
function (by analogy with xrange) has been discussed and rejected
before now.






From DavidL at NOraqiaSPAMMY.com  Tue Dec 19 14:23:19 2000
From: DavidL at NOraqiaSPAMMY.com (David Lees)
Date: Tue, 19 Dec 2000 14:23:19 -0500
Subject: FIFO in python?
References: <3A3FAB7A.DC63610C@NOraqiaSPAMMY.com>
Message-ID: <3A3FB5A7.9EA37CAF@NOraqiaSPAMMY.com>

Actually, I just found that if I try to write to the fifo (which is
SAMBA mounted) from the NT side, I get an error:
IOError: (0, 'Error')

Meanwhile on the Linux, side, our local perl guru had no trouble getting
a pair of perl scripts to pass strings via the same fifo that I am
unable to open when I run Python from Linux.

Presumably this is some kind of configuration problem with python?

david lees


David Lees wrote:
> 
> I am trying to learn how to use FIFO and am unable to open a fifo under
> Linux RedHat 6.2.  I hang on the open (open(fn,'rw'), even though I have
> done a successful check for file existence (os.path.exists(fn)).  The
> same code works under NT.  Any thoughts on what is going on?
> 
> Thanks in advance,
> 
> david lees
> 
> here is my code:
> -------------------------
> import string, os, sys
> 
> def OpenFileRW(n):
>     if os.path.exists(n):
>         fn=open(n,'rw')
>         return fn
>     else:
>         print 'Filename '+name+' does not exist'
>         sys.exit(-1)
> 
> fifoName='junkfifo'
> fifo=OpenFileRW(fifoName)
> fifo.close()
> print 'done'


From db3l at fitlinxx.com  Wed Dec 13 16:16:06 2000
From: db3l at fitlinxx.com (David Bolen)
Date: 13 Dec 2000 16:16:06 -0500
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <kuzoi15d60.fsf@lasipalatsi.fi> <Pine.BSF.4.10.10012122310510.48551-100000@tcfreenet.org>
Message-ID: <u7l54ggyx.fsf@ctwd0143.fitlinxx.com>

Aaron Jon Reichow <reichowa at tcfreenet.org> writes:

> On 13 Dec 2000, Erno Kuusela wrote:
> 
> > the thing that struck me as a little foreign was the apparent
> > isolation of the smalltalk image from the rest of the operating system
> > and the filesystem. i want to be able to grep code!
> 
> While I'm not a proponent of Smalltalk isolationism, grep is a useless
> tool (more or less), simply an artifact, ye, a symptom of the
> sickness known as file-based development.

I'm not sure I'd call it a "sickness" since any development model is
just that - a model - it may be better or worse than some other model,
but at least for me, a file-based model works well in many cases.

My biggest problem with the integrated models (with database or other
repository back-ends) tends to be that often they aren't yet fleshed
out enough, or in many cases flexible enough, to replicate all the
functionality I can get out of a file based system with a current
collection of file based tools that have had lots of time to evolve
for efficiency and flexibility.

With file-based environments, I generally feel safe that if the
environment itself doesn't support something, there are enough tools
that understand the back-end repository (e.g., files) that I can get
what I want even if I have to go outside the integrated system.  That
may not be possible with a closed back-end repository.

As was noted elsewhere in this thread, and to tie a little more back
into Python, this can also be an issue with Zope.  For example, Zope
may provide grep-like functionality that works through the object
store, but it may be difficult to replicate what in a filesystem might
be a grep piped into an awk to produce summary information.  Now maybe
the right answer is that the same functionality should be done in some
other manner with Zope, but it may also be a point of needed
functionality just not yet being supported by the development model.
(I know in my case there's also a little "queasiness" just because of
a feeling of lack of control over the store, which I think is just an
side-effect of my primary file-based focus for so long).

But if the repository model is flexible enough to permit the eventual
creation of the smaller tools and then their coupling on an as-need
basis (which Zope may well be for example), then it's more a question
of a minimal set of functionality to feel comfortable to those of us
more used to the file system and utilities and tools that we've gained
proficiency in - that tool proficiency generally leading to
development efficiency.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From tbryan at python.net  Fri Dec 29 00:08:02 2000
From: tbryan at python.net (Thomas A. Bryan)
Date: Fri, 29 Dec 2000 00:08:02 -0500
Subject: Python Database Interfaces - Any Standard?
References: <RUo16.204045$RP.37382354@news1.rdc1.md.home.com> <kuk88pu0cd.fsf@lasipalatsi.fi> <Gwu16.205577$RP.37521696@news1.rdc1.md.home.com> <3A483B53.ACB0C04D@python.net> <p6qu27pe5cn.fsf@informatik.hu-berlin.de>
Message-ID: <3A4C1C32.B648F571@python.net>

Martin von Loewis wrote:
> 
> "Thomas A. Bryan" <tbryan at python.net> writes:
> >
> > To change to PostgreSQL, you'd have to change the import and the connection.
> 
> If you anticipate change to another database, you write
> 
> import DCOracle as DB
> dbh = DB.Connect( connect_string )
> query = dbh.cursor()
> query.execute( query_string )
> 
> For PostgreSQL, you only need to write
> 
> import pg as DB

Thanks, but note that the connect_string probably still needs to change 
even if one tries to keep the database details (machine, instance, 
port, username, password) the same because the API doesn't specify the 
connection string, and I'm not sure how many modules follow its 
recommendation.  I hope to find out next year when I start playing 
with PostgreSQL.  :)

> Now, it is common that Python provides a wrapper module, like the os
> module or the anydbm module. However, there was no demand for such a
> wrapper in Python, yet.

It seems like there was some discussion of a generic wrapper on the db-sig 
this year, but I don't know what became of it.  

because-I-wasn't-paying-attention-at-the-time-ly yours
---Tom


From empii at cc.tut.fi  Thu Dec 14 14:44:27 2000
From: empii at cc.tut.fi (Pekkarinen Mikko)
Date: 14 Dec 2000 19:44:27 GMT
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <3A2DD2CA.14A5D7E2@engcorp.com> <90l3j9$4bn$1@kapsel.intranet> <udzoi9cedt.fsf@box.home.de>
Message-ID: <91b7ur$msd$1@baker.cc.tut.fi>

Jochen K?pper <jochen at pc1.uni-duesseldorf.de> hat geschrieben:

: Where's the foreign language;-))

You should have used a search engine before asking. Try:

http://www.google.com/search?q=language

:)

-- 
mikko pekkarinen empii at cc.tut.fi http://www.cs.tut.fi/~empii/
What is a rhetorical question?


From erno-news at erno.iki.fi  Thu Dec 28 22:34:14 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 29 Dec 2000 05:34:14 +0200
Subject: Python vs PHP (& Java?)
References: <92f3kp01cqe@news1.newsguy.com> <mailman.977897649.4877.python-list@python.org> <slrn94l86m.a3h.tim@degree.ath.cx> <mailman.978021012.21874.python-list@python.org> <slrn94nk02.jk.amk@207-172-57-128.s128.tnt2.ann.va.dialup.rcn.com>
Message-ID: <kuelyrrjd5.fsf@lasipalatsi.fi>

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

| On 28 Dec 2000 16:22:03 -0000, Moshe Zadka <moshez at zadka.site.co.il> wrote:
|| (BTW: Are there plans for quixote/mod_snake integration? That'd
|| be ultra-cool: one of the benefits of HTML::Mason is that it
|| integrates really well with mod_perl)

| Not at this point in time; we use FastCGI, which has been an adequate
| performance boost so far.

sadly the mod_fastcgi license isn't very free. i wonder if there's
another module with a nicer license somehwere?

  -- erno


From m.faassen at vet.uu.nl  Sat Dec 30 20:59:34 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 31 Dec 2000 01:59:34 GMT
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <slrn94shbe.150.grey@teleute.rpglink.com> <X8s36.2731$Qb7.305339@newsb.telia.net> <slrn94smds.150.grey@teleute.rpglink.com> <FOt36.17$vd1.2224@e420r-atl2.usenetserver.com> <slrn94t1vj.150.grey@teleute.rpglink.com>
Message-ID: <92m3u6$cfm$1@newshost.accu.uu.nl>

[snip flamewar type stuff]

And now, we will return to our regularly scheduled flame-free newsgroup.
This was all just a misunderstanding, right? We can disagree without
flaming.

And-if-not-the-PSU-will-come-and-get-you-ly yours,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From erno-news at erno.iki.fi  Tue Dec 12 20:21:27 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 13 Dec 2000 03:21:27 +0200
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu>
Message-ID: <kuzoi15d60.fsf@lasipalatsi.fi>

In article <915qlh$89fa$1 at nntp6.u.washington.edu>, "Russell E. Owen"
<owen at astroNOJNK.washington.edu.invalid> writes:

| I think of Python as a very advanced scripting language.

why a "scripting languagage"?

| Much cleaner and more regular than Perl, not as powerful as Smalltalk. It
| is a great language and should appeal to many Smalltalkers for
| smaller tasks.

i have only had a peek at smalltalk, namely Squeak. it seemed very
nice.

the thing that struck me as a little foreign was the apparent
isolation of the smalltalk image from the rest of the operating system
and the filesystem. i want to be able to grep code!

but i guess this is just an artifact of that particular
implementation. (it was the only free one that i came across however)

| On the other hand, it is not as powerful, clean and regular as 
| smalltalk. Limitations include:
| - The language is somewhat cluttered and irregular because it has 
| procedures, functions and methods. Hence arbitrary decisions had to be 
| made in some areas as to how to implement things, which can lead to some 
| hunting through the manuals.

hmm? there are only functions and methods, if you want to make the
distinction. the only difference is that if you put a function
definition inside a class definition, it will get passed the class
instance as first argument.

| - garbage collection is reference counted, though now enhanced in Python 
| 2.0 so it can handle most cases of circular references

i don't think there are any situations in which the 2.0 system
would not handle them.

in fact the refcounting + "real" gc combination is much nicer than
pure gc, because of the predictable finalisation behaviour.

| - Mediocre GUI support (a mediocre interface to Tk and a different 
| interface to wxWindows that isn't available on my platform--Mac)

(i prefer python-gtk/python-gnome)

| - Mediocre IDE. I hope this will be improved over time, along with the 
| GUI support.

emacs!

| - Code indentation is significant. This tends to annoy folks at first, 
| but almost everybody comes to find it reasonable or good after a short 
| time. It shocked me at first, but I like it just fine.

look at it this way: you indent your code anyway. this way
you don't need to communicate the information about block
structure twice to the computer. 

| It's an absolutely stunning language for free, and although not as fast 
| as Perl

(it's as fast in my experience)

| still quite reasonable for basic work. Every Smalltalker who 
| needs a smaller-scale language for small projects should look at it, but 
| don't expect it to be a substitute for Smalltalk.

  -- erno


From aahz at panix.com  Sun Dec 17 09:41:30 2000
From: aahz at panix.com (Aahz Maruch)
Date: 17 Dec 2000 06:41:30 -0800
Subject: Interrupting a thread...
References: <91hqgk$pu0$1@bob.news.rcn.net>
Message-ID: <91ijaq$4pf$1@panix6.panix.com>

In article <91hqgk$pu0$1 at bob.news.rcn.net>,
David Allen <s2mdalle at titan.vcu.edu> wrote:
>
>I'm writing a program which does some networking, and I have a thread
>that is devoted to grabbing things off of the network so that the
>program won't hang while it is reading data.
>
>The problem is, I also occasionally need to interrupt this data
>fetching operation, so I need a way to either raise an exception in the
>other thread, or otherwise signal to it to stop what it's doing.  Since
>the download thread calls other functions that are not sensitive to the
>data in the original object, I can't check to see if some value has
>changed each time I'm reading in a loop.

Sounds like you should split the download thread into two threads, with
critical section locks to prevent contamination of the data.  The second
thread can use threading.Event() to learn when it's supposed to do work.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

"Using LOCs as a measure of programmer productivity is like estimating the 
productivity of a welder by the amount of acetylene used."  


From robin at jessikat.fsnet.co.uk  Thu Dec 14 09:11:30 2000
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Thu, 14 Dec 2000 14:11:30 +0000
Subject: split weirds
References: <QI2l1FAXBMO6EwD6@jessikat.demon.co.uk>
Message-ID: <GZXwhJASUNO6EwxV@jessikat.demon.co.uk>

In article <QI2l1FAXBMO6EwD6 at jessikat.demon.co.uk>, Robin Becker
<robin at jessikat.fsnet.co.uk> writes
>Can somebody explain why strip behaves differently with and without an
>argument?
>
>>>> from string import split
>>>> split(' ')
>[]
>>>> split(' ',' ')
>['', '']
>>>> split('  ')
>[]
>>>> split('  ',' ')
>['', '', '']
>>>> 
>
>with the default whitespace arg seems as though a run of whitespace is
>being treated as a single character.
several people have pointed out that arbitrary whitespace in the docs
means that so the no arg case really is special. I still find it a bit
strange that

split(' word ') ==> ['word'], but then the docs use the term separated
for the no arg case which can't be done for absent words. 
-- 
Robin Becker


From dale at out-think.NOSPAMco.uk  Mon Dec  4 06:48:59 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Mon, 04 Dec 2000 11:48:59 +0000
Subject: General default exception handler
Message-ID: <lj0n2to9vs8bmv3oo49mqgoc6h7esh8t37@4ax.com>

We have a substantial set of COM objects in Python that run a Web site.

Currently, if an exception is raised, the user gets an unfriendly message in their browser and we
may never know about it.

I want to log all unhandled exceptions but I'd rather not have to put a try/except block around
every entry point - because there are dozens of them.

Is there a general way to catch all unhandled exceptions so I can log them and pass back something
nice and soothing to the user?

I'm planning to use the traceback module to record the exception info.

Thanks
--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants




From lpointal at planete.net  Sat Dec  9 16:27:39 2000
From: lpointal at planete.net (Laurent Pointal)
Date: 9 Dec 2000 21:27:39 GMT
Subject: creating an odbc link with python
References: <gSVX5.55717$yR4.1401853@news1.rdc1.tx.home.com>
Message-ID: <9005E4099lpointalplanetenet@195.134.198.12>

slhath at DownWithSpam.home.com (Scott Hathaway) wrote in
<gSVX5.55717$yR4.1401853 at news1.rdc1.tx.home.com>: 

>Hello Everyone,
>
>I am using python to create an installation program for a web app that
>we have at our company.  The web app uses an access database that is
>hooked up through odbc.  How can I programatically create the dsn in
>python (I guess I need to edit the windows registry)?

If your question is about manipulating registry, see Win32 python 
extension, registry manipulation functions are here.

If its about DSN structure on windows, maybe its not the right newsgroup. 
See comp.database.* if a newsgroup has better chance to fill your request, 
or see microsoft knowledge base.

Good luck.

Laurent.


-- 
Laurent POINTAL - lpointal at planete.net


From emile at fenx.com  Tue Dec 19 11:34:44 2000
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 19 Dec 2000 08:34:44 -0800
Subject: Walk through dictionary keys?
References: <3A3F8B89.1DD9E01F@schlund.de>
Message-ID: <3A3F8E24.8040307@fenx.com>

I'm not sure when or how long this has been in, but the cvs version of 
dict's has a popitem method that looks like it does what you want.

d = {1:11,2:22,3:33,4:44}
for i in xrange(len(d)):
     print d.popitem()

HTH,

Emile van Sebille
emile at fenx.com

Carsten Gaebler wrote:

> Hi!
> 
> Is there a way of 'walking' through the keys of a dictionary? If I have
> a large dictionary I'd like to prevent 
> 
> for key in d.keys():
>     do_something
> 
> because this makes a copy of all the keys. I'm looking for something
> like Perl's "each".
> 
> Regards
> Carsten.



From arhiv at freemail.org.mk  Thu Dec  7 07:53:19 2000
From: arhiv at freemail.org.mk (Damjan)
Date: Thu, 7 Dec 2000 13:53:19 +0100
Subject: Really, Really Confused...
In-Reply-To: <NCBBLGDECCBLJILGIFMPIECPCJAA.ben@semio.com>; from ben@semio.com on Wed, Dec 06, 2000 at 05:49:38PM -0800
References: <NCBBLGDECCBLJILGIFMPIECPCJAA.ben@semio.com>
Message-ID: <20001207135319.A1612@localhost.localdomain>

> Hi Folks,
> 
> I've encountered a very strange problem that'll take a little space to
> explain.  I hope some of you will brave your way to the end though because
> I'm well and truly stumped by this one.

Let's simplify your examples:

class A:
	myDict={}
	def __init__(self):
		if self.myDict.has_key('foo'):
			raise 'Already set'
		self.myDict['foo']='bar'

class B:
	myString=None
	def __init__(self):
		if self.myString != None
			raise 'Already set'
		self.myString = 'bar'

a1=A()
a2=A()
b1=B()
b2=B()

Here is what happens. 

'myDict', in Class A, is a Class wide dictionary (so called 'static' in some
languages). It's created when the class is defined, and here it can be 
accessed through either of 'A.myDict' or 'a1.myDict', it's the same object.

'myString' = None, also creates a Class wide None object, and that object is
accesed in the 'if self.myString...' statement. But later a string object is
created ('bar') and referenced to 'self.myString'. Now, 'self.myString' hides
'B.myString'.

Check this:
While 'a1.myDict', 'a2.myDict' and 'A.myDict' is the same object, 
'b1.myString', 'b2.myString' and 'B.myString' are diferent objects 
(only two of them actually being strings).

Perhaps, what you needed was this:

class C:
	myDict = {}
	def __init__(self):
		# create a new dictionary object and
		# reference it to self.myDict
		self.myDict = {'foo':'bar'}	
c1=C()

Now, 'C.myDict' and 'c1.myDict' are diferent objects.

 
-- 
Damjan Georgievski		|           ?????? ???????????
Skopje, Macedonia		|           ??????, ??????????



From DanielK at aracnet.com  Mon Dec 25 12:40:47 2000
From: DanielK at aracnet.com (Daniel Klein)
Date: Mon, 25 Dec 2000 09:40:47 -0800
Subject: soundex (revisited)
References: <Var16.263$LU6.109277@typhoon.aracnet.com> <YPE16.190877$U46.5871952@news1.sttls1.wa.home.com>
Message-ID: <JyL16.269$LU6.117754@typhoon.aracnet.com>

Thanks for the critique Greg.

Looks like I just had the one bug with consecutive letters. I was originally
comparing the next with previous soundex codes (rather than next/previous
alpha character) but it was not handling names like 'LLOYD'. I also took
your advice on char.isalpha() and ord('A').

Thanks again,
Daniel Klein
Beaverton, Oregon USA


"Greg Jorgensen" <gregj at pobox.com> wrote in message
news:YPE16.190877$U46.5871952 at news1.sttls1.wa.home.com...
> "Daniel Klein" <DanielK at aracnet.com> wrote in message
> news:Var16.263$LU6.109277 at typhoon.aracnet.com...
> > After seeing the post from several days ago on soundex, I gave it whirl
to
> > see if I could come up with something different (and possibly better),
> > following the rules laid down by Knuth:
> >
> > def get_soundex(name, digits = 3):
> >     soundexcodes = "01230120022455012623010202"
> >     #               ABCDEFGHIJKLMNOPQRSTUVWXYZ
> >     instring = name.upper()
> >     soundex = instring[0]
> >     last = soundex
> >     instring = instring[1:]
> >     for char in instring:
> >         if 65 <= ord(char) <= 90:
> >             sx = soundexcodes[ord(char) - 65]
> >             if int(sx) and char != last:
> >                 soundex += sx
> >                 last = char
> >     if len(soundex) < (digits + 1): soundex = (soundex + ("0" * digits))
> >     return soundex[:digits + 1]
>
> I see a few problems, mainly in the handling of consecutive consonants.
You
> are checking for consecutive characters, but the Soundex algorithm
specifies
> that consecutive character codes be treated as a single code. Both 'mm'
and
> 'mn' are considered consecutive codes because both 'm' and 'n' are coded
as
> 5.
>
> You can (and probably should) use the isalpha() string method to check for
> alpha characters, rather than the 'magic numbers' 65 through 90. Likewise
> ord(char) - ord('A') is a bit more clear.
>
> Here's a version I wrote. I'm open to any criticisms, suggestions, etc. I
> compared my version to the module announced here a while back (I think
mine
> is a lot more readable; it is certainly shorter). I also compared it to a
> Perl version I found and I think my implementation is more robust and
> smaller.
>
> def soundex(name, len=4):
>     """ soundex module conforming to Knuth's algorithm
>         implementation 2000-12-24 by Gregory Jorgensen
>         public domain
>     """
>
>     # digits holds the soundex values for the alphabet
>     digits = '01230120022455012623010202'
>     sndx = ''
>     fc = ''
>
>     # translate alpha chars in name to soundex digits
>     for c in name.upper():
>         if c.isalpha():
>             if not fc: fc = c   # remember first letter
>             d = digits[ord(c)-ord('A')]
>             # duplicate consecutive soundex digits are skipped
>             if not sndx or (d != sndx[-1]):
>                 sndx += d
>
>     # replace first digit with first alpha character
>     sndx = fc + sndx[1:]
>
>     # remove all 0s from the soundex code
>     sndx = sndx.replace('0','')
>
>     # return soundex code padded to len characters
>     return (sndx + (len * '0'))[:len]
>
>
> --
> Greg Jorgensen
> Deschooling Society
> Portland, Oregon, USA
> gregj at pobox.com
>
>




From pbjorn at uswest.net  Wed Dec 13 12:31:38 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Wed, 13 Dec 2000 10:31:38 -0700
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <29lZ5.138780$U46.4638547@news1.sttls1.wa.home.com> <uhf49kf8r.fsf@roundpoint.com> <qXFZ5.142740$U46.4719533@news1.sttls1.wa.home.com>
Message-ID: <3A37B27A.B18E8497@uswest.net>

Greg Jorgensen wrote:

> "Ben Hutchings" <ben.hutchings at roundpoint.com> wrote in message
> news:uhf49kf8r.fsf at roundpoint.com...
> > "Greg Jorgensen" <gregj at pobox.com> writes:
> > <snip>
> > > A flame war broke out a few years ago in comp.lang.c++ over which
> > > was better: for ( i=0; i < N; i++ ) or for (i=0; i < N; ++i ). The
> > > result is exactly the same, and K&R use the first (postincrement)
> > > idiom, but it turns out that ++i is usually a tiny bit more
> > > efficient than i++.
> > <snip>
> >
> > For iterators of built-in types (like int) it's unlikely to matter,
> > but for iterators of user-defined types the latter may be less
> > efficient as it's harder for the compiler to optimise the code for the
> > overloaded post-increment operator.
>
> The problem is the temporary variable. The value of i++ is i before the
> increment, so the compiler has to save the current value before
> incrementing:
> [snip]
> The difference is negligible in most cases, and modern compilers easily
> optimize the temporary out anyway. I was using the flame war over this as an
> example of programmers defending a familiar idiom.

In these cases the compiler doesn't need a temporary variable. If you look at
the (unoptimized) output for your favorite compiler you'll see that they're
identical [I only tested it on msvc, but if it can do it, I'm assuming other
compilers can also <wink> ]

As the previous poster pointed out, it is much more difficult for the compiler
to figure out whether it can optimize something like:

   for (vector<int>::iterator i = v.begin(); i != v.end(); i++)

although msvc gets this "right" also... So the moral is, while it doesn't hurt
to have coding standards saying you should use pre-increment whenever possible,
most people will guess wrong about what the compiler will actually do.

now-back-to-non-infinite-infinite-loops'ly y'rs
-- bjorn




From glenfant at equod.com.nospam  Wed Dec 27 05:17:42 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Wed, 27 Dec 2000 11:17:42 +0100
Subject: Python vs PHP (& Java?)
References: <mailman.977897649.4877.python-list@python.org>
Message-ID: <92cfbh$nlr$1@reader1.imaginet.fr>

Some answers in short:
You can do in Python as in PHP (see
http://sourceforge.net/projects/modsnake/ ).
But, whatever language you choose (even Python), it's not a "clever"
approach in application centric Web servers because you mix HTML
(presentation layer) and scripts (data process layer) in the same file.
This site programmation approach leads to some mess in big projects
involving both programmers and web designers.
Keep this kind of approach for small application sites.

Java vs Python... well.

My pro java half brain says that it's a statically typed language (some bugs
that raise at compile time in java when they raise at run-time in python).
Prefer java if the security is a critical issue.
You can make transparent distributed web software in Java (applets + RMI)
because most browsers include a Java engine.
If you got lots of $$, there is number of solution providers based on common
java API (J2EE, JMS...) when Zope is the only app server solution based on
Python.

My pro python half brain say's it's twice faster to go from design to a full
featured application.
You can't make system specific code in Java.
You can do in Python (almost) all you can do in Java with less computer
resources.

In the project I'm working on, there's a mix of Java (WebObjects application
server) and Python.

We use WebObjects for all applications with HTML GUI, when we use Python for
batch processes that do not need to be fast and that are executed behind a
Firewall.



From dale at out-think.NOSPAMco.uk  Mon Dec  4 11:41:34 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Mon, 04 Dec 2000 16:41:34 +0000
Subject: Printable time
References: <7m9n2tcp5ff10o97sr1d57qokakjqsa2m9@4ax.com> <3A2BA9A3.63214210@schlund.de>
Message-ID: <u7in2t8kqrqtr74er1o69c0e9oo66l2toh@4ax.com>

Carsten Gaebler <cg at schlund.de> wrote:

>Dale Strickland-Clark wrote:
>
>> Is there a more succinct way of getting the current local time as text than:
>> time.strftime('<format>', time.localtime(time.time()))
>
>
>time.ctime(time.time())
>
>
>cg.

Thanks. 

It's always struck me as a bit strange that none of the time functions will use the current time as
a default if none is specified.

--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants




From teyc at bigfoot.com  Tue Dec 26 05:03:51 2000
From: teyc at bigfoot.com (Chui Tey)
Date: Tue, 26 Dec 2000 20:03:51 +1000
Subject: Season's greetings (was Re: Some Python 2.1 ideas)
References: <mailman.977765348.13376.python-list@python.org> <928otj02l39@news1.newsguy.com> <oQX16.71486$df5.2622172@news1.crdva1.bc.home.com>
Message-ID: <929qhl$p36$1@bunyip.cc.uq.edu.au>

gdchart doesn't do scatter plots yet.
Jpython's example scripts includes the chart applet rewritten in python and
awt






From stefanb at mail.vega.bg  Fri Dec  8 10:55:38 2000
From: stefanb at mail.vega.bg (Stefan Boiadjiev)
Date: Fri, 08 Dec 2000 17:55:38 +0200
Subject: Refactoring browser
Message-ID: <3A31047A.A7E880EF@mail.vega.bg>

Hi,
Do in python world exist something similar to Smalltalk Refactoring
browser ?
stefan boiadjiev



From mliss at exmail.de  Sat Dec  9 17:15:19 2000
From: mliss at exmail.de (Mirko =?iso-8859-1?Q?Li=DF?=)
Date: Sat, 09 Dec 2000 23:15:19 +0100
Subject: Help/example of bit testing
References: <90llq7$8fc@dispatch.concentric.net>
Message-ID: <3A32AEF7.59E52A92@exmail.de>

Bryan Webb wrote:
> I am reading a file (binary)of 1024 char recs (bitmaps) . I can read the
> file , bit im having trouble finding examples of testing bits. I need to
> look at each bit of the above record.

Supposedly, you already use struct.unpack() from the struct-module to
convert
chunks of your binary string data into numbers.

If speed and memory doesn't matter much, you might want to store
the bits in a user-defined list-object, for example:

>>> import UserList
>>> class Bitfield(UserList.UserList):
...   def __init__(self,number):
...       self.data = []
...       while number:
...          self.data.append(number & 1)
...          number = number >> 1
...   def __getitem__(self, key):
...          if key < len(self.data):
...             return self.data[key]
...          else:
...             return 0 
... 
>>> bf=Bitfield(0xc02)
>>> bf
[0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1]

bf[0] is the least significant bit.
Because of the redefined __getitem__() method, you don't need to worry
about
illegal indices:

>>> len(bf),bf[15],bf[13],bf[12],bf[11]
(12, 0, 0, 0, 1)

You might as well redefine the __getslice__() method if you need to look
at 
various bits at once.


I hope you like this,

Mirko Liss


From aleaxit at yahoo.com  Tue Dec 26 03:49:15 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 26 Dec 2000 09:49:15 +0100
Subject: classes (was Re: Same again please for OOP)
References: <928m5l02iuo@news1.newsguy.com>, <928569$6ckoi$1@ID-11957.news.dfncis.de>, <92708d0g4g@news1.newsguy.com>, <mailman.977623149.8184.python-list@python.org> <9257mf0sl1@news1.newsguy.com> <roy-E5B925.13324224122000@news.panix.com> <925khi019hs@news1.newsguy.com> <kubsu1t81x.fsf@lasipalatsi.fi> <mailman.977739488.28224.python-list@python.org> <mailman.977785029.26926.python-list@python.org> <20001226154218.E87F4A84F@darjeeling.zadka.site.co.il>
Message-ID: <002c01c06f19$265f01c0$14c02dd5@martelli>

"Moshe Zadka" <moshez at zadka.site.co.il> writes:

> > That would be handy, and I like 'Boost Python' approach to that --
> > naming conventions rather than complex correspondences (see
> > www.boost.org, but i guess C++-loathers would rather ignore it:-).
>
> In spite of your warning, I went to the site, but couldn't figure
> out the relevance. Can you summarize what is the Boost Python approach?

See: http://www.boost.org/libs/python/doc/special.html, right
at the end of this longish page.

Any BPL-wrapped C++ class is a metaclass that lets Python
classes inheriting from it define per-attribute accessors;

"""
BPL checks the special names

__getattr__<name>__
__setattr__<name>__
__delattr__<name>__
to provide functional access to the attribute <name>. This facility can
be used from C++ or entirely from Python. For example, the following
shows how we can implement a ``computed attribute'' in Python:

>>> class Range(AnyBPLExtensionClass):
...    def __init__(self, start, end):
...        self.start = start
...        self.end = end
...    def __getattr__length__(self):
...        return self.end - self.start
...
>>> x = Range(3, 9)
>>> x.length
6
"""

I'd rather have the accessor-name shorter (__get_length__, for
example), but the key thing is simple naming convention.

At the C++-level, only, BPL also provides finer-grained control
of this through functions def_getter and def_setter to get read
and write access respectively for a given attribute name, and the
simpler def_readonly (which provides read access AND blocks any
write-access) and def_read_write (which provides r/w access).

I don't think we need such fine grain from the Python side, at
least not in 80-90% of the cases.

If one DID want it, I think the Pythonic way to provide it would
be through special named attributes, e.g. a dictionary from
attribute-names to (getter, setter) pairs -- or two dictionaries,
one for getters and one for setters.  The latter (and more besides,
admittedly) is basically what Mark Hammond's excellent win32com
does to map between COM properties and Python attributes; in
that case, something of that ilk is surely needed since several
little key pieces of metadata must be associated with each prop,
not just one or two methods.

While Abrahams and K?the's Boost Python Library (BPL) uses
the metaclass hook to provide this, Hammond doesn't -- he does
it via __getattr__/__setattr__ (actually, he generates Python
code from COM type-libraries on the fly, caching it for future
accesses, and that Python code in turn uses __getattr__ &c -- a
neat hack indeed, but that's another issue:-).


Maybe I should clarify the implicit-delegation issue too.
Sticking to getters for brevity, this could be it today:

class GetterDelegatorMixin:
    def __init__(self):
        self._deldic_={}

    def delegateTo(self, delegate, *names):
        for name in names:
            self._deldic_[name] = delegate

    def removeDelegation(self, *names):
        for name in names:
            del self._deldic_[name]

    def removeDelegate(self, delegate):
        for name, whosit in self._deldic_.items():
            if whosit == delegate:
                del self._deldic_[name]

    def __getattr__(self, name):
        delegate = self._deldic_.get(name,None)
        if delegate is None:
            raise AttributeError, name
        return getattr(delegate, name)


As long as the __init__ and __getattr__ of this mixin are
appropriately called, any inheriting instance can choose
to delegate certain named attributes to equally named
ones of other objects (actually, the getattr for their
lookup might usefully be anticipated to delegateTo time,
but doing it at the very last instant might allow for even
_further_ fluidity...:-).  All sorts of design decisions are
somewhat arbitrary (_deldic_ maps name->delegate, so
to find everything mapped to a given delegate, e.g. for
total removal of that delegate's responsibilities, we need
to loop over all of its items -- I think that's OK because,
in the use-cases I have in mind, the map is not all that
big for any given object anyway, but considering other
use-cases might lead to different architectures), but as a
proof-of-concept I hope it may serve.


Alex



_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com




From cookedm at physics.mcmaster.ca  Fri Dec 29 23:33:55 2000
From: cookedm at physics.mcmaster.ca (David M. Cooke)
Date: 29 Dec 2000 23:33:55 -0500
Subject: sniffing btw my web browser and a server
References: <92aush$rhc$1@news.nuri.net> <92fs3l$5gg$01$1@news.t-online.com>
Message-ID: <qnk7l4ittn0.fsf@arbutus.physics.mcmaster.ca>

At some point, "Eduard Hiti" <KatMouse at gmx.de> wrote:

> You can try 'Proxomitron' as a proxy. It's basically an ad filtering
> (shonenware!?) software (but capable of much more), and has a very nice log
> window, in which you can view all HTTP headers sent.
> 
> The URL: http://members.tripod.com/Proxomitron/
> 
> 
> There's also a Python only solution: 'WebDebug'. It also acts as a proxy,
> logging all transfers and generating a nice and comprehensive HTML report.
> 
> The URL: http://www.cyberclip.com/webdebug/download.html
> 

Also check out this pure Python proxy code by Amit J. Patel at

http://theory.stanford.edu/~amitp/proxy.html

I use Proxy 3: it has many options (like changing Slashdot from green
to blue, show HTTP headers, etc.), and is very modular. Well worth a look.

-- 
|>|\/|<
/--------------------------------------------------------------------------\
|David M. Cooke
|cookedm at mcmaster.ca


From claird at starbase.neosoft.com  Fri Dec 29 14:04:54 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 29 Dec 2000 13:04:54 -0600
Subject: Python vs. PHP (& Java?)
References: <69BC4C6AE83ED311BC9400805FA7B14AF84D2C@il93exp01.css.mot.com> <92fv4a$l2l$1@nnrp1.deja.com> <92gghv02iq1@news2.newsguy.com> <92iie9$lg3$1@nnrp1.deja.com>
Message-ID: <80D1D7A4D98F27A8.5722E62E65394869.6E689B85B4FC3BB1@lp.airnews.net>

In article <92iie9$lg3$1 at nnrp1.deja.com>,  <rturpin at my-deja.com> wrote:
>  "Alex Martelli" <aleaxit at yahoo.com> wrote:
			.
			.
			.
>> Nothing "is EXACTLY like" anything else (Korzibsky would
>> insist on such firm avoidance of "the IS of identity", but,
>> as he doesn't seem to be following this newsgroup regularly,
>> I'll have to play his role, I guess). ..
>
>I'm surprised no one brought up the good Count when Clinton
>uttered his infamous question.
I assume, Russell, that you are just being provocative
here.  OF COURSE some folks brought it up, including
<URL:http://members.aol.com/evolvertcb/intro.htm>
(while you're roaming through the Web, check out the
*TIME-BINDING Newsletter*).
			.
			.
			.
>> [Perl's] design goals DID apparently include rich,
>> exhuberant, redundant abundance -- it seems to have
>> matched them closely. ..
>
>That is a most unfortunate goal.
I think you harsh in that.  It seems to suit some
mentalities, and the worst I can bring myself to say
is that it's been an informative (and not-yet-finished!)
experiment.
			.
			.
			.
-- 

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


From stevee at recombinant.demon.co.uk  Sat Dec 23 16:30:22 2000
From: stevee at recombinant.demon.co.uk (Stephen D Evans)
Date: Sat, 23 Dec 2000 21:30:22 +0000
Subject: coding convention conversion
References: <20001222122015.A7987@llnl.gov> <mailman.977522226.5104.python-list@python.org> <oSV06.176$g03.67749@sapphire.mtt.net> <mailman.977550847.24965.python-list@python.org>
Message-ID: <3A45196E.26B5BA6A@recombinant.demon.co.uk>

Issac

There is a utility by the name of fixcid.py with the python distribution in the
Tools\scripts directory that does 'massive identifier substitution on C source
files'. It looks like it will work on C++ too. You'd probably have to get the
(unmangled) function names from the linker.

Stephen D Evans


Issac Trotts wrote:

> OK, that should do the trick for python.  Any ideas about how to
> do it for C++ (using python, so that this isn't off-topic ;).
>
> Issac
>
> On Sat, Dec 23, 2000 at 04:29:08AM +0000, Bruce Dodson wrote:
> > Better yet, import the file as a module, and scan through its public symbols
> > (via dir()) to find out which names need to be changed, and then use that
> > information to build a sed script which can be applied globally.  You can
> > easily take advantage of other meta-information such as the type, e.g. limit
> > it to just functions if you want.  This approach is good because the parser
> > will recognize more variations than a regexp.
> >
> >
> > Something like this is what I had in mind:
> >
> > from types import ModuleType, ClassType
> >
> > def print_symbols(symbols, prefix=''):
> >     for sym in dir(symbols):
> >         if sym[:2] == '__': continue
> >
> >         v = getattr(symbols, sym)
> >         if type(v) is ModuleType: continue
> >
> >         print prefix + sym
> >
> >         if type(v) is ClassType:
> >             print_symbols(v, prefix + sym + '.')
> >
> >
> > See also the pyclbr module for a way to do something similar without
> > executing the module.
> >
> > Bruce
> >
> >
> >
> > "D-Man" <dsh8290 at rit.edu> wrote in message
> > news:mailman.977522226.5104.python-list at python.org...
> > >
> > > This shouldn't be too hard with a python script and some regexes with
> > > some more string processing in between (or maybe multiple pass).
> > >
> > > Everything you want to change is bounded by "def" on one side and "("
> > > on the other.  Then replace the capitals with _ followed by the lower
> > > case version.
> > >
> > > A quick (partial) solution:
> > >
> > > re = re.compile( "def ([a-z]+)([A-Z])([a-z]*)\(" )
> > >
> > > for line in source.readlines() :
> > > match = re.match( line )
> > > if ( match ) :
> > > res = "def " + re.group( 0 ) + "_" + \
> > > string.lower( re.group( 1 ) + re.group( 2 ) + "("
> > >
> > >
> > > This isn't complete and won't have the arguments in the result (nor
> > > preserve indentation before the "def").
> > >
> > > I think the re.sub() function will do more of what you want it to.
> > >
> > >
> > > HTH,
> > > -D
> > >
> > >
> > > On Fri, Dec 22, 2000 at 12:20:15PM -0800, Issac Trotts wrote:
> > > > Does anyone know of a utility to change between the coding conventions
> > > > of
> > > >
> > > > def fooBarBaz(): pass
> > > >
> > > > and
> > > >
> > > > def foo_bar_baz(): pass
> > > >
> > > > It is a pain to do this by hand.
> > > >
> > > > Thanks,
> > > > Issac
> > >
> >
> > --
> > http://www.python.org/mailman/listinfo/python-list



From pbjorn at uswest.net  Fri Dec 22 14:39:24 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Fri, 22 Dec 2000 12:39:24 -0700
Subject: ANNOUNCE: soundex.py
References: <mailman.977338109.32467.python-list@python.org> <3A4119C3.2E5E54E1@uswest.net> <mailman.977501169.22862.python-list@python.org> <slrn94787k.1q7.amk@207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com> <3A43A006.F2CCC36@uswest.net> <slrn94793q.1qk.amk@207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com>
Message-ID: <3A43ADEC.66C0487E@uswest.net>

"A.M. Kuchling" wrote:

> On Fri, 22 Dec 2000 11:40:06 -0700, Bjorn Pettersen <pbjorn at uswest.net> wrote:
> >If we could get the soundex.py as part of the standard library, this
> >would definitely get my vote.
>
> That's not what I'm suggesting; it was there once and was removed as
> being of marginal usefulness to most people. Modules/soundex.c is
> still around as a relic of this, though it's not compiled by default.
> I'm suggesting it's now safe to delete soundex.c.

Well, since the .py module now has more functionality than the .c module I can
definitely see problems with keeping soundex.c. OTOH, removing it would force
people to download something extra instead of just editing the Setup file...
Speaking from a software maintenance perspective, I wouldn't expect a deprecated
module to disappear until the next _major_ release.

Perhaps it's time to create an "extras" package that contain all the marginal
modules that don't have enough of a userbase to be in the standard library?

-- bjorn



From israelt at optushome.com.au  Wed Dec 13 05:24:44 2000
From: israelt at optushome.com.au (israel raj thomas)
Date: Wed, 13 Dec 2000 10:24:44 GMT
Subject: Smalltalk and Python 
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de>
Message-ID: <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com>

>Do you think that it would be possible to use Python as a Smalltalk
>replacement if an IDE existed that hid some of Python's limitations?

Would it perhaps not be better to use Ruby ?
See below for information and for comparisons of Ruby, Python, Eiffel
and Smalltalk:

http://www.bagley.org/~doug/shootout/
http://www.pragmaticprogrammer.com/ruby/articles/index.html
http://www.pragmaticprogrammer.com/talks/introruby/introruby.htm
http://www.ddj.com/articles/2001/0101/0101b/0101b.htm
http://www.s-direktnet.de/homepages/neumann/rb/download_ruby.html
http://www.s-direktnet.de/homepages/neumann/ruby_en.html
http://www.s-direktnet.de/homepages/neumann/lang_cmp.en.htm
http://www-4.ibm.com/software/developer/library/ruby.html
http://www.sunworld.com/swol-02-1999/swol-02-regex.html



From pearu at cens.ioc.ee  Fri Dec 22 02:37:31 2000
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Fri, 22 Dec 2000 09:37:31 +0200
Subject: ?Why is 0e-6 invalid?
In-Reply-To: <mailman.977460545.25728.python-list@python.org>
References: <3A428C7C.5CFBB181@email.sps.mot.com> <mailman.977460545.25728.python-list@python.org>
Message-ID: <Pine.LNX.4.10.10012220922540.14068-100000@kev.ioc.ee>


On Thu, 21 Dec 2000, D-Man wrote:

> AFIAK scientific notation is [1-9]e-?[0-9]+
> (a 1 thru 9 followed by e followed by optional negative sign followed
> by any sequence of 0 thru 9)

Scientific notation in Python is a bit more involved:

[+-]?((\d+[.]\d*|[.]\d+|[1-9]+)[eE][+-]?\d+ # may not be a complete pattern

that allows to represent number `0' also.

Though, I don't understand why 0e-6 is invalid in Python, indeed.

Pearu



From syver at NOSPAMcyberwatcher.com  Fri Dec  8 21:41:36 2000
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Sat, 9 Dec 2000 03:41:36 +0100
Subject: Unit Testing tool
Message-ID: <90s69s$f85$1@troll.powertech.no>

What is the Python unit testing tool preferred by readers/writers from
comp.lang.python?




From jadestar at idiom.com  Mon Dec 11 06:40:13 2000
From: jadestar at idiom.com (James T. Dennis)
Date: 11 Dec 2000 11:40:13 GMT
Subject: Script vs Program
References: <y39V5.524$OC.159868@typhoon.aracnet.com> <bHnV5.96503$U46.3094705@news1.sttls1.wa.home.com> <86F97286C31D847A.BFF08774F46BB2D0.C6D33E45597C67FC@lp.airnews.net>
Message-ID: <912eet$1g72$1@news.idiom.com>

Cameron Laird <claird at starbase.neosoft.com> wrote:

> In article <bHnV5.96503$U46.3094705 at news1.sttls1.wa.home.com>,
> Greg Jorgensen <gregj at pobox.com> wrote:
> 			.
>>Python is often called a scripting language, but I've never heard Java
>>called a scripting language. Perhaps Python seems more script-like because
>>you can start up the interpreter and get it to interpret and execute code
>>directly, whereas Java tools don't have immediate execution modes (at least
>>not that I've seen).
> 			.
> I have--heard Java called "a scripting language", that
> is.  'Happens all the time, particularly among those
> with shaky educations.


	I think the difference between "scripts" and 
	"programs" relates to the human processes that 
	are involved in their development.

	A script is basically written directly, possibly
	going through several iterative cycles of refinement.
	Scripts generally have a reasonably simple architecture
	(even if they are quite long and complex in actuality).

	A program is *designed* and code is written to
	implement that design.  

	I mostly write scripts.  Generally this is no real
	"design" involved.  Certainly there is some 
	analysis of my needs (desired outputs and/or 
	effects) and my inputs and resources (expected
	inputs, available tools or data, etc).  However, 
	the necessary filtering, transformation and 
	manipulations of my inputs are generally pretty
	straightforward enough of obviate any formal
	design process. 

	Naturally I do my scripting using languages that
	are particularly well suited to this process
	(interpreted, high-level languages that offer
	flexible manipulations of strings, lines of input,
	whole files, etc).  Obviously there are many 
	scripting languages that are particularly apropos
	for tasks in specific domains (awk and text
	processing, expect for automation of interactive
	processes, TCL/Tk for GUI interfaces, procmail for
	e-mail filtering, sorting and auto-response generation,
	etc).

	I don't think of Python as a "scripting" language
	though it certainly has "scripting" features.




From dbrueck at edgix.com  Tue Dec  5 09:16:49 2000
From: dbrueck at edgix.com (Dave Brueck)
Date: Tue, 5 Dec 2000 07:16:49 -0700
Subject: Binary vs. Text mode
In-Reply-To: <20001202001141.B13916@corrada.com>
Message-ID: <009201c05ec6$024f4960$760214ac@PRODUT2KDAVE>

Absolutely. I hope I didn't make it sound like this was Python's doing;
thanks for making that clear.

-Dave

> As I have read in previous postings, this is not a feature of
> Python but rather a feature of C. This 'feature' regularly bites people
> working on Macs and PCs. On Unix, C's behavior is transparent but on Macs
> '\n' is converted to '\r' while on Win32 '\n' is converted to '\r\n'.
>




From sholden at holdenweb.com  Tue Dec 19 08:07:17 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 19 Dec 2000 08:07:17 -0500
Subject: tar files
References: <3A3B89BF.51B2C26C@ex.ac.uk> <yaybsuc8ccx.fsf@einn.ifi.uio.no> <mailman.977177822.8335.python-list@python.org>
Message-ID: <S5J%5.18020$ma.344495@e420r-atl2.usenetserver.com>

Rodrigo Senra <rodsenra at correionet.com.br> wrote in message
news:mailman.977177822.8335.python-list at python.org...
> Vetle Roeim wrote:
> >
> > * M. J. Fisher
> > > Hi,
> > >
> > > I am trying to find some code for unpacking tarred files and I am not
> > > sure why to look, does anybody have any ideas? I would also appreciate
> > > an example if any one had one,
> >
> > well.. I started making a tar-module in pure Python, but I never had
> > time to finish it. you can use it if you want, and is able to
> > understand it.. it was written wuick and dirty by examining the
> > documentation and source code of GNU Tar.
>
> Couldn't the 'tar functionality' be achieved by using
> the zipfile module without setting ZIP_DEFLATED ?
>
bzzzzt.

A simple experiment shows this is not true:

>>> import zipfile
>>> z = zipfile.ZipFile(r"C:\Temp\metakit-2_01.tar")
Traceback (innermost last):
  File "<interactive input>", line 1, in ?
  File "d:\python20\lib\zipfile.py", line 113, in __init__
    self._GetContents()
  File "d:\python20\lib\zipfile.py", line 137, in _GetContents
    raise BadZipfile, "File is not a zip file, or ends with a comment"
BadZipfile: File is not a zip file, or ends with a comment
>>>

The default compression is ZIP_STORED.  While WinZip, having been programmed
to know about tar files, can handle this file, the ZipFile module clearly
can't.

regards
 Steve





From thomas at xs4all.net  Sat Dec 23 12:17:41 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Sat, 23 Dec 2000 18:17:41 +0100
Subject: A Suggestion for Python Colon Syntax
In-Reply-To: <slrn947q64.jkg.zeitlin@seth.lpthe.jussieu.fr>; from zeitlin@seth.lpthe.jussieu.fr on Fri, Dec 22, 2000 at 11:50:28PM +0000
References: <mailman.977446446.16022.python-list@python.org> <slrn947q64.jkg.zeitlin@seth.lpthe.jussieu.fr>
Message-ID: <20001223181741.F5515@xs4all.nl>

On Fri, Dec 22, 2000 at 11:50:28PM +0000, Vadim Zeitlin wrote:

>  BTW, what was the real rationale for adding composed assignment operators
> (+=) to Python 2.0 (which I'm really extremely glad to have now)? Wasn't
> it meant to make Python more attractive to the people coming from C-lie
> languages?

No, honestly not. If that was the case, '++' and '--' would have been added
too. The funny part is that I never saw a concious decision not to add them
being made, everyone involved (Guido, Michael Hudson, me, the rest of
python-dev) seemed to naturally accept that the auto-inc/dec operators had
no place in Python, and that in itself is plenty of validation, don't you
think ? :-)

Guido-probably-made-all-decisions-about-Python-years-ago--he's-just-waiting-
 -for-people-to-_implement_-them-ly y'rs,
-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From amk at mira.erols.com  Fri Dec 22 13:45:43 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 22 Dec 2000 18:45:43 GMT
Subject: ANNOUNCE: soundex.py
References: <mailman.977338109.32467.python-list@python.org> <3A4119C3.2E5E54E1@uswest.net> <mailman.977501169.22862.python-list@python.org> <slrn94787k.1q7.amk@207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com> <3A43A006.F2CCC36@uswest.net>
Message-ID: <slrn94793q.1qk.amk@207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com>

On Fri, 22 Dec 2000 11:40:06 -0700, Bjorn Pettersen <pbjorn at uswest.net> wrote:
>If we could get the soundex.py as part of the standard library, this
>would definitely get my vote.

That's not what I'm suggesting; it was there once and was removed as
being of marginal usefulness to most people. Modules/soundex.c is
still around as a relic of this, though it's not compiled by default.
I'm suggesting it's now safe to delete soundex.c.

--amk









From katz at Glue.umd.edu  Sat Dec  9 19:12:01 2000
From: katz at Glue.umd.edu (Roy Katz)
Date: Sat, 9 Dec 2000 19:12:01 -0500
Subject: Dijkstra's Shortest Path algorithm
In-Reply-To: <90t78p$5r1$03$1@news.t-online.com>
References: <Pine.GSO.4.21.0012090042370.15110-100000@y.glue.umd.edu> <90t78p$5r1$03$1@news.t-online.com>
Message-ID: <Pine.GSO.4.21.0012091850170.28179-100000@z.glue.umd.edu>

On Sat, 9 Dec 2000, J?rgen Hermann wrote:

> Maybe you should try to implement it in Python and NOT take 2-3 days. ;))


I've looked around on google and found a visual graph solver applet coded
(presumably) in Python; I have to check it out.  I've also checked out
www.python.org/docs/essay/graphs.html;  it lists a beginnings of a Graph 
class (sans Dijkstra's algorithm).  

consider something simple:



 Class Graph:


   adjMatrix = None  # array.array? all we need are int's. (NxN matrix)
   cost      = []    # cost from nodeFrom to all other nodes
   pred      = []    # predecessors of every node
   visited   = []    # boolean array symbolizing who has been processed

   # add a vertex
   def addVertex( node ): pass

   # add a (weighted) edge
   def addEdge( nodeFrom, nodeTo, cost ): pass

   # returns array of nodes representing a path;
   # empty path denotes nodeFrom == nodeTo; 
   # None represents no connection between 
   # nodeFrom and nodeTo.

   def shortestPath( nodeFrom, nodeTo ): pass

   # computes cost from nodeFrom to every other node and
   # stores it in self.cost; also builds 'self.pred' array.
   # uses a heap (for the more moderately-sized graphs),
   # otherwise a simple array.

   def dijkstraHelper( nodeFrom ): pass



coded as a C++ extension (I say C++ because I used C++ for my
assignment).  Conceptually, the class would involve an adjacency matrix to
store the edge weights, and some small list (or dictionary) to act as a
lookup table.  shortestPath, after calling dijkstraHelper(),  would use
The lookup table to map vertex integers to objects as it builds the path. 

I can imagine many situations in which I'd want to use something like 
this.  Unfortunately, as finals approach, I don't think I have the time to
code this up :).  



Roey


2nd year, Computer Engineering 
University of Maryland, USA




From eblakes at drwtrading.com  Fri Dec  8 22:21:18 2000
From: eblakes at drwtrading.com (Edward Blakes)
Date: Fri, 8 Dec 2000 22:21:18 -0500
Subject: Implementing IPersistStorage using PythonCOM
References: <t308m58pkecra3@corp.supernews.com> <3A307D8C.10909@ActiveState.com> <t325urdmtq1t07@corp.supernews.com> <3A317651.5000109@ActiveState.com>
Message-ID: <t339eb6pvurf67@corp.supernews.com>

On its way!

Ed


"Mark Hammond" <MarkH at ActiveState.com> wrote in message
news:3A317651.5000109 at ActiveState.com...
> Edward Blakes wrote:
>
>
> > My understanding was that this simply converts a string to a PyIID, not
> > generates a brand new GUID. As you will see below, I directly placed the
> > string per your suggestion, but to no avail.
>
> My mistake - sorry.  Your code was fine.
>
> > I made the changes suggested, got more output but with the same result.
> > Given your response I believe I clipped important information that, in
its
> > absence, mislead you. So that I won't do that again, I am placing the
entire
> > source here so that you can run it directly. I hate the length but I
want
> > you to have all the information needed.
>
> Can you mail it to me - all the lines are wrapped!
>
> Mark.
>




From hinsen at cnrs-orleans.fr  Tue Dec 19 08:22:47 2000
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 19 Dec 2000 14:22:47 +0100
Subject: Is Stackless and uthread.py SMP friendly?
References: <91br9v$qa4$1@nnrp1.deja.com> <ku1yva5ocq.fsf@lasipalatsi.fi> <91dp4j$5fj$1@news.gte.com> <91dt39$fbk$1@nnrp1.deja.com>
Message-ID: <m38zpcwno8.fsf@chinon.cnrs-orleans.fr>

jbalding at hotmail.com writes:

> Okay. My next question: Is there a simple way of taking advantage of
> multiple processors using Python, or is this something that is going to

Unless your code spends most of its time in purpose-written C code,
threads are not the way to go, so you need multiple independent Python
processes. There are a couple of packages out there that can help with
coordinating such processes. For example, you can use an MPI
implementation (there are at least two free ones, MPICH and LAM) with
a Python interface (e.g. the one in ScientificPython) and add some
high-level code for task administration etc.

This is not as complicated as it sounds; the most difficult part is
installing everything. And MPI makes sense on shared-memory machines
as well; a good MPI implementation handles message exchange via shared
memory when possible. However, for some problems message-passing is
more difficult to use than shared memory communication.
-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.56.24
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------


From claird at starbase.neosoft.com  Sun Dec 31 11:49:10 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 31 Dec 2000 10:49:10 -0600
Subject: Python vs. PHP (& Java?)
References: <mailman.977897470.4734.python-list@python.org> <92m5ar$cfm$2@newshost.accu.uu.nl>
Message-ID: <1BF13176BDBA46CA.F610B8E8417D60A4.AA7406CC253BD8A7@lp.airnews.net>

In article <92m5ar$cfm$2 at newshost.accu.uu.nl>,
Martijn Faassen <m.faassen at vet.uu.nl> wrote:
>Ben Ocean <beno at thewebsons.com> wrote:
>
>[snip]
>> Python has the advantage of byte-code and
>> compilation, but it?s server-side. PHP might not be as fast on the server,
>> but it doesn?t have to make as many trips back and forth between server and
>> client.
>
>It appears as if you're misunderstanding something here. Firstly, 
>I don't know which of the languages is faster; it probably depends on the
>task as much on the language itself. I don't know much about PHP, but I thought
Worth repeating.
>it could be compiled to bytecode or at least some intermediate format as well?
I quote <URL: http://www.zend.com/zend/whats-new.php >:
  PHP 4.0 is more efficient, more reliable,
  and much, much faster than PHP 3.0.  . . .

  This superior performance of PHP 4.0 is a
  result of the PHP scripting engine's
  redesign: ...  The PHP 4.0 scripting engine
  ... uses a much more efficient "compile-then-
  execute" paradigm, instead of the
  "execute-while parsing" model employed in PHP
  3.0.
Plenty (sufficient motivation could probably quantify
that usefully) of sites remain at PHP3; that probably
yields some of the confusion.
-- 

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


From junaftnoon at nospamplzyahoo.com  Thu Dec 28 05:25:14 2000
From: junaftnoon at nospamplzyahoo.com (June Kim)
Date: Thu, 28 Dec 2000 19:25:14 +0900
Subject: Interesting behaviour of the assignment
Message-ID: <92f4gk$qk6$2@news.nuri.net>

>>> a=100
>>> b=100
>>> a is b
0
>>> a=99
>>> b=99
>>> a is b
1

Is this an intended/expected result? What is the lying structure beneath
this?




From kragen at dnaco.net  Sat Dec 30 17:48:08 2000
From: kragen at dnaco.net (Kragen Sitaker)
Date: Sat, 30 Dec 2000 22:48:08 GMT
Subject: turtle.py bugfix & questions
Message-ID: <ICt36.11$vd1.1285@e420r-atl2.usenetserver.com>

turtle.py, which I believe is in the standard distribution (at least of
PythonWin) has a minor bug: when tracing is turned on, lines shorter
than a pixel don't get drawn at all.  If you're building long curves
made of such lines, this is a bad thing.

Here's a unidiff patch I think fixes the problem.  

--- turtle.py	Sat Dec 30 14:24:17 2000
+++ nturtle.py	Sat Dec 30 14:26:09 2000
@@ -221,6 +221,8 @@
                         self._canvas.coords(item, x0, y0, x, y)
                         self._canvas.update()
                         self._canvas.after(10)
+                    # in case nhops==0
+                    self._canvas.coords(item, x0, y0, x1, y1)
                     self._canvas.itemconfigure(item, arrow="none")
                 except Tk.TclError:
                     # Probably the window was closed!

The range controlling the loop above is range(1, 1+nhops) instead of
range(1, nhops); presumably, if it were range(1, nhops), then the last
pixel's worth of drawing would be left off the line.  This should now
be unnecessary, as the coords() call after the loop should take care of
that.

Larger questions I have: is anyone using turtle.py?
http://www.python.org/1.5/NEWS-152.txt says:

	- turtle.py is a new module for simple turtle graphics.  I'm still
	working on it; let me know if you use this to teach Python to children 
	or other novices without prior programming experience.

But it's not mentioned anywhere else I can find --- not even under
"Undocumented modules" in the library reference.

Perhaps, if nobody is using it, there is no documentation, and it's
buggy (although it looks pretty stable in the hour or so I played with
it, I did find this one bug), we should remove it from the standard
distribution.

(I found it while looking for examples of the usage of Tkinter.Canvas,
because there is no documentation for Tkinter in PythonWin --- not even
the Tcl man pages, as far as I can tell.)
-- 
<kragen at pobox.com>       Kragen Sitaker     <http://www.pobox.com/~kragen/>
Perilous to all of us are the devices of an art deeper than we possess
ourselves.
       -- Gandalf the White [J.R.R. Tolkien, "The Two Towers", Bk 3, Ch. XI]




From neelk at alum.mit.edu  Thu Dec 21 21:40:31 2000
From: neelk at alum.mit.edu (Neelakantan Krishnaswami)
Date: 22 Dec 2000 02:40:31 GMT
Subject: reading numbers from file
References: <slrn9457uq.9tl.jp@localhost.localdomain>
Message-ID: <slrn945jg1.p7a.neelk@alum.mit.edu>

On 22 Dec 2000 00:27:06 GMT, Jacek Pop?awski <jp at ulgo.koti.com.pl> wrote:
>I have file with numbers:
>
>23 41 24 0 2 42 532 53 234
>34 234 53 23 4 323 53 (etc...)
>
>1.How to read one line of this file and write it to list of numbers?
>  (like: a=[23,41,24...])

This code will put a big list of numbers into the array allnums

import string

def string_to_numbers(line):
    return map(int, string.split(line))

allnums = []
file = open("foo.txt")
for line in file.readlines():
    numbers_on_line = string_to_numbers(line)
    allnums.extend(numbers_on_line)

The function string_to_numbers can also be written in two other
ways:

def string_to_numbers(line):
    return [int(x) for x in string.split(line)]

def string_to_numbers(line):
    lst = []
    for x in string.split(line):
        lst.append(int(x))
    return lst

Any of these is fine -- use whichever is most clear to you.

> 2.How to read only 5 numbers, write it to list, then another 5, and
> write it to second list?

lst1 = []
lst2 = []
file = open("foo.txt")

current = lst1
count = 1
for line in file.readlines():
    numbers_on_line = string_to_numbers(line)
    #
    # Each number should be written to either lst1 or lst2. We'll
    # track the number of numbers written to both lists and flip
    # to the other one each time we hit a multiple of 5.
    #
    for n in numbers_on_line:
    	current.append(n)
    	count = count + 1
    	if (count % 5) == 0:
    	    if current == lst1:
    		current = lst2
    	    else:
    		current = lst1

> 3.What will change if I want to add comments (#) and empty lines to
> my file?

You'll need to change the function string_to_numbers() so that it can
account for these two changes to your format. string.split() already
returns an empty list if it receives a line of whitespace, so it
already works on blank lines. To handle '#' to the end of line
comments, we can either use regexps or string.split() again. I'll do
both.

The basic idea is that if we do string.split on '#' then we get a list
of elements that are divided by the '#'. We only need to look at the
first element of that list since it is the string that precedes the
first '#' on the line.

def string_to_numbers(line):
    comment_free_line = string.split(line, "#")[0]
    return map(int, string.split(comment_free_line))

# And then the prior solutions follow.

Using regexps, the solution might look like this:

import re

comment_free = re.compile("^([^#]*).*$")

def string_to_numbers(line):
    m = comment_free.match(line)
    comment_free_line = m.group(1)
    return map(int, string.split(comment_free_line))

I hope this helped.


Neel


From fredrik at effbot.org  Thu Dec 21 12:26:14 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Thu, 21 Dec 2000 17:26:14 GMT
Subject: A Suggestion for Python Colon Syntax
References: <91tc8h$2pp$1@hecate.umd.edu>
Message-ID: <W2r06.1076$AH6.86249@newsc.telia.net>

"William Djaja Tjokroaminata" wrote:
> However, a lot of time when I ran my Python program, I got syntax
> error, because I forgot to put colon (:) after my if (or for, etc.) clause.
> My question is, is the colon really necessary, syntactically, in Python?

Short answer:  The colon isn't there for the computer, it's
there for human readers.  Just like the colons in your post:

> My suggestion is as follows:
>
> 1) If the if (or for, etc.) clause is followed by a block, then the colon
> is optional (for backward compatibility):
>
>    if condition is true
>        statement_1
>        statement_2
>        ....

(Using colons and indentations in the same way as in other written
languages is just one of many ideas Python has borrowed from the
ABC language: http://www.cwi.nl/~steven/abc )

:::

Longer answer: ... nah, I'll leave that to the "while true" crowd.
I'm sure they'll keep you entertained until the end of next year!

:::

Happy holidays, everyone!

</F>




From jp at ulgo.koti.com.pl  Thu Dec 28 03:26:43 2000
From: jp at ulgo.koti.com.pl (Jacek =?iso-8859-2?Q?Pop=B3awski?=)
Date: 28 Dec 2000 08:26:43 GMT
Subject: split hangs
References: <slrn94l1pu.4so.jp@localhost.localdomain> <92e3lm$pni$1@panix3.panix.com> <slrn94l4d3.550.jp@localhost.localdomain> <yv2n1dhufo2.fsf@lionsp093.lion-ag.de>
Message-ID: <slrn94lua3.67q.jp@localhost.localdomain>

Harald Kirsch wrote:
>Yes, something is very strange with this type of regex. Run the
>following program. It shows strange behaviour with 1.5.2 on Linux as
>well as with 2.0 on IRIX.

wow, looks like every step is few times slower than latest, so it
doesn't hang, it just compute and compute... but why regex are _so_
slow ?

-- 
Changes comes slowly 
But this is what you'll be
Somehow it snows in the summer  
And the sun is freezing me                     "Otherworld" - Ronnie James Dio


From sh at ttsoftware.co.uk  Fri Dec 15 09:37:43 2000
From: sh at ttsoftware.co.uk (Steve Horne)
Date: Fri, 15 Dec 2000 14:37:43 +0000
Subject: Python advocacy
References: <91bln3$lmr$1@nnrp1.deja.com> <91d63k0msl@news1.newsguy.com>
Message-ID: <8iak3t4a3lrqcqp9iqftmekdoahuico2b0@4ax.com>

On Fri, 15 Dec 2000 14:24:27 +0100, "Alex Martelli"
<aleaxit at yahoo.com> wrote:

><jschmitt at vmlabs.com> wrote in message news:91bln3$lmr$1 at nnrp1.deja.com...
>> I just read this good article about Perl advocacy.  I think it applies
>> to Python equally.  I found the link mentioned from /. this morning.
>>
>> http://www.perl.com/pub/2000/12/advocacy.html?wwwrrr_20001213.txt#why i
>> hate advocacy
>
>Very interesting (for some reason, a not-yet-Pythonista Perl-using
>coworker forwarded me an e-mail copy too -- wonder if he's trying
>to tell me something?-).  Highly recommended reading for anybody
>interested in really understanding why human beings are this way,
>Matt Ridley, "The Origins of Virtue".  There's a long detailed
>summary at http://www.geocities.com/Athens/Ithaca/4388/origins.html,
>but you wouldn't be doing yourself a favor if you limited your
>reading to those 23 pages -- Ridley writes very well, and argues
>persuasively; get the original!

I've read it and, if it's the one I think it is (lots of stuff about
the prisoners dilemma, for instance) then I agree. I'd also recommend
the books by Steven Pinker (one on the origins of language, the other
more generally on human nature - I forget the titles).

<SNIP>

>Makes me wonder why Larry Wall didn't think to include at least
>ONE strangeness, wart, or inconsistency *somewhere* in Perl...
>thus making it so boringly perfect and un-idiosyncratic as to
>leave its advocates without a suitable 'scapegoat feature'!

Erm - either you haven't used Perl, or your an advocate yourself!
Personally, if I was selling Python, Perl would be involved in my
scapegoat manouvre.

In fact, one such wart was mentioned in that article above (a length
function that gives the 'wrong' result due to a scalar context).

-- 
Steve Horne
Home : steve at lurking.demon.co.uk
Work : sh at ttsoftware.co.uk


From obou at my-deja.com  Wed Dec  6 05:05:22 2000
From: obou at my-deja.com (obou at my-deja.com)
Date: Wed, 06 Dec 2000 10:05:22 GMT
Subject: SAP R/3 automation with win32com
References: <90flpu$gpu$1@nnrp1.deja.com> <kl5n2tsfk3hud1c37gjo56290sueccuegk@4ax.com> <90i6n6$hg0$1@nnrp1.deja.com> <90icc6022q9@news2.newsguy.com> <90j81i$c4a$1@nnrp1.deja.com> <90jjgi010ib@news1.newsguy.com>
Message-ID: <90l310$sga$1@nnrp1.deja.com>

In article <90jjgi010ib at news1.newsguy.com>,
  "Alex Martelli" <aleaxit at yahoo.com> wrote:
> <obou at my-deja.com> wrote in message
news:90j81i$c4a$1 at nnrp1.deja.com...
>     [snip]
> In the type-library's "helpstring" propriety.  But that is the same
> place from which Ole View, AND Visual Basic's "Project/References..."
> dialog, also get it, so, if you can find them with any one of these
> tools, the others should be no problem.
>
> Alternatively, the EnsureDispatch method of gencache can (probably)
> work on an object you already have instantiated (if that object is
> reasonably well-behaved) and perform the same work as makepy
> would.  Starting with a clean-slate gen_py, you can just then open
> with your favourite text-editor the only long-named .py file you'll
> find there.
>
> Or, you can do it by studying the registry.  From a ProgId (such as
> "feefie.foofum", in the registry under HKEY_CLASSES_ROOT you'll
> find the CLSID; under HKEY_CLASSES_ROOT/CLSID/thatclsid, you'll
> find the type-library's own GUID, which you can in turn use as a
> key under HKEY_CLASSES_ROOT/Libraries to find out everything
> about the location of the specific type-library of interest.  OleView
> does that for you, by the way, if you start your browsing with an
> object rather than going directly for the libraries.
>
> Alex
>
>

Thanks Alex for this great informations!

I finally used the gencache.EnsureDispatch method and finally got the
member and paramter list.

Olivier


Sent via Deja.com http://www.deja.com/
Before you buy.


From smigowsky at dspace.de  Fri Dec 22 01:42:24 2000
From: smigowsky at dspace.de (Stefan Migowsky)
Date: Fri, 22 Dec 2000 07:42:24 +0100
Subject: AW: COM / Variants / VT_BOOL
Message-ID: <84257D042AA7D411B3F700D0B7DB9B7C13A8BD@newsintern.dspace.de>

You do not need to create a VARIANT ! The win32com library automatically
converts almost all Python objects to the corresponding VARIANT type. 
Since VT_BOOL is just an int which is transmitted through the COM channel 
you can use the simple Python 1 or 0, e.g. 
server = win23com.client.Dynamic("XXX")

server.Enabled = 1 # Set to true
server.Enabled = 0 # Set to false


  Stefan

-----Urspr?ngliche Nachricht-----
Von: frankmcgeough at my-deja.com [mailto:frankmcgeough at my-deja.com]
Gesendet: Freitag, 22. Dezember 2000 03:44
An: python-list at python.org
Betreff: COM / Variants / VT_BOOL


Simple questions. How do I create a variant in Python? How do I set
that variant to hold a VT_BOOL?
More complicated question. Why is this info so hard to find? I would
think it would come up a lot if folks are trying to use the
win32com.client.
Thanks for any clues, Frank


Sent via Deja.com
http://www.deja.com/
-- 
http://www.python.org/mailman/listinfo/python-list



From tdelaney at avaya.com  Tue Dec 12 20:07:09 2000
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 13 Dec 2000 12:07:09 +1100
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B0FBCA6@natasha.auslabs.avaya.com>

This could get complicated to understand, so I'm going to use some examples
...

If a reference is bound as a result of being assigned a reference with
mutable access to an object, the new reference has mutable access.

If a reference is bound as a result of being assigned a reference with
immutable access to an object, the new reference has immutable access.

The immutable() function and immutable keyword can change the access of a
reference being assigned to immutable.

There should also be a mutable keyword for function declarations, but not a
mutable() function. This would throw an exception if a reference is passed
which has immutable access. Basically, this is saying that the contents of
the parameter *will* be changed.

# var1 becomes a reference to a list with mutable access
var1 = []

# var2 becomes a reference to the same list, with immutable access
var2 = immutable(var1)

# var3 becomes a reference to the same list, with immutable access
var3 = var2

# the following operation will throw an exception
try:
	var3.append(1)
except ImmutableException:
# the following line will be reached
	pass

# var3 becomes a reference to the same list, with mutable access
var3 = var1

# the following operation will not throw an exception
try:
	var3.append(1)
except ImmutableException:
# the following line will not be reached
	pass

# var4 may have either mutable or immutable access
def m_or_i (var4):
	return var4

# var5 will always have immutable access
def i_only (immutable var5):
	return var5

# var6 must have mutable access or an exception will be thrown
def m_only (mutable var6):
	return var6

# var3 becomes a reference with mutable access
var3 = m_or_i(var1)

# var3 becomes a reference with immutable access
var3 = i_only(var1)

# var3 becomes a reference with immutable access
var3 = m_or_i(var2)

# var3 becomes a reference with immutable access
var3 = i_only(var2)

# var3 becomes a reference with mutable access
var3 = m_only(var1)

# an exception will be thrown and var3 will be unchanged
var3 = m_only(var2)

Tim Delaney
Avaya Australia
+61 2 9352 9079

> Timothy,
> 
> Thank you very much.  I played around with Python a bit and then
> this started making sense to me. 
> 
> How would immutable work for return values? anything special there?
> 
> Roey
> 
> 
> On Wed, 13 Dec 2000, Delaney, Timothy wrote:
> 
> 
> > Objects are generally documented as mutable or immutable. 
> As an example:
> > 
> > 	tuple is immutable
> > 	list is mutable
> > 
> > In both cases, you can pass a reference to the object to a 
> function. The
> > function aquires a *separate* reference to the object.
> > 
> > Now, if you assign something else *to that reference* you 
> change what that
> > reference is referring to. When you exit the function the 
> original reference
> > still refers to the original object.
> > 
> > However, you can modify the contents of the list that the 
> reference is
> > referring to. In this case, the contents of the object that 
> the original
> > reference is referring to (which is the same object as the 
> reference in the
> > function is pointing to) will change.
> > 
> > If you try to modify the contents of the tuple that a 
> reference is referring
> > to you will have an exception thrown, because it is immutable.
> > 
> > Likewise, you pass around references to other immutable 
> items such as
> > integers. The difference is that they are immutable *and* 
> they don't have
> > "contents" as such. You can change object the reference to 
> the integer is
> > pointing to, but every other reference to that integer 
> remains unchanged.
> > 
> > Note: you list string as "call-by-reference". Yes, it is. 
> However, you can't
> > modify the contents of a string (they are immutable). You 
> can assign a
> > different string to the reference - but the original 
> reference still points
> > to the original string.
> > 
> > > I want to know where this is documented.  Specifically, 
> where is the
> > > dichotomy of ints/longs/floats vs. 
> > > objects/strings/lists/dictionaries. 
> > 
> > There is no dichotomy. Everything behaves the same.
> > 
> > You appear to be thinking of python in C terms. In this 
> case, consider
> > python references as pointers. You *only* pass pointers. 
> Some of these
> > pointers may point to a const memory location. Some don't. 
> If you change
> > what a copy of the pointer is pointing to, you don't change what the
> > original is pointing to. If you change the contents of the 
> memory location
> > is that is being pointed to, dereferencing any pointer 
> which points there
> > will expose the change.
> > 
> > > While I'm brainstorming with Python grammar changes, what 
> > > about a real &
> > > operator.  No more of this hokey is-it-modifiable-in-place-or-not
> > > business.  Explicit control is important. 
> > 
> > IMO there should be an explicit immutable() function and an 
> immutable
> > keyword.
> > 
> > The function would return the original object, but mark it 
> as immutable
> > (thus, you could simulate tuples by calling immutable() on a list).
> > Attempting to modify such an object would throw an exception.
> > 
> > The keyword would be used to mark parameters to functions.
> > 
> > Note that the immutability would be based on *which 
> reference you accessed
> > the contents through*. Thus you could have a multi-threaded 
> app which has
> > one thread with a mutable reference, and all the others 
> with immutable
> > references. Suddenly you don't need to worry about locking 
> an object to
> > write to it!
> > 
> > In both cases, I envisage that it would go from one 
> dictionary holding
> > references in each namespace, to two - one for references 
> to immutable
> > objects, the other holding references to (possibly) mutable 
> objects (i.e.
> > any which have not been marked as immutable).
> > 
> 



From mjackson at wc.eso.mc.xerox.com  Thu Dec  7 14:20:15 2000
From: mjackson at wc.eso.mc.xerox.com (Mark Jackson)
Date: 7 Dec 2000 19:20:15 GMT
Subject: Attention NYC Shoppers & Visitors - Beware of Rip-off SY Stores!!! ...... kHP2pr9I
References: <90onmn$goo$1@news.wrc.xerox.com>
Message-ID: <90ontf$goo$2@news.wrc.xerox.com>

> watch for multiple postings of (early) identical material.

watch for multiple postings of (NEARLY) identical material.

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	I know nothing about this subject but I do have
	prejudices, which I am more than happy to share with you.
					- Leon Botstein




From root at rainerdeyke.com  Sun Dec 17 12:22:35 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Sun, 17 Dec 2000 17:22:35 GMT
Subject: Singleton vs Proxies DP (was Re: Solution: Direct access to Printer I/O lines)
References: <917k9b$9i9$1@nnrp1.deja.com> <slrn93hrjn.2vp.ssthapa@localhost.localdomain> <91cnp3$g59$1@nnrp1.deja.com> <91d23f0jk7@news1.newsguy.com> <91gfnp$7r4$1@animus.fel.iae.nl> <91gkh801e49@news2.newsguy.com> <InW_5.37707$x6.18119119@news2.rdc2.tx.home.com> <91hrbn$v7a$1@slb3.atl.mindspring.net>
Message-ID: <vD6%5.41846$x6.19189815@news2.rdc2.tx.home.com>

"Andrew Dalke" <dalke at acm.org> wrote in message
news:91hrbn$v7a$1 at slb3.atl.mindspring.net...
> Rainer Deyke wrote:
> >My preference for standardized singletons in Python is this:
> >
> >singleton C:
> >  pass # class definition here
> >
> >The singleton statement would be equivalent to the following:
> >
> >class __tmp:
> >  pass # class definition here
> >C = __tmp()
> >del __tmp
>
> Wouldn't the cleaner equivalent be
>
> class C:
>   pass
> C = C()

Good point.

> If that's the case, do we really need a new keyword for
> what can be done with one extra line of code and no new
> functionality?

Well, no, we don't actually *need* a new keyword.  If you have lots of small
singletons, the above idiom is still somewhat cumbersome.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From News at Titanic.co.uk  Wed Dec 27 10:08:54 2000
From: News at Titanic.co.uk (Simon Faulkner)
Date: Wed, 27 Dec 2000 15:08:54 GMT
Subject: Controlling the Parallel Port
References: <mvuj4tkmn22gbth1a6t8fag39902qeeu9l@4ax.com> <mailman.977928790.25688.python-list@python.org>
Message-ID: <pe1k4tsr7u017n45gf09f20fc8k1int2ci@4ax.com>

Cheers Bill,


I have found the link and, indeed, several other c links.

So, pardon my Python ignorance, but what is SWIG?

I am already searching...


Simon

On Wed, 27 Dec 2000 09:52:43 -0500, Bill Scherer
<Bill.Scherer at VerizonWireless.com> wrote:

>Simon
>
>There is a parallel port library for linux at
>
>http://www.circlemud.org/~jelson/software/parapin/
>
>that is simple to SWIG into a Python module.
>I can send you the .i file for this if you want.
>
>I haven't really used it, but it loads and reads the port status just
>fine.
>
>
>-Bill
>
>Simon Faulkner wrote:
>> 
>> Is it easy to control (read/write) the Parallel Port in linux using
>> Python?
>> 
>> I am hoping to use it to drive some simple electronics
>> 
>> (Usual disclaimer understood - If it smokes then you probably got the
>> wiring wrong!)
>> 
>> Simon
>> Simon Faulkner
>> --
>> http://www.python.org/mailman/listinfo/python-list

Simon Faulkner


From sholden at holdenweb.com  Mon Dec  4 08:27:44 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 4 Dec 2000 08:27:44 -0500
Subject: setsockopt and urllib
References: <slrn92moue.hc.frank.neuber@mars.private.de>
Message-ID: <Z%MW5.10936$lF4.165258@e420r-atl1.usenetserver.com>

Look for the timeoutsocket module by Timothy O'Malley <timo at alum.mit.edu> --
sorry I don't have a URL to hand.  This will raise errors on socket timeouts
in library code.

regards
 Steve


Frank Neuber <frank.neuber at gmx.de> wrote in message
news:slrn92moue.hc.frank.neuber at mars.private.de...
> Hi group,
> I have to set special timeouts on receive and send calls in conjunction
with
> urllib-calls.
> Is there a simple solution to do this without modifying the urllib itself.
> For Example:
> I wish to set the timeout to 60 s for the following command:
> (tmpfile, hdr) = urllib.urlretrive(urlname)
> It is possible to to this with the socks module call
> setsockopt(SOL_SOCKET,SO_RCVTIMEO, ttime) over the urllib module.
>
> thanks in advance
> Frank
>
>
> --
>      _/_/_/_/ _//   _/ Frank Neuber
>     _/       _/_/  _/  frank.neuber at gmx.de (private)
>    _/_/_/   _/ _/ _/
>   _/       _/  _/_/    neuber at opensource-systemberatung.de
>  _/       _/    // http://www.opensource-systemberatung.de
>





From aleaxit at yahoo.com  Tue Dec  5 04:36:14 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 5 Dec 2000 10:36:14 +0100
Subject: Python programs (or scripts) as drop targets
References: <hv5k2tg9cdf33j8v333qpkk7jo45ht596n@4ax.com> <90d9hi029hm@news1.newsguy.com> <3A2ACF21.4060406@ActiveState.com> <90fl26022in@news2.newsguy.com> <3A2C73CE.5060506@ActiveState.com>
Message-ID: <90icu9023as@news2.newsguy.com>

"Mark Hammond" <MarkH at ActiveState.com> wrote in message
news:3A2C73CE.5060506 at ActiveState.com...
> Alex Martelli wrote:
>
> > If saving a trivial file is indeed the only advantage of implementing
> > the shell extension, then you may be right (though quibbling about the
> > "registry additions" needed to _register_ a shell extension seems pretty
> > peculiar to me).
>
> Hrm - I hope you didn't think I was quibbling about that!  You won't
> find me arguing against that.  I apologize if that is what I was saying!
>
> While we are at it, the registry information needed to the shell
> communicates to a running Pythonwin via DDE (instead of a new Pythonwin
> startng and _it_ using DDE) would be nice too ;-)

<shudder> if we're talking DDE, there is *another* issue which I
find more pressing: DDEML is *NOT* usable in professional apps
(although many apps that call themselves professional do use it).

Reason: DDEML's functions (those which need to broadcast) use
SendMessage, WITHOUT a timeout!, to broadcast their messages.  So,
DDEML-using applications tend to hang if anybody, anywhere on the
system, has a top-level window whose messages it isn't serving.
And this happens very often (hung processes, etc, etc).  It's a
shame to let ANY other buggy app on the system hang yours...!

DDE itself is a very bad protocol, but the hang problem is quite
specifically with the DDEML helper library -- it's a bug which
MS has acknowledged (5 years ago, I think) yet will never be
fixed (MS apparently does NOT want to encourage DDE use in any
way, shape, or form).  The only fix I can think of is to
rearchitect code that needs to use DDE in order to have it
use the low-level DDE messages (AND SendMessageTimeout any
time it needs to broadcast...!) instead of DDEML.  Better would
be to root out any DDE use whatsoever, of course (relying instead
totally on the COM interfaces to the shell -- best documented
in Dino Esposito's book published by Wrox Press).


Alex





From fredrik at effbot.org  Tue Dec 12 04:36:29 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 12 Dec 2000 09:36:29 GMT
Subject: Inheriting the @ sign from Ruby
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <m3lmtm2dit.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <xkmZ5.3533$Kd1.409065@newsb.telia.net>

Michael Hudson wrote:
> Also, I think it looks horrible, but that could just be me.

nope.

</F>





From wolfson at midway.uchicago.edu  Fri Dec 29 21:40:57 2000
From: wolfson at midway.uchicago.edu (Ben Wolfson)
Date: Sat, 30 Dec 2000 02:40:57 GMT
Subject: Great New Book
References: <t4o1vr8hktk531@corp.supernews.com> <92ij47$n3p$1@bob.news.rcn.net> <slrn94pnrk.irm.grey@teleute.rpglink.com> <92inl3$a5o$1@panix2.panix.com>
Message-ID: <ZWb36.104$x3.1604@uchinews>

In article <92inl3$a5o$1 at panix2.panix.com>, Aahz Maruch <aahz at panix.com> wrote:
>In article <slrn94pnrk.irm.grey at teleute.rpglink.com>,
>Steve Lamb <morpheus at here.not.there> wrote:
>>On 29 Dec 2000 17:54:15 GMT, Fred Yankowski <fcy at enteract.com> wrote:
>>>
>>>I actually prefer heavier paper since it allows me to wield my
>>>highlighter with abandon without worrying about marring the backside
>>>of the marked page.
>>
>>    Not to mention 2'5" isn't all that thick.  I have fictional paperbacks
>>that are that thick.  :)
>
>Really?  You have paperbacks more than two feet thick????  (No fair
>counting all the Robert Jordans as a single book.)

He did say the books were fictional.



-- 
Barnabas T. Rumjuggler
I say, an object can move whether or not an "external (male, masculine) 
force" chooses to act upon (rape) it!
 -- David Zacuto


From pulsar at qks.com  Fri Dec 15 20:35:21 2000
From: pulsar at qks.com (David Simmons)
Date: Sat, 16 Dec 2000 01:35:21 GMT
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <3bFZ5.382$Y2.6323@news> <3A3AC0C4.8E13D4CC@home.com>
Message-ID: <tFz_5.33542$M5.1233497@news1.frmt1.sfba.home.com>

"James A. Robertson" <jarober at home.com> wrote in message
news:3A3AC0C4.8E13D4CC at home.com...
> Chris Ryland wrote:
> >
> > Russell--
> >
> > Overall, good show! Some questions/observations.
> >
> > "Russell E. Owen" <owen at astroNOJNK.washington.edu.invalid> wrote in
message
> > news:915qlh$89fa$1 at nntp6.u.washington.edu...
> > > [...]
> > > On the other hand, it is not as powerful, clean and regular as
> > > smalltalk. Limitations include:
> > > - The language is somewhat cluttered and irregular because it has
> > > procedures, functions and methods. Hence arbitrary decisions had to be
> > > made in some areas as to how to implement things, which can lead to
some
> > > hunting through the manuals.
> >
> > But, as we all know, languages that give you an "on ramp" from other
more
> > familiar paradigms often succeed against "more clean/powerful"
languages.
> > (E.g., Java was an easy "reach" for many C++ programmers.)
> >
> > I.e., Smalltalk's fierce O-O cleanliness may work against it for general
> > acceptance. (That, and the fact that every Smalltalk vendor in the world
> > (all 2 of them ;-) fell down and played dead at the first appearance of
the
> > (apparent) Java juggernaut.)
>
> All two of them?
>
> Cincom VisualWorks, ObjectStudio, VSE
> IBM VAST
> Gemstone Gemstone/S
> Object-Arts Dolphin
> ObjectConnect Smalltalk/MT
> eXept Smalltalk/X
> QKS SmalltalkAgents, SmallScript (.NET)
>
> And Disney has Squeak open sourced

And you really want to add:

    OTI Smalltalk (for embedded systems -- now owned by IBM and the basis
for VAST)

    PocketSmalltalk (for PDA/embedded systems)

    Bistro Smalltalk for JVM

-- Dave Simmons [www.qks.com / www.smallscript.com]
  "Effectively solving a problem begins with how you express it."

>
> Somewhat more than 2.
>
>
>
> --
> James A. Robertson
> Technical Product Manager (Smalltalk), Cincom
> jarober at mail.com
>
> <Talk Small and Carry a Big Class Library>




From benpark at my-deja.com  Fri Dec 15 12:45:00 2000
From: benpark at my-deja.com (Ben Park)
Date: Fri, 15 Dec 2000 17:45:00 GMT
Subject: comments of python by a perl user
Message-ID: <91dlar$7s3$1@nnrp1.deja.com>


A comparison of Perl and Python by a perl user:
http://www.perl.com/pub/language/versus/python.html

Here is what he/she considers as the annoying things about python:

One of those most annoying things about Python for many people is that
it intuits block structure from indentation alone.

Another thing people consider a feature in Python over Perl is that no
one is ever going to turn around and write the whole thing in an
entirely different way as they might in Perl.

I am not starting a war here. Just for information.


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


From urner at alumni.princeton.edu  Wed Dec 20 14:50:17 2000
From: urner at alumni.princeton.edu (Kirby Urner)
Date: Wed, 20 Dec 2000 11:50:17 -0800
Subject: real life examples
References: <u48q19.cth.ln@adsl.bapnt>
Message-ID: <76324tok2522ngc0viv0v1uqnsr8qn655l@4ax.com>

huber <nospam at wanted.com> wrote:

>i try to teach python to my son (13 years old)!
>but instead of stupid for loops, classes nobody ever needs and so on .....
>are there some good real life examples a child could be interested in ?
>
>thanks
>udo 

Cryptology offers some ideas for doing Python (from
easy to not-so-easy):  

http://www.inetarena.com/~pdx4d/ocn/clubhouse.html

Although Alice is fine, you can also go a long with 
with Povray -- still pictures.

Using Python to study basic math concepts is my approach
for the younger set.  The following essay is too hard 
for a 13 year old -- is directed more at math teachers 
-- but has ideas for using Python to climb a math 
ladder:

http://www.inetarena.com/~pdx4d/ocn/numerarcy0.html

Kirby



From mikael at isy.liu.se  Wed Dec  6 07:11:54 2000
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Wed, 06 Dec 2000 13:11:54 +0100 (MET)
Subject: Calculating factorial
In-Reply-To: <ya0vgsx3h2f.fsf@levding.ifi.uio.no>
Message-ID: <XFMail.001206131154.mikael@isy.liu.se>

On 06-Dec-00 Roger Hansen wrote:
 >  Agree! it's a nice version. But I think the recursive version is even
 >  clearer. 
 >  
 > >>> def fact(n):
 >  ...     if n > 1:
 >  ...         return n*fact(n-1)
 >  ...     return 1
 >  ... 
 > >>> fact(4)
 >  24
 >  
 >  You can change line 3 with 
 >  ...         return long(n)*fact(n-1)
 >  
 >  and you have long integers

The result of the proposed function with long(n) is of type long for
all n>1. I would actually prefer that the function returns type long 
for all legal inputs. The following does that.

>>> def fact(n):
...     if n > 1:
...         return n*fact(n-1)
...     return 1L
... 


/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    06-Dec-00
Time:    13:05:48

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

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


From qdlaty at wielun.dhs.org  Fri Dec 22 02:20:39 2000
From: qdlaty at wielun.dhs.org (QdlatY)
Date: Fri, 22 Dec 2000 08:20:39 +0100
Subject: saving dictionaries
Message-ID: <3A4300C7.FA4192AF@wielun.dhs.org>

Hello!

How to easily save and load contents of dictionaries (2 elements + key)
to file?

Yours
QdlatY


From shindich at my-deja.com  Fri Dec  8 21:06:41 2000
From: shindich at my-deja.com (shindich at my-deja.com)
Date: Sat, 09 Dec 2000 02:06:41 GMT
Subject: using execfile() in class/instance methods
References: <slrn932qna.rtk.wcuddy@lserv.ja10629.home>
Message-ID: <90s43f$jn0$1@nnrp1.deja.com>

In article <slrn932qna.rtk.wcuddy at lserv.ja10629.home>,
  wcuddy at lserv.ja10629.home (Wayne Cuddy) wrote:
> I am attempting to use the execfile() function to "source" a file
that
> has valid python code to create a diction object:
>
> bn { 'key' : 'value1',
>         'key2' : 'value2' }
>
> I am issueing this command from a class constructor method and would
> like bn to be bound the class instance.  So in the future I can say:
>         print self.bn['key']
>
> Is this possible?  I assume I am screwing up the global() and locals
()
> arguments to the function which I can say that I don't fully
> understand.  Also is there another better/worse way to do this???just
> so I can see what else is out there..
>
> Thanks in advance,
>
> Wayne
>
>
Try the following:
>>> class Foo: pass

>>> f = Foo ()
>>> exec "f.bn = { 'key' : 'value1',  'key2' : 'value2' }"
>>> f.bn
{'key': 'value1', 'key2': 'value2'}

I hope this helps!

Regards,

Alex Shindich
visit http://www.shindich.com/


Sent via Deja.com http://www.deja.com/
Before you buy.


From loewis at informatik.hu-berlin.de  Sat Dec  9 05:44:47 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 09 Dec 2000 11:44:47 +0100
Subject: Link problems embedding python in Linux. HELP!
References: <90r5kf$q9j$1@nnrp1.deja.com> <p6qr93iiqf4.fsf@informatik.hu-berlin.de> <90rki8$796$1@nnrp1.deja.com>
Message-ID: <p6qy9xpan6o.fsf@informatik.hu-berlin.de>

jvanar1 at my-deja.com writes:

> Thanks Martin but that didn't help any. 

Why not? What exactly did you do (after my suggestion), and what
exactly was the result?

> What values do PYTHONHOME && PYTHONPATH be set to??

When linking an executable, these variables have no significance at
all.  They are best left unset.

When running the embedded Python, then that should be able to find its
files in the python-configure-time --prefix, but again, during
linking, the PYTHONPATH does not matter at all.

Regards,
Martin


From richard at NOstarfighterSPAM.freeuk.com  Thu Dec  7 15:20:33 2000
From: richard at NOstarfighterSPAM.freeuk.com (Richard)
Date: Thu, 7 Dec 2000 20:20:33 -0000
Subject: PIL binary for Python 2.0 for Windows ?
References: <3a2f472f$0$235@hades.is.co.za> <KZOX5.2015$Kd1.294314@newsb.telia.net>
Message-ID: <IlSX5.24763$R77.1876538@nnrp4.clara.net>

I have downloaded the dll & the PIL tar file Imaging-1.1.1.tar.gz.

Please can you tell me where these file should be copied? (My Python 2.0 is
istalled under C:\Python20).

thanks,
Richard


"Fredrik Lundh" <fredrik at effbot.org> wrote in message
news:KZOX5.2015$Kd1.294314 at newsb.telia.net...
> Leon Booyens wrote:
> > Where can I find a binary distribution for Windows of the Python Image
> > Library 1.1.1. (It is not as yet available from Pythonware as a binary
and I
> > unfortunatly have no means of compiling the code on a windows platform)
>
> http://effbot.org/pil
>
> </F>
>
>




From gregj at pobox.com  Wed Dec  6 02:46:24 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Wed, 06 Dec 2000 07:46:24 GMT
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com>
Message-ID: <k9mX5.117682$U46.3888112@news1.sttls1.wa.home.com>

"Nigel" <nigeljennison at cs.com> wrote in message
news:90k3pq$50p$1 at nnrp1.deja.com...
> I am doing some research and I would be grateful if anybody could give
> me some history on the origins of Python.

Amazing. Public education strikes again.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From aleaxit at yahoo.com  Sun Dec 10 12:43:48 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 10 Dec 2000 18:43:48 +0100
Subject: coding standards in cpython
References: <8vvvmp$93f$1@news.kth.se> <slrn9379is.h4.amk@207-172-146-162.s162.tnt3.ann.va.dialup.rcn.com>
Message-ID: <910fg50165q@news2.newsguy.com>

"A.M. Kuchling" <amk at mira.erols.com> wrote in message
news:slrn9379is.h4.amk at 207-172-146-162.s162.tnt3.ann.va.dialup.rcn.com...
    [snip]
> cc-mode in Emacs or XEmacs, c-set-style "python" will use the right
> style -- 8-space indents, and blocks in this style:
>
>         if (tstate == NULL) {
>                 PyInterpreterState_Delete(interp);
>                 return NULL;
>         }
>
> I forget what that style of brace placement is called.

K&R?  Or, almost -- that's what they used except for
the space between the 'if' and the open-parenthesis,
I think.


Alex





From joconnor at mail.cybermesa.com  Thu Dec  7 16:01:07 2000
From: joconnor at mail.cybermesa.com (Jay O'Connor)
Date: Thu, 07 Dec 2000 21:01:07 GMT
Subject: Server scripting on NT?
Message-ID: <200012072101.OAA23608@mail.cybermesa.com>

An embedded and charset-unspecified text was scrubbed...
Name: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20001207/8b7313de/attachment.ksh>

From vr at acm.org  Sat Dec 16 13:11:58 2000
From: vr at acm.org (Vetle Roeim)
Date: 16 Dec 2000 19:11:58 +0100
Subject: tar files
References: <3A3B89BF.51B2C26C@ex.ac.uk>
Message-ID: <yaybsuc8ccx.fsf@einn.ifi.uio.no>

* M. J. Fisher
> Hi,
> 
> I am trying to find some code for unpacking tarred files and I am not
> sure why to look, does anybody have any ideas? I would also appreciate
> an example if any one had one,

well.. I started making a tar-module in pure Python, but I never had
time to finish it. you can use it if you want, and is able to
understand it.. it was written wuick and dirty by examining the
documentation and source code of GNU Tar.

<URL:http://www.ifi.uio.no/~vetler/PyTar_0.1.tgz>

maby I'll get around to finishing the darn thing someday.


vr
-- 
touch eyballs to screen for cheaper laser surgery


From gtalvola at NameConnector.com  Tue Dec 19 12:44:52 2000
From: gtalvola at NameConnector.com (gtalvola at NameConnector.com)
Date: Tue, 19 Dec 2000 17:44:52 GMT
Subject: Source Code Management
References: <fb4f3tsgffnmvme64aol1cn397te7ln349@4ax.com> <91aq96$tba$1@nnrp1.deja.com> <698_5.3930$Kd1.455411@newsb.telia.net> <3a3d1470.5243319@news.laplaza.org>
Message-ID: <91o6qi$681$1@nnrp1.deja.com>

In article <3a3d1470.5243319 at news.laplaza.org>,
  mats at laplaza.org (Mats Wichmann) wrote:
> Perforce, in fact, provided my incentive to learn Python.  I picked
> Perforce for a particular client project because it was the easiest to
> deploy and the small number of developers involved made it
> cost-effective.  When we needed some scripts to extend functionality,
> several of them were in Python...

You're not the only one -- I too learned Python because of Perforce.
I had no idea at the time how widely useful and productive (and fun)
Python would turn out to be...

- Geoff


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


From dale at out-think.NOSPAMco.uk  Wed Dec  6 08:25:03 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Wed, 06 Dec 2000 13:25:03 +0000
Subject: Rummaging around inside an Access DB
References: <c5es2ts2l66bhg8qb7ne28cj0qa3gi9595@4ax.com>
Message-ID: <hefs2t8tpsu3f9h746o84ns6sspa2p99dp@4ax.com>

Dale Strickland-Clark <dale at out-think.NOSPAMco.uk> wrote:

>
>I've been trying to dig some information out of an Access DB using Python. I need to access as much
>information as possible about the composition of all the reports. 
>
>After a few false starts and lots of head scratching (I can't find a help file on the object model
>anywhere!!), I found something in the Python Win32 distribution called daodump.py, presumably
>written by Mark H.
>
>This helped a lot and produced a listing of all the reports but I can't seem to break into the
>Report object.
>
>This is what I've come up with so far:
>
>
>from win32com.client import Dispatch
>import sys
>eng = Dispatch("DAO.DBEngine.36")
>db = eng.OpenDatabase(sys.argv[1])
>rep = db.Containers.Item('Reports')
>for doc in rep.Documents:
>	print doc.Name
>
>
>Does anyone know how I can get at the cuts of the Report object? The only other properties I can

No! Not "cuts" but "guts", as in "internals". Sorry about that. I'm sur my tryping is getitng wosre.
--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants




From jcm at bigskytel.com  Thu Dec 21 02:36:38 2000
From: jcm at bigskytel.com (David Porter)
Date: Thu, 21 Dec 2000 00:36:38 -0700
Subject: Replay commands
In-Reply-To: <91rep9$tev$1@nnrp1.deja.com>; from pinhong@eecs.berkeley.edu on Wed, Dec 20, 2000 at 11:19:11PM +0000
References: <91rep9$tev$1@nnrp1.deja.com>
Message-ID: <20001221003637.A17989@bigskytel.com>

* pinhong at eecs.berkeley.edu <pinhong at eecs.berkeley.edu>:

> When using Python interactively, I need to save my command history
> for editing and reuse. Is that possible?

Yes.

An example of it being done can be found in IPP, the Interactive Python
Prompt at <~ > mc

~ > locate PythonInterpreter
~ > locate PythonPrompt     
~ > locate interpreter
/suse/usr/doc/packages/python/html/lib/interpreter-objects.html
/usr/lib/gaby/plug-ins/interpreter
/usr/lib/gaby/plug-ins/interpreter/libpython.la
/usr/lib/gaby/plug-ins/interpreter/libpython.so
/usr/lib/gaby/plug-ins/interpreter/libpython.so.0
/usr/lib/gaby/plug-ins/interpreter/libpython.so.0.0.0
/usr/share/doc/gaby/C/interpreters.html
/usr/share/doc/gaby/README.plug-ins_interpreter_python
/var/www/python/html/lib/interpreter-objects.html
/win/Python20/Doc/lib/interpreter-objects.html
/win/Python20/Doc/mac/interpreter.html
~ > mc

~ >   
~ > 
~ > 
~ > mc 

You have mail in /var/spool/mail/dap
~ > pyrun ipp
Reading /home/dap/.pyrun.gz
Building script list.
Rebuilding /home/dap/.pyrun.gz
Building file list... this can take a while.
Reading /home/dap/.pyrun.gz
Building script list.
Script found at /home/dap/py/ipp/ipp.py
Executing.
Python 1.5.2 (#0, Apr  3 2000, 14:46:48)  [GCC 2.95.2 20000313 (Debian
GNU/Linux)] on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
(IPP)
Type ? for more help

>>> ?
        ipp -> the interactive python prompt
        
        An enhanced console for Python.
        Features are:
        - readline support if present
        - completion in the local namespace, eg. type TAB twice at the prompt.
        - logging of input , see command-line options
        - Systemshell escape by the ! , eg !ls
        - Pythonshell commands, starting with a @
        - some of these are  @ls, @pwd, @cd
        - Keeps track of locally defined variables @who, @whos
        - Show documentation with a ? , eg ?list
        - evaluation of a personal config file which can be used to extend
the shell
        - expansion of history lines by a starting '.'
        
        Command line options:
        -i      | execute remaining files and than jump to the prompt
        -l      | logging in standard-log file
        -L file | name a special log file and log into it
        -s      | don't read $HOME/.ipprc
        -c file | read this as ipprc

        
>>> 
~ > pyrun ipp -l
Reading /home/dap/.pyrun.gz
Building script list.
Script found at /home/dap/py/ipp/ipp.py
Executing.
Python 1.5.2 (#0, Apr  3 2000, 14:46:48)  [GCC 2.95.2 20000313 (Debian
GNU/Linux)] on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
(IPP)
Type ? for more help

>>> 
~ > pyrun ipp -L pylog
Reading /home/dap/.pyrun.gz
Building script list.
Script found at /home/dap/py/ipp/ipp.py
Executing.
Python 1.5.2 (#0, Apr  3 2000, 14:46:48)  [GCC 2.95.2 20000313 (Debian
GNU/Linux)] on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
(IPP)
Type ? for more help

>>> agd
Traceback (innermost last):
  File "<console>", line 1, in ?
NameError: agd
>>> reg
Traceback (innermost last):
  File "<console>", line 1, in ?
NameError: reg
>>> hg
Traceback (innermost last):
  File "<console>", line 1, in ?
NameError: hg
>>> trgh
Traceback (innermost last):
  File "<console>", line 1, in ?
NameError: trgh
>>> th
Traceback (innermost last):
  File "<console>", line 1, in ?
NameError: th
>>> gre
Traceback (innermost last):
  File "<console>", line 1, in ?
NameError: gre
>>> re
Traceback (innermost last):
  File "<console>", line 1, in ?
NameError: re
>>> 45yt
  File "<console>", line 1
    45yt
       ^
SyntaxError: invalid syntax
>>> 
>>> htr
Traceback (innermost last):
  File "<console>", line 1, in ?
NameError: htr
>>> ht
Traceback (innermost last):
  File "<console>", line 1, in ?
NameError: ht
>>> ht
Traceback (innermost last):
  File "<console>", line 1, in ?
NameError: ht
>>> gr
Traceback (innermost last):
  File "<console>", line 1, in ?
NameError: gr
>>> rg
Traceback (innermost last):
  File "<console>", line 1, in ?
NameError: rg
>>> rge
Traceback (innermost last):
  File "<console>", line 1, in ?
NameError: rge
>>> 
~ > less py
py     pylog  pyrun  
~ > less pylog 
F10 key ==> File   Edit   Search   Buffers   Windows   System   Help
* pinhong at eecs.berkeley.edu <pinhong at eecs.berkeley.edu>:

> When using Python interactively, I need to save my command history
> for editing and reuse. Is that possible?

Yes.

An example of it being done can be found in IPP, the Interactive Python
Prompt at <http://starship.python.net/crew/jhauser/miscellany.html>.


David



From kragen at dnaco.net  Sat Dec 30 18:28:05 2000
From: kragen at dnaco.net (Kragen Sitaker)
Date: Sat, 30 Dec 2000 23:28:05 GMT
Subject: Great New Book
References: <t4o1vr8hktk531@corp.supernews.com> <3A4D40F3.134D839B@NOtheworldSPAMMY.com>
Message-ID: <9cu36.10$9s1.981@e420r-atl2.usenetserver.com>

In article <3A4D40F3.134D839B at NOtheworldSPAMMY.com>,
David Lees  <debl at NOtheworldSPAMMY.com> wrote:
>Actually, another new Python book has also come out recently, called
>Python Developers's Handbook by Lessa.

Does she include tips on dragon-riding?

Let us know how it is after you read it.


-- 
<kragen at pobox.com>       Kragen Sitaker     <http://www.pobox.com/~kragen/>
Perilous to all of us are the devices of an art deeper than we possess
ourselves.
       -- Gandalf the White [J.R.R. Tolkien, "The Two Towers", Bk 3, Ch. XI]




From tim.one at home.com  Wed Dec 20 05:45:41 2000
From: tim.one at home.com (Tim Peters)
Date: Wed, 20 Dec 2000 05:45:41 -0500
Subject: Python's scope rules
In-Reply-To: <3A40813B.C04D21C@et.dtu.dk>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEMCIEAA.tim.one@home.com>

[Brian Elmegaard]
> I am not a computer scientist, just an engineer knowing how to
> program in Fortran, Pascal,... So I after reading some of the
> python material on the web, I decided to give it a try and is
> now using the language some, finding it fun and easy.

It will stay that way too -- provided you don't pay too much attention to
computer scientists <wink>.

> On usenet I have now learned from skilled scientists, that Python
> has some deficiencies regarding scope rules, and I am not capable
> of telling them why it has not (or that it has).

Some of the arguments pro and con are summarized in PEP (Python Enhancement
Proposal) 227, at

    http://python.sourceforge.net/peps/pep-0227.html

Python's current scheme is very usable in practice; some people will never
like it, though, because it's not exactly like anything else.

> In Aaron Watters 'The wwww of python' I have read that Python uses
> lexical scoping with a convenient modularity. But, other people
> tell me it does not.  Some say it uses dynamic scoping, some say
> it uses it own special 'local-global-builtin' scope. What is right?

The rules are detailed in the language reference manual's section 4.1 ("Code
blocks, execution frames, and namespaces"), at

    http://www.python.org/doc/current/ref/execframes.html

The fundamental rule is that a name is local to a block if and only if it
(a) is assigned to somewhere in the block, and (b) does not appear in a
"global" statement within the block.  Note that this applies to *all* names!
In particular, there's nothing special about function, class or module names
as opposed to variable names.

It's proper to call this lexical scoping because the local vs non-local
decision is made wholly at compile-time; but most people who use terms like
"lexical scoping" have backgrounds in languages where lexical scopes nest
arbitrarily deep (the subject of PEP 227), and are offended that Python is
simpler than that <wink>.

> The above mainly is a theoretical question. A more practical
> example which I agree seems a bit odd is:

I'll explain this in detail.  That may make it *look* complicated.  But it's
not:  the detail is needed in order to bring out how uniform it is.  Things
are going on that you don't realize at first.  There is in fact no reason
you ever need to realize some of these, but if you're going to plumb the
depths of name resolution in Python then you need to know them:

> Python 1.5.2 (#1, Mar  9 2000, 17:40:34)  [GCC 2.95.2 19991024
> (release)] on hp-uxB
> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>> x=24

x is assigned to here, so x is local to *its* code block.  Its code block
happens to be the module whose name is "__main__", which module Python
invented for you because you're in an interactive session.  Code always
executes in the context of *some* module.  Modules have the peculiar
property that their local and global namespaces are the same.

> >>> def foo():

Something you weren't expecting:  "def" is an executable statement in
Python, and also acts like an assignment.  The appearance of "def foo()"--
just like "x=24" --makes foo a name local to module __main__ too.  A def
statement also introduces a new code block.

> ...     y=2
> ...

y is assigned to within foo's code block, so y is a name local to foo.

> ...     print x+y

x is not assigned to within foo's code block, so x is not local to foo.  It
will be looked up in foo's global namespace instead (== the local namespace
of module __main__).

> >>> def bar():

Introduces "bar" as a name local to __main__, and starts a new code block.

> ...     x=11

x is assigned to within bar's code block, so this x is local to bar (and has
nothing whatsoever to do with the name x that's local to __main__).

> ...     foo()
> ...

Note!  The name "foo" is not assigned to within bar's code block, so the
name foo is not local to bar.  It will be looked up in bar's global
namespace instead (== the local namespace of module __main__).  This is
*exactly* the same rule that caused us to look up "x" in the global
namespace within function "foo":  there's nothing special about function
names.  All names get resolved by the same little set of rules.

> >>> bar()

bar is local to __main__, so gets resolved in __main__'s local namespace.

> 26
> >>>

> Well, I can live with it.

More easily than you may think, too.

> In Fortran I would always put all the variables to be
> passed in the parameter list.

Or in common blocks.  While in Pascal, you would declare names local to a
function within that function.  But Python has no declarations (ignore
"global" for now), and needs *some* rule to determine what is and isn't
local.  The "local if and only if bound somewhere in the block" rule usually
works out just right by magic.  The consequence that may be unique to Python
is that, from within a function contained in a module, it's impossible to
assign directly to a module-local name unless that name is listed in a
"global" statement within the function (indeed, that's why the "global"
statement exists).  I think that's a Good Thing:  mucking with global
variables makes code hard to understand, so if you *have* to go changing the
bindings of globals, it's Good that the language makes you say something
extra to point that out.

> Other people say that Fortran is not easy, and this example shows how
> stupid python is. It does?

I suspect you've got Fortran77 in mind -- and F77 also has simple rules.

> Hopefully not, but probably, this is a faq, but I have not been
> able to find anything on the matter.

It's hard to find.  Guido (van Rossum, Python's father) has written that
Python's "2-level scope nesting" was in part a reaction against the abuses
of deep lexical nesting common in large Pascal programs of the time.  I
didn't like Python's scope rules when I first saw the language (I didn't
think "simple", I thought "simple-minded"!), but that argument persuaded me
to give it a chance -- since I had also been on the receiving end of other
peoples' insanely nested Pascal monsters.  Within a year I learned how to
make good use of modules and classes in Python to manage names, and Python's
scope rules never bothered me again.  I think they're great!  Unfortunately,
they're not what anyone coming from other languages expect, and that's why
PEP 227 is being seriously considered for Python 2.1.

it's-all-lambda's-fault-but-that-tale-of-woe-is-for-another-day-ly
    y'rs  - tim




From root at rainerdeyke.com  Fri Dec 15 12:16:33 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Fri, 15 Dec 2000 17:16:33 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <r0pzoi11lhz.fsf@black.cs.washington.edu> <5jvZ5.3588$Kd1.418770@newsb.telia.net> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com> <kuu28971kp.fsf@lasipalatsi.fi> <3A380E4E.8537A0FB@san.rr.com> <91a1vr02b02@news2.newsguy.com> <Co9_5.25465$x6.13756579@news2.rdc2.tx.home.com> <91bbq102mtk@news1.newsguy.com> <rWd_5.26820$x6.14119694@news2.rdc2.tx.home.com> <91d5el0me7@news1.newsguy.com> <rOr_5.31120$x6.15241856@news2.rdc2.tx.home.com>
Message-ID: <Rls_5.31136$x6.15277726@news2.rdc2.tx.home.com>

"Rainer Deyke" <root at rainerdeyke.com> wrote in message
news:rOr_5.31120$x6.15241856 at news2.rdc2.tx.home.com...
> Yes, in-place modification should be supported.  I had something like this
> in mind:
>
> class UserInt:
>   def __init__(self, value):
>     self.data = value
>   def __add__(self, other):
>     if isinstance(other, UserInt):
>       return self.__class__(self.data + other.data)
>     else:
>       return self.__class__(self.data + other)
>   __radd__ = __add__
>   def __iadd__(self, other):
>     if isinstance(other, UserInt):
>       self.data += other.data
>     else:
>       self.data += other
>   # Repeat for all other operations
>
> UserInt as written is more properly UserNumber, since it supports longs
ints
> and floats.

Of course __iadd__ actually belongs in MutableInt instead of UserInt, but
that didn't occur to me until after I'd written it.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From cjensen at bioeng.ucsd.edu  Tue Dec  5 17:08:33 2000
From: cjensen at bioeng.ucsd.edu (Curtis Jensen)
Date: Tue, 05 Dec 2000 14:08:33 -0800
Subject: parallel port access?
References: <3A268378.4DA1B831@sympatico.ca> <3A272881.6F444750@engcorp.com>
Message-ID: <3A2D6761.814ABE8E@bioeng.ucsd.edu>

Peter Hansen wrote:
> 
> DOC wrote:
> >
> > I would like to be able to control the individual
> > lines of the parallel port using Python.
> >
> > Anyone know of any Class packages to do this or
> > do I need to write one myself?
> 
> What platform?

There has been so many questiona about sieral and parallel port access
throuhg Python.  How come it hasn't been put into the distrabution?  I
realize that it is platform specific, but so are other things that are
in Python.  Or even put it in as a platform specific module like the gl
module.

-- 
Curtis Jensen
cjensen at bioeng.ucsd.edu
http://www-bioeng.ucsd.edu/~cjensen/
FAX (425) 740-1451


From emile at fenx.com  Thu Dec  7 22:57:55 2000
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 07 Dec 2000 19:57:55 -0800
Subject: Stupid Newbie Question:  Python FAQ?
References: <3A304F1C.DF1224C2@springhaven.org>
Message-ID: <3A305C43.3090601@fenx.com>

========================================================================

Everything you want is probably one or two clicks away in these pages:

       Python.org's Python Language Website is the traditional
       center of Pythonia
           http://www.python.org although PythonLabs.com bills itself as 
"The Python Source",
       which is becoming increasingly true in 2000
           http://www.pythonlabs.com/  PythonWare complements the digest 
you're reading with the
       daily python url
            http://www.pythonware.com/daily comp.lang.python.announce 
announces new Python software.  Be
       sure to scan this newly-revitalized newsgroup at least weekly.
           http://deja.com/group/comp.lang.python.announce  Andrew 
Kuchling writes marvelous summaries twice a month of
       the action on the python-dev mailing list, where the future
       of Python is truly determined
           http://www.kuchling.com/python/dev/  The Vaults of Parnassus 
ambitiously collects Python resources
           http://www.vex.net/~x/parnassus/ Much of Python's real work 
takes place on Special-Interest Group
       mailing lists
           http://www.python.org/sigs/  Python Consortium emerges as an 
independent nexus of activity
           http://www.python.org/consortium  Cetus does much of the same
           http://www.cetus-links.de/oo_python.html  Python FAQTS
           http://python.faqts.com/  Python To-Do List anticipates some 
of Python's future direction
           http://www.python.org/cgi-bin/todo.py  Python Journal is at 
work on its second issue
           http://www.pythonjournal.com  Links2Go is a new 
semi-automated link collection; it's impressive
       what AI can generate
           http://www.links2go.com/search?search=python  Archive probing 
trick of the trade:
 
http://www.dejanews.com/dnquery.xp?QRY=&DBS=2&ST=PS&defaultOp=AND&LNG=ALL&format=threaded&showsort=date&maxhits=100&groups=comp.lang.python 
   Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
   http://purl.org/thecliff/python/url.html or
 
http://www.dejanews.com/dnquery.xp?QRY=~g%20comp.lang.python%20Python-URL%21 
   Suggestions/corrections for next week's posting are always welcome.
[http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
e-mail us ideas directly.]

To receive a new issue of this posting in e-mail each Monday morning,
ask <claird at neosoft.com> to subscribe.  Mention "Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.



Johnson \"da'fox\" Earls wrote:

> Hello all,
> 
>     Chiming in with my first Stupid Newbie Question:  Is there a FAQ
> available for python, so that I can avoid asking more Stupid Newbie
> Questions? :)



From pblanchette at pixelsystems.com  Tue Dec  5 13:44:47 2000
From: pblanchette at pixelsystems.com (Patrick Blanchette)
Date: Tue, 05 Dec 2000 18:44:47 GMT
Subject: win32security.AdjustTokenPrivileges
Message-ID: <3A2D378E.7050504@pixelsystems.com>

I,
  I'd like to reboot a windows NT station using a Python script.  It is 
supposed to be possible using the Python extensions for Windows.  I try 
the following code but an exception is thrown after the call to 
AdjustTokenPrivileges:
-----------------------------------------------
hProcess = GetCurrentProcess()
processToken = OpenProcessToken(hProcess, TOKEN_ADJUST_PRIVILEGES | 
TOKEN_QUERY)
procPrivValue = LookupPrivilegeValue (None, SE_SHUTDOWN_NAME)
AdjustTokenPrivileges (processToken, 0, (provValue, SE_PRIVILEGE_ENABLED) )
ExitWindowsEx(EWX_REBOOT | EWX_FORCE, 0)
-->
Traceback (innermost last):
 File "<interactive input>", line 1, in ?
TypeError: A TOKEN_PRIVILEGES object must be a tuple of (LARGE_INTEGER, 
int)
-----------------------------------------------

How should I use the AdjustTokenPrivileges function?

thanks



From fredrik at effbot.org  Wed Dec 13 15:22:01 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 13 Dec 2000 20:22:01 GMT
Subject: unwrapping argument tuples
References: <3A37D870.F3D01433@kla-tencor.com>
Message-ID: <JTQZ5.3841$Kd1.441313@newsb.telia.net>

Prashanth Mundkur wrote:
> is there a way to unwrap a tuple argument to a function and
> match it to the parameters?
>
> that is, if
>
> >>>lst = [1, 2, 3]
>
> >>>def func (p1,p2,p3):
>        return p1+p2+p3
>
> then I want to apply func somewhat as
>
> >>>func(lst)
> 6
>
> such that the elements of lst are unwrapped and and matched
> to p1, p2, p3

    apply(func, lst)

or

    func(*lst) # 2.0 and later

also see:

    http://www.python.org/doc/FAQ.html#4.31
    "How do I call a function if I have the arguments in a tuple"

(in your case, lst is a list, not a tuple.  but in more recent version
of Python, apply accepts lists as well...)

</F>




From t.keil at zvs.zgs.de  Tue Dec 12 17:22:10 2000
From: t.keil at zvs.zgs.de (Thomas Keil)
Date: Tue, 12 Dec 2000 14:22:10 -0800
Subject: S: source for python 1.0.1 (DOS)
Message-ID: <3A36A512.151D@zvs.zgs.de>

Hello,

I'm looking for modules (os, string, ...) for Python 1.0.1 (it works
fine on my HP 200LX :-)
Where can I get these files?
Thank you in advance!

 _  o         
| )_/#        
 T  <T        Th. Keil


From support at internetdiscovery.com  Sat Dec  9 18:40:00 2000
From: support at internetdiscovery.com (Mike Clarkson)
Date: Sat, 09 Dec 2000 23:40:00 GMT
Subject: reading X resources from Tkinter
References: <90n3at$130f$1@chaos.obstruction.com> <p6qg0jyk5yw.fsf@informatik.hu-berlin.de>
Message-ID: <3a32a784.4490775@24.0.228.33>

On 08 Dec 2000 21:34:15 +0100, Martin von Loewis
<loewis at informatik.hu-berlin.de> wrote:

>guy at obstruction.com (Guy Middleton) writes:
>
>> It's not clear to me (maybe I am not reading carefully enough) if one can
>> query the X resources directly, or if you can copy X resources into the
>> option database.  Is this possible?

You can't query the X resources directly, but you can override then
with the [option add] command.

>I believe that's not possible. The X resource database is an Xt
>feature, and Tk does not use Xt (nor does it emulate access to the
>resource database).

No that's not true. Look at the Tk documentation for the option
command.

Mike.


From Bill.Scherer at VerizonWireless.com  Wed Dec 27 10:48:11 2000
From: Bill.Scherer at VerizonWireless.com (Bill Scherer)
Date: Wed, 27 Dec 2000 10:48:11 -0500
Subject: Controlling the Parallel Port
References: <mvuj4tkmn22gbth1a6t8fag39902qeeu9l@4ax.com> <mailman.977928790.25688.python-list@python.org> <pe1k4tsr7u017n45gf09f20fc8k1int2ci@4ax.com>
Message-ID: <3A4A0F3B.59DD7AD2@VerizonWireless.com>

SWIG, the Simplified Wrapper and Interface Generator, is available at
http://www.swig.org/
It makes it easier, sometimes trivial, to make a module for your
favorite interpreter from a c lib.

For the parapin lib, it was trivial...

If you get SWIG and the parapin lib, you can have a working parapin
module for Python in about five minutes...


-Bill

Simon Faulkner wrote:
> 
> Cheers Bill,
> 
> I have found the link and, indeed, several other c links.
> 
> So, pardon my Python ignorance, but what is SWIG?
> 
> I am already searching...
> 
> Simon
> 
> On Wed, 27 Dec 2000 09:52:43 -0500, Bill Scherer
> <Bill.Scherer at VerizonWireless.com> wrote:
> 
> >Simon
> >
> >There is a parallel port library for linux at
> >
> >http://www.circlemud.org/~jelson/software/parapin/
> >
> >that is simple to SWIG into a Python module.
> >I can send you the .i file for this if you want.
> >
> >I haven't really used it, but it loads and reads the port status just
> >fine.
> >
> >
> >-Bill
> >
> >Simon Faulkner wrote:
> >>
> >> Is it easy to control (read/write) the Parallel Port in linux using
> >> Python?
> >>
> >> I am hoping to use it to drive some simple electronics
> >>
> >> (Usual disclaimer understood - If it smokes then you probably got the
> >> wiring wrong!)
> >>
> >> Simon
> >> Simon Faulkner
> >> --
> >> http://www.python.org/mailman/listinfo/python-list
> 
> Simon Faulkner
> --
> http://www.python.org/mailman/listinfo/python-list

-- 
William K. Scherer
Sr. Member of Applications Staff - Verizon Wireless
Bill.Scherer_at_VerizonWireless.com



From aleaxit at yahoo.com  Mon Dec 11 08:00:50 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 11 Dec 2000 14:00:50 +0100
Subject: Python wrapper for C++ STL?
References: <3A31760E.D157EB54@erols.com> <npfY5.1850$x6.2577114@news2.rdc2.tx.home.com> <90s2lm$ig1$1@nnrp1.deja.com> <bDhY5.1999$x6.2840573@news2.rdc2.tx.home.com> <90s9s1$5oj$1@slb6.atl.mindspring.net>
Message-ID: <912j6c063p@news2.newsguy.com>

"Andrew Dalke" <dalke at acm.org> wrote in message
news:90s9s1$5oj$1 at slb6.atl.mindspring.net...
>
> Rainer Deyke wrote
> >I tend to use a dictionary where all values are None for sets.
>
> For a bit more performance, use a number, like 0 or 1.
> Using None requires a name lookup.

Nice point!  Copying None to a local variable (when you
know you'll need it a zillion times in a loop) may
ameliorate some of this, and the ability to check with
'is' rather than with '==' may give a tiny advantage:

-- finti.py:

def acheck_oneis(r):
    for c in r:
            if c is 1: return

import time
r = range(2,1000000)

for fun in (acheck_noneeq, acheck_noneis,
            acheck_noneleq, acheck_nonelis,
            acheck_oneeq, acheck_noneis):
    t1=time.time();fun(r);t2=time.time()
    print fun.func_name, t2-t1

-- end of finti.py

D:\PySym>python finti.py
acheck_noneeq 2.41399991512
acheck_noneis 2.05300009251
acheck_noneleq 2.01300001144
acheck_nonelis 1.73199999332
acheck_oneeq 1.75299990177
acheck_noneis 2.0830000639

D:\PySym>python finti.py
acheck_noneeq 2.41399991512
acheck_noneis 2.03299999237
acheck_noneleq 1.98200011253
acheck_nonelis 1.68299996853
acheck_oneeq 1.72200000286
acheck_noneis 2.04299998283

D:\PySym>

I'm not sure why 'is' should be measurably
faster for None and myNone, while being a
bit slower for 1 -- maybe it's crucially
dependent on the values one is testing.  But,
at least on my machine, using a local variable
set to None, and testing with 'is', seems to
be a WEE bit faster than the second-best way,
using 1 and testing with '=='.


Alex





From aleaxit at yahoo.com  Tue Dec 12 05:55:25 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 12 Dec 2000 11:55:25 +0100
Subject: function pointer
References: <3A358485.FC22A29@bioeng.ucsd.edu> <slrn93b4bd.8bv.andrew.spambgone.chatham@fizgig.dorm.duke.edu>
Message-ID: <91507302pog@news2.newsguy.com>

"Andrew H. Chatham" <andrew.spambgone.chatham at duke.edu> wrote in message
news:slrn93b4bd.8bv.andrew.spambgone.chatham at fizgig.dorm.duke.edu...
> You could use a CObject, which is what I've done before, though it has
> limitations. A CObject is really just a Python object that holds a C
pointer.
> See http://python.org/doc/current/api/cObjects.html . Of course, doing it
this
> way, you'd have to already know what kind of arguments and return values
the
> function takes and there's also the possibility that someone could pass a
bad
> pointer if they had a CObject from another module, since you cast it to a
"void
> *" to store and retrieve it from the CObject. So if you already know what
kind
> of function it will be a pointer to and don't expect people to go out of
their
> way to break your module, a CObject would be a good choice.

Just a pedantic observation -- there is NO guarantee in the
C standards that a void* can hold a pointer to FUNCTION.  It
IS guaranteed to be able to hold any pointer to DATA, in
such a way that stashing the pointer into a void* and casting
it out again will give the same value back; but, no such
guarantee for pointers to FUNCTIONS.

For example, there used to be, on 16-bit environments for
PC's, compilation modes in which pointers to data would
only use 16 bits, while pointers to functions could use
32 bits.  Here, stashing a function-pointer into a void*
and back again could lead to disastrous truncation.

I guess this is unlikely to bite you in practice, but, if
you rely on that, do please document it clearly, at least!


Alex





From per at sbc.su.se  Fri Dec 22 03:34:39 2000
From: per at sbc.su.se (Per Kraulis)
Date: Fri, 22 Dec 2000 09:34:39 +0100
Subject: [ANN] crng 1.1: Random-number generator extension types (minor upgrade)
Message-ID: <3A43121F.7C9484FE@sbc.su.se>

crng 1.1: Random-number generators (RNGs) implemented as Python
extension types coded in C.

This is a minor upgrade from version 1.0.

The most important changes in version 1.1:

- The script (or module) 'crng_pickle.py' has been added, which
  performs setup for using the crng module with the pickle/cPickle
modules.
- A new utility function 'stir' has been added; it shuffles a list 
  in-place.
- The types Ranlux and MT19937 now have an initialization argument
  'state' and the corresponding attribute, which allows save/retrieve 
  operations.
- The 'repr' function has been properly implemented for all RNG types. 
- A script 'setup.py' has been contributed for installation using the
  Distutils tools.

Official website:  http://www.sbc.su.se/~per/crng/

Documentation:  http://www.sbc.su.se/~per/crng/doc.html

The Python module crng implements random-number generators (RNGs)
based on several different algorithms producing uniform deviates in
the open interval (0,1), i.e. exclusive of the end-point values 0 and
1. A few continuous and integer-valued non-uniform deviates are also
available. Each RNG algorithm is implemented as a separate Python
extension type. The RNG types are independent of each other, but have
very similar interfaces. The entire module is implemented as one
single C source code file.

The GNU General Public License (GPL) applies.

-- 
Per J. Kraulis, Ph.D.                        per at sbc.su.se
Stockholm Bioinformatics Center (SBC)        http://www.sbc.su.se/~per
Dept. Biochemistry, Stockholm University     phone +46 (0)8 - 674 78 17
SE-106 91 Stockholm, SWEDEN                  fax   +46 (0)8 -  15 80 57


From gregj at pobox.com  Wed Dec  6 06:58:52 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Wed, 06 Dec 2000 11:58:52 GMT
Subject: Sort
References: <NlnX5.1444$yv1.2700363@nnrp5.proxad.net>
Message-ID: <0SpX5.117794$U46.3921711@news1.sttls1.wa.home.com>

"Alfred" <scj at mcom.mcom.fr> wrote:

> I'm trying to sort a List of List, on the third col, but I don't find how
to
> do this.

>>> def itemcmp(a, b):
        return cmp(a[2], b[2])

>>> t = [ ['a',2,4], ['g',1,7], ['b',0,2], ['z',8,5] ]
>>> t.sort(itemcmp)
>>> t
[['b', 0, 2], ['a', 2, 4], ['z', 8, 5], ['g', 1, 7]]
>>>

The itemcmp function takes two elements (in this case, lists of at least 3
elements) and compares the third element using the built-in cmp function.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From fidelman at world.std.com  Thu Dec 14 09:39:37 2000
From: fidelman at world.std.com (Miles R. Fidelman)
Date: Thu, 14 Dec 2000 14:39:37 GMT
Subject: Python for REAL Projects ?
References: <3a380d1e.409238@bendigo.news.origin.net.au>
Message-ID: <G5KBE1.8F3@world.std.com>

dragon at hitech.net.au wrote:
: Hello Python People !
: I have been looking at, and trying out Python (under Win NT) and
: although Python SEEMS interesting can it "cut the mustad" on a real
: client project ??

It seems like Zope is being used on "real client projects" (and was
developed to support them). Since Zope runs on Python....




From jp at ulgo.koti.com.pl  Thu Dec 21 22:32:36 2000
From: jp at ulgo.koti.com.pl (Jacek =?iso-8859-2?Q?Pop=B3awski?=)
Date: 22 Dec 2000 03:32:36 GMT
Subject: reading numbers from file
References: <slrn9457uq.9tl.jp@localhost.localdomain> <slrn945jg1.p7a.neelk@alum.mit.edu>
Message-ID: <slrn945iqk.nab.jp@localhost.localdomain>

Neelakantan Krishnaswami wrote:
>I hope this helped.

thanks a lot to you and Bjorn Pettersen, I think I understand everything
this group is *very* helpful and friendly


-- 
Same story, different eyes
Here's where the hero never, never dies

                                          "Hollywood Black" - Ronnie James Dio


From junaftnoon at nospamplzyahoo.com  Fri Dec  8 12:27:26 2000
From: junaftnoon at nospamplzyahoo.com (June Kim)
Date: Sat, 9 Dec 2000 02:27:26 +0900
Subject: The secrets of operator.add
References: <g08zpq51vy.fsf@scumbag.ecs.soton.ac.uk>
Message-ID: <90r5nk$i1r$1@news.nuri.net>

>>> profile.run('map(lambda x:x+1,range(5000))')

         5002 function calls in 0.584 CPU seconds

   Ordered by: standard name

   ncalls  tottime  percall  cumtime  percall filename:lineno(function)
     5000    0.257    0.000    0.257    0.000 <string>:1(<lambda>)
        1    0.325    0.325    0.582    0.582 <string>:1(?)
        1    0.002    0.002    0.584    0.584 profile:0(map(lambda
x:x+1,range(5000)))
        0    0.000             0.000          profile:0(profiler)


>>> profile.run('map(operator.add,(1,)*5000,range(5000))')

         2 function calls in 0.022 CPU seconds

   Ordered by: standard name

   ncalls  tottime  percall  cumtime  percall filename:lineno(function)
        1    0.021    0.021    0.021    0.021 <string>:1(?)
        1    0.002    0.002    0.022    0.022
profile:0(map(operator.add,(1,)*5000,range(5000)))
        0    0.000             0.000          profile:0(profiler)


"Jacek Generowicz" <jmg at ecs.soton.ac.uk> wrote in message
news:g08zpq51vy.fsf at scumbag.ecs.soton.ac.uk...
> Where can I find out about the ways in which the functions in the
> operator module are optimized, and how to use them to best effect in my
> code ?
>
> For example, the following bit of code:
>
> from operator import add
> from profile import run
> N = 50000
> run ('reduce(lambda x,y:x+y, range(N))')
> run ('reduce(add, range(N))')
>
> already throws up a surprise for me:
>
>          50001 function calls in 1.600 CPU seconds
>
>    Ordered by: standard name
>
>    ncalls  tottime  percall  cumtime  percall filename:lineno(function)
>     49999    0.840    0.000    0.840    0.000 <string>:1(<lambda>)
>         1    0.720    0.720    1.560    1.560 <string>:1(?)
>         0    0.000             0.000          profile:0(profiler)
>         1    0.040    0.040    1.600    1.600 profile:0(reduce(lambda
x,y:x+y, range(N)))
>
>
>          2 function calls in 0.060 CPU seconds
>
>    Ordered by: standard name
>
>    ncalls  tottime  percall  cumtime  percall filename:lineno(function)
>         1    0.060    0.060    0.060    0.060 <string>:1(?)
>         0    0.000             0.000          profile:0(profiler)
>         1    0.000    0.000    0.060    0.060 profile:0(reduce(add,
range(N)))
>
>
> It looks as if multiple calls to operator.add are avoided somehow
> . . . or am I failing to understand the profiler's output ?
>
> Is there a way of getting similar advantages in
>
> run ('map(lambda x:x+1, range(N))')  ?
>
> Is there a Python way of currying add?
>
> Jacek



From magnus.heino at rivermen.se  Sun Dec 10 04:27:46 2000
From: magnus.heino at rivermen.se (Magnus Heino)
Date: Sun, 10 Dec 2000 10:27:46 +0100
Subject: imputils problem
References: <aGzY5.6237$x6.4746529@news2.rdc2.tx.home.com>
Message-ID: <90viai$slg$1@merlin.rembrandtstr.bocholt.de>

> I've been having a problem with imputils: a module imported through a
> custom
> importer seems unable in import .pyd modules.  In particular, if I import
> the zipfile module though a custom importer, it is unable to find the zlib
> module.  Does anybody here know how to fix this?

I have it working, but I had to tweek imputils.py a bit.

How are you doing things?

/Magnus


From moshez at zadka.site.co.il  Fri Dec 29 02:11:09 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Fri, 29 Dec 2000 09:11:09 +0200 (IST)
Subject: Python vs. PHP (& Java?)
In-Reply-To: <69BC4C6AE83ED311BC9400805FA7B14AF84D2C@il93exp01.css.mot.com>
References: <69BC4C6AE83ED311BC9400805FA7B14AF84D2C@il93exp01.css.mot.com>
Message-ID: <20001229071109.CE0EBA84F@darjeeling.zadka.site.co.il>

On Thu, 28 Dec 2000, Kemp Randy-W18971 <Randy.L.Kemp at motorola.com> wrote:

> 1.  Could we argue that English is a better language to use around the world 
>     then French or Spanish?

Yes. Personally, I think that would be wrong -- English has a very
hard to learn spelling system.

> 2.  Could we argue that an artificial language, such as Esperanto, is better 
>     to use around the world then a natural language, like English?

Yes. I think this would be wonderful, except political constraints dictate
otherwise.

>   Isn't there an axiom in physics that an experimenter influences the 
>   outcome of his experiment (physicists, help me out here)?  

Widely overused by laymen.
It says that in *some* experiments, measuring one quantity *has* to
interfere with another quantity some specified amount. 

ok-thomas-is-this-off-topic?-ly y'rs, Z.
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From jonathan.gilligan at vanderbilt.edu  Tue Dec  5 15:46:19 2000
From: jonathan.gilligan at vanderbilt.edu (Jonathan Gilligan)
Date: Tue, 5 Dec 2000 14:46:19 -0600
Subject: How do packages and .pth files work under Windows (Py 2.0)
References: <90i4uc$lin$1@news.vanderbilt.edu> <z26X5.15943$R77.1248181@nnrp4.clara.net>
Message-ID: <90jk6r$rga$2@news.vanderbilt.edu>

Thanks, but this doesn't tell me about when python reads .pth files,
which is the crux of my problem. Between this and writing a
__init__.py file which properly initializes the variable __all__, I
managed to work around my immediate problem.

I have managed to resolve another part of my problem---it seems that
Python reads the default value of the registry key
"HKLM\Software\Python\CurrentVersion\2.0\PythonPath" only if it cannot
resolve PYTHONHOME. If it can resolve PYTHONHOME, then it ignores the
default value in "HKLM\Software\Python\CurrentVersion\2.0\PythonPath"
and substitutes some default directories instead for the PYTHONPATH.
It does add any values in keys below PythonPath, but not the value at
the top of the tree.

This is undocumented, but can be found in the C code in getpathp.c.

I'd still like to know how to use .pth files properly, though, if
anyone can help.

Jonathan

"Stuart" <sgm103 at cs.york.ac.uk> wrote in message
news:z26X5.15943$R77.1248181 at nnrp4.clara.net...
> > It would help me enormously if there were a way to get Python to
print
> > out a list of where it looks when it tries to resolve an "import".
Is
> > this possible?
>
> I'm not certain, but i think this does the trick:
>
> import sys
> print sys.path





From glenfant at equod.com.nospam  Fri Dec 15 11:26:33 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Fri, 15 Dec 2000 17:26:33 +0100
Subject: cron with sched module
Message-ID: <91dgfa$q54$1@reader1.imaginet.fr>

I'm trying hardly to make a python "cron" like utility that may run
functions/methods periodically (say every x seconds) and predetermined
hours.

I can't use Window "AT" or Unix "cron" cause some functions/methods must be
executed evry 15 seconds. Launching Python executable periodically such a
manner will eat the disk I/O and

So, I'm trying with the "sched" module to do this but I can't add  "enter"
to a scheduler object already running.
So, what's the workaround ?
Use directly lower level "threading" module in lace of sched ?

TIA

Gilles Lenfant




From timr at probo.com  Thu Dec  7 02:31:10 2000
From: timr at probo.com (Tim Roberts)
Date: Wed, 06 Dec 2000 23:31:10 -0800
Subject: Freeze.py and Python 2.0?
Message-ID: <h2fu2tsqltgh0ujcj224jo62d2dpnu2hnp@4ax.com>

The Python doc says that the Freeze tool is present in the Tools/freeze
directory of the Python source distribution.  Does this mean I *MUST* build
from source in order to get Freeze?  Is there any path to Freeze available
to someone using a binary distribution?
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From paulp at ActiveState.com  Tue Dec 12 02:57:17 2000
From: paulp at ActiveState.com (Paul Prescod)
Date: Mon, 11 Dec 2000 23:57:17 -0800
Subject: [Q] Are Exceptions used that much in practice?
References: <3A35D546.A4731B5B@enteract.com>
Message-ID: <3A35DA5D.CD5FE73B@ActiveState.com>

Jerome Mrozak wrote:
> 
> I realize that Python allows me to catch exceptions.  But being a novice
> I'm not sure if they are used much in the 'real world'.

Yes. For instance if 

	* a file cannot be opened, 
	* a network socket goes down,
	* a user hits Ctrl-C, 
	* an XML file has a typo in it,
	* ...

For various reasons you might want to catch all of these circumstances.

 Paul Prescod



From jadestar at idiom.com  Thu Dec 28 06:06:52 2000
From: jadestar at idiom.com (James T. Dennis)
Date: 28 Dec 2000 11:06:52 GMT
Subject: ~Python ?
References: <3a4833f8.346848@news.t-online.de> <slrn94if5e.3bu.jepler@potty.housenet>
Message-ID: <92f6sc$1v7n$1@news.idiom.com>

Jeff Epler <jepler at inetnebr.com> wrote:

> On Tue, 26 Dec 2000 06:03:40 GMT, Gerson Kurz
>  <gerson.kurz at t-online.de> wrote:
>>I understand that ~ is probably handled by the shell, but if Python
>>doesn't deal with it, its hard to use Python in some kinds of scripts.
>>(I think Perl, Pythons ugly twin-brother, does)

> I'd have to say that string-substitution-based languages screw
> me up about ~2/3 of the time.  It must be something about the way my mind
> works.      ^^^^
>             hee hee, get it?


> Seriously, the os.path module includes goodies such as os.path.expanduser
> and os.path.expandvars.

> If you want shorthand for this when you do an 'open', try the following:
> 	import __builtin__, os
> 	def open(file, *args):
> 		file = os.path.expanduser(file)
> 		file = os.path.expandvars(file)
> 		return __builtin__.open(file, *args)

>>>> open("~/.bashrc").readlines()
> ['# .bashrc\012', '\012', '# User specific aliases and functions\012', '\012', '# Source global definitions\012', 'if [ -f /etc/bashrc ]; then\012', '\011. /etc/bashrc\012', 'fi\012']

> Jeff

   What!?!  You think there's something wrong with:

open(os.path.expanduser(os.path.expandvars("$HOME/.bashrc"))).readlines()




From loewis at informatik.hu-berlin.de  Fri Dec  8 15:34:15 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 08 Dec 2000 21:34:15 +0100
Subject: reading X resources from Tkinter
References: <90n3at$130f$1@chaos.obstruction.com>
Message-ID: <p6qg0jyk5yw.fsf@informatik.hu-berlin.de>

guy at obstruction.com (Guy Middleton) writes:

> It's not clear to me (maybe I am not reading carefully enough) if one can
> query the X resources directly, or if you can copy X resources into the
> option database.  Is this possible?

I believe that's not possible. The X resource database is an Xt
feature, and Tk does not use Xt (nor does it emulate access to the
resource database).

You can emulate most of that yourself, except that Tk does not expose
XGetWindowProperty which you'd need to get at RESOURCE_MANAGER property.

Regards,
Martin



From aahz at panix.com  Thu Dec 28 13:03:16 2000
From: aahz at panix.com (Aahz Maruch)
Date: 28 Dec 2000 10:03:16 -0800
Subject: uthreads and context-switching (was Re: Erlang)
References: <mailman.977984412.30211.python-list@python.org> <92f0il01a5u@news1.newsguy.com> <86wvckqw4h.fsf@localhost.my.domain>
Message-ID: <92fv94$lpb$1@panix2.panix.com>

In article <86wvckqw4h.fsf at localhost.my.domain>,
Brad Knotwell  <knotwell at ix.netcom.com> wrote:
>
>I'm now wondering if Stackless Python's uthreads are cooperatively or
>pre-emptively multi-tasked.

Both.  :-/  That is, I believe it's pre-emptively multi-tasked in the
sense that threads do not explicitly yield up control, but it's
cooperative in the sense that the context-switch can only occur between
byte-codes, so a blocking I/O call will prevent context-switching.

If you put the blocking I/O into Python threads (which are OS-level
threads and therefore definitely pre-emptive), that might be a good way
of handling uthreads.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

Ninth Virtual Anniversary: 3 days and counting


From db3l at fitlinxx.com  Fri Dec 29 14:11:41 2000
From: db3l at fitlinxx.com (David Bolen)
Date: 29 Dec 2000 14:11:41 -0500
Subject: How to import from within a module
References: <mailman.978059610.14528.python-list@python.org>
Message-ID: <u66k3jb4i.fsf@ctwd0143.fitlinxx.com>

Le Snelson <le_snelson at transtechinc.cc> writes:

> I am attempting to import a module from in a function within a sort of
> root module [init] - it contains only an initialization function.  The
> intent is that configuration files will provide data to derive the name
> of the module to be imported. All of the scripts provide the same
> function names but for different underlying devices.
> 
> I am unable to get __import__( xxx ) to actually add xxx to any visible
> dictionary. However, in the same file if I uncomment a line at the top
> (global scope) that reads:
> import xxx
> Sure enough the xxx is in the list from dir(init).

The "import xxx" statement both imports the module (or finds it in
sys.modules if it has already been loaded), but also acts as a
definition in the local namespace for that module using it's name.  So
"import xxx" both finds/loads module xxx, and then sets a label "xxx"
to reference that module.

When you use "__import__" it just returns a reference to the loaded
module, but doesn't change your namespace.  If you want to keep the
reference, you need to handle your own assignment.  So, typically,
using "__import__" to mimic "import" would be:

    xxx = __import__('xxx')

Note that in both cases, the label created is in the local namespace,
which might lead to some confusion if the import is done within a
function but you expect to use the module elsewhere.  While the module
itself will be loaded (and could be found in sys.modules), no
reference will exist outside of the namespace where the reference was
created.  For example, if in your init module you had:

    def somefunction():

	import xxx

	# Use xxx here
	xxx.otherfunction()


    # Call the function
    somefunction()

    # xxx is not a variable here - this gives a NameError on xxx
    xxx.otherfunction()


If you needed the module reference to be available in both places, you
could either import it normally in the global scope (which the
somefunction() reference would then find), or you could import it
later in the module just above when your global scope needed it.  The
later import would just reference the existing copy in sys.modules,
creating the reference to it.  (You could actually achieve the same
thing with "xxx = sys.modules['xxx']" but unless you were sure it had
previously been imported, you'd want to protect that sys.modules
reference against a KeyError).

If the issue is that somefunction() is dynamically figuring out what
module to load (thus you can't just import it in global module scope),
and then you want to insert it into the module level scope while still
within the function, that gets trickier.  It's probably cleaner to
have the function return a reference to the module which the caller
(from the module level scope) uses to assign to the variable.  There
have been other posts in this newsgroup covering accessing your own
module's dictionary to make such modifications, or explicitly
installing such references in the global scope, if that's really what
you want to do.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From linux at sinesoft.com  Sun Dec 10 09:52:27 2000
From: linux at sinesoft.com (Linux - Sinesoft)
Date: Sun, 10 Dec 2000 14:52:27 -0000
Subject: Python Problem - Could not find platform independent
Message-ID: <9105de$20l$1@venus.telepac.pt>

hi!

I had installed python in my linux box, but i had the following error:

bash-2.03# python
Could not find platform independent libraries <prefix>
Could not find platform dependent libraries <exec_prefix>
Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>]
'import exceptions' failed; use -v for traceback
Warning!  Falling back to string-based exceptions
'import site' failed; use -v for traceback
Python 1.5.2 (#1, Apr 28 2000, 06:45:56)  [GCC 2.95.2 19991024 (release)] on
linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>>


Can anyone please help me corrected this error message ?  Please answer me
by email.





From aleaxit at yahoo.com  Fri Dec 15 11:07:14 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 15 Dec 2000 17:07:14 +0100
Subject: __repr__ policy problem: should module be prefixed to output?
References: <3A3A3A6D.2C3B1A91@sbc.su.se>
Message-ID: <91dfl0014sa@news1.newsguy.com>

"Per Kraulis" <per at sbc.su.se> wrote in message
news:3A3A3A6D.2C3B1A91 at sbc.su.se...
> Having just released a Python extension module crng
> (http://www.sbc.su.se/~per/crng/), I have received some constructive
> criticism, one issue being save/restore functionality. My plan is to
> define a __repr__ for each type in the module.

That may give problems with state that's in floats, so I'd rather
have separate checkpointing save/restore mechanisms (returning a
tuple of picklable data).  Still, the repr issue IS interesting:

> However, I have realized that there is a policy problem. Should __repr__
> add the module name as prefix to the output string, or not? Or is there
> some other solution?

I have exactly the same problem in gmpy (I do supply a portable
binary picklable string, but also a repr output that is eval'uable
to restore something 'close to' the original).

> The problem is that Alt 1 is appropriate in the context of "import
> crng", while Alt 2 is better when having done "from crng import *".

If it was only the latter, it would not be worth supporting except
for those very rare modules that are _designed_ to be from-import-*'ed.

But a more reasonable usage, such as:

    from themodule import specificthing, anothergoodie, classname

would also favour repr strings of the form 'classname(avalue)' over
'themodule.classname(avalue)'.

Existing modules seem to do it both ways, so it's hard to
find specific precedents for guidance.

One thing favouring the 'undecorated' form is that an eval
_using the module's dictionary as global-dictionary_ should
then work anyway (as well as string-manipulations before
eval, which are of course possible in either case).


> A suggestion is that there should be a settable module variable that
> controls this. Feels like a kludge to me...

I originated that suggestion, and I agree it's a kludge (gmpy, in
its pre-alpha stage, is accumulating a bit too many 'module-level
settings' to try to keep everybody happy -- I'll have to do
something about it!-).

If nothing else, opinions about the best _default_ for such a
setting would be interesting to hear.


Alex





From scarblac at pino.selwerd.nl  Wed Dec 20 06:16:19 2000
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 20 Dec 2000 11:16:19 GMT
Subject: Need Help sorting alphabetically.
References: <3A3F1A91.A7ED9182@araidesign.com> <slrn93u8qc.5lh.scarblac@pino.selwerd.nl> <OlY%5.172268$U46.5446765@news1.sttls1.wa.home.com>
Message-ID: <slrn94152t.74g.scarblac@pino.selwerd.nl>

Greg Jorgensen <gregj at pobox.com> wrote in comp.lang.python:
> "Remco Gerlich" <scarblac at pino.selwerd.nl> wrote in message
> news:slrn93u8qc.5lh.scarblac at pino.selwerd.nl...
> > Chris Arai <chris at araidesign.com> wrote in comp.lang.python:
> 
> > > I would like to sort alphabetically (I'm not sure that is the correct
> > > term) so that strings of alpha numerics sort alphabetically instead of
> > > by ASCII order.
> [snip]
> > It seems you want to ignore case and ignore everything outside a-z.
> > So make a function that throws out all of the other characters and
> > turns the rest into a single case, and then compares those:
> >
> > def compare_alphabetically(a, b):
> >    a = filter(lambda x: x.isalpha(), a.upper())
> >    b = filter(lambda x: x.isalpha(), b.upper())
> >    return cmp(a,b)
> 
> The filter is unnecessary; the upper() and lower() methods of string only
> change alphabetic characters. A simpler solution is:
> 
> def compare_alpha(a, b):
>     return cmp(a.lower(), b.lower())

It's unclear what he wanted exactly, but I think he wanted to sort
purely alphabetically, ignoring the numbers and other characters.
For instance, 'Pg95A_100.gif' sorts before 'Pg100_100.gif'. With
my version it compares "PGAGIF" and "PGGIF" and it works.

But it was pretty unclear what he wanted exactly and why, so who knows :)

-- 
Remco Gerlich



From ingoogni at home.nl  Mon Dec 18 14:12:34 2000
From: ingoogni at home.nl (ingo)
Date: Mon, 18 Dec 2000 19:12:34 GMT
Subject: Numeric & map() ?
Message-ID: <Xns900EC6C24seed7@news1.zwoll1.ov.home.nl>

When using an array as a sequence in the map() an axis of the array got 
lost.

from Numeric import *
def Func(Data3d):
    print Data3d.shape

Arr=reshape(arange(27),(3,3,3))
print Arr.shape

a=map(Func,Arr)

result:
(3,3,3)
(3,3)
(3,3)
(3,3)

To me, being quite new with all this, it's somewhat unexpected. What is 
happening and how do I get the third axis back?

TIA,

Ingo


From aleaxit at yahoo.com  Wed Dec 13 04:52:25 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 13 Dec 2000 10:52:25 +0100
Subject: A question.
References: <9152a0$acb$1@panorama.wcss.wroc.pl> <915elr0681@news2.newsguy.com> <9166uv$8rs$1@animus.fel.iae.nl> <916cd20b3b@news1.newsguy.com> <jMAZ5.170$x3.2004@uchinews>
Message-ID: <917gta020ig@news2.newsguy.com>

"Ben Wolfson" <wolfson at midway.uchicago.edu> wrote in message
news:jMAZ5.170$x3.2004 at uchinews...
> In article <916cd20b3b at news1.newsguy.com>,
> Alex Martelli <aleaxit at yahoo.com> wrote:
>
> > [I don't wink, I don't adverbialize, I iconize
> >only very occasionally, the polysyllabic eruditeness of my hallmark
>                                           ^^^^^^^^^^^
> >circumlocutions pushes my Gunning Fog AND my Flesch-Kincaid up
> >in the high 20's -- what more could I do to shew hostility...?]
>
> *Surely* you mean "erudition"!

Abso-bloody-lutely not, o Ben.  Why use a relatively common
9-letter word when a perfectly correct 11-letter synonym of
it, that is used much more rarely, offers itself?  Hey,
'eruditeness' gets 68 Google hits in all (including those
in Merriam-Webster's Online Collegiate Thesaurus, American
Heritage, etc, which just serve to proclaim it as a fully
correct and legitimate synonym for 'erudition'), while your
preferred 'erudition' get 46.600 -- talk about hoi polloi!

Glad I landed a nicely-sized fish with this one.  I was
rather betting on 'shew', though -- doesn't it just _look_
like a typo?-)


Alex





From sill at localhost.kitenet.net  Fri Dec  8 17:40:30 2000
From: sill at localhost.kitenet.net (Oldayz)
Date: Fri, 08 Dec 2000 22:40:30 GMT
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu> <slrn92tu7t.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com> <3A2FE630.5C113AB@bioeng.ucsd.edu> <oOVX5.95$U4.1605@newsread1.prod.itd.earthlink.net> <MhWX5.2529$O%.394850@newsc.telia.net>
Message-ID: <yrdY5.50233$II2.4521078@newsread2.prod.itd.earthlink.net>

On Fri, 08 Dec 2000 00:53:00 GMT, Fredrik Lundh <fredrik at effbot.org> wrote:
>Oldayz wrote:
>> in other web searches you'd probably be even less lucky, while search on
>> 'history of python' came up with self-descriptive    [40]History of:
>> GreatSoftware - Python.
>
>http://wiki.cs.uiuc.edu/GreatSoftware/HISTORY/Python ?
>
>Did you follow that link?
>
>Did you notice the colon in the "self-descriptive" title?
>
>(It's a revision history for a page about Python, not a
>page about Python's history).

Oops.

However, after about 5 minutes of lurking on python.org,
I found this one: http://python.org/doc/essays/foreword.html.
It may be tricky to find, but as one post here said, the
original poster could have asked 'I've looked at python.org
and searched the web and I can't find anything! help!'.
In that case I'm sure nobody would be mean to him.

>
>> while a question like 'guys I'm writing a paper on python,
>> can you tell me where it comes from?' shows laziness
>
>Looks like you're pretty lazy yourself, but that didn't
>stop you from posting...

No, I showed that I jump to conclusions (which is pretty
bad, but nevertheless an honest mistake). I think that
explains the reaction of people in newsgroups and on
irc: when you see someone who lacks in knowledge,
you try to teach him a bit, if someone shows laziness,
most will have a hostile reaction. I've seen this alot
in linux irc channels. I personally prefer the neutral
'go read the docs' because swearing at people leads
to pointless flamewars and helping the lazy is like
trying to drink the ocean: they will keep coming back
again and again spending more of their own time than
it'd take to look it up themselves. 

I think we all do this to some extent at first (I
know I did, but after *some* research) but this is 
something you have to get over as soon as possible.

>
></F>
>
>


-- 

	Andrei


From ngps at madcap.dyndns.org  Wed Dec 13 12:16:33 2000
From: ngps at madcap.dyndns.org (Ng Pheng Siong)
Date: 13 Dec 2000 17:16:33 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <WntZ5.3566$Kd1.414873@newsb.telia.net>
Message-ID: <918ath$15f$1@violet.singnet.com.sg>

According to Fredrik Lundh <fredrik at effbot.org>:
> Realize that there are more than one way to do it.
                     ^^^
"There are more than one way..."

"There is more than one way..." aka TIMTOWTDI.

Hmmm...


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



From syver at NOSPAMcyberwatcher.com  Fri Dec  8 21:02:43 2000
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Sat, 9 Dec 2000 03:02:43 +0100
Subject: SSL how to?
References: <yZ3Y5.3016$9B.89007@news1.oke.nextra.no> <p6qu28eiqig.fsf@informatik.hu-berlin.de>
Message-ID: <90s40v$f6d$1@troll.powertech.no>

"Martin von Loewis" <loewis at informatik.hu-berlin.de> wrote in message
news:p6qu28eiqig.fsf at informatik.hu-berlin.de...

> Yes, httplib supports SSL in Python 2.

I am using Python 2 under Win 2000 Pro by the way.

If I understand it right you say that it's a part of the Be Open Source
distribution but not on Windows?




From rob at ZOOstation.cc  Fri Dec  8 03:13:31 2000
From: rob at ZOOstation.cc (Rob Brown-Bayliss)
Date: 07 Dec 2000 20:13:31 -1200
Subject: Tuple - List - Whats the difference?
In-Reply-To: <Pine.LNX.4.21.0012071349420.22360-100000@relax.dk>
Message-ID: <200012080813.eB88DWu01392@ZOOstation.cc>

Hi, brand new to python and I am trying to understand why there are
tuples and lists.  

They both look to serve the same purpose.  Can some one enlighten me?

Thanks  


-- 

  Rob Brown-Bayliss
 ---======o======---
  www.ZOOstation.cc



From chicago_george at my-deja.com  Mon Dec  4 14:30:58 2000
From: chicago_george at my-deja.com (chicago_george at my-deja.com)
Date: Mon, 04 Dec 2000 19:30:58 GMT
Subject: JPython - is it what I need to...
Message-ID: <90grdi$err$1@nnrp1.deja.com>

Hello - Here's what I need to do.. I thought JPython would be good for
this, but wanted to get the opinions of others first..  ANY feedback
would be really appreciated;

Using a high level scripting interface, I need to subclass a Java
class.  We hope we can make this script easy enough so the people in
our support department can modify the script without needing the help
of the developers.  This script should define the details of how data
from a flat file will be inserted into a database.

Would I be better off using a different tool?
Opinions/Questions/Comments..

Thanks!
George.
NewToDevelopment at hotmail.com


Sent via Deja.com http://www.deja.com/
Before you buy.


From root at rainerdeyke.com  Sat Dec 30 00:23:26 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Sat, 30 Dec 2000 05:23:26 GMT
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <3A4D6A05.EEC97FA6@uswest.net> <slrn94qqvg.mgg.grey@teleute.rpglink.com>
Message-ID: <ije36.96813$x6.43604873@news2.rdc2.tx.home.com>

"Steve Lamb" <grey at despair.rpglink.com> wrote in message
news:slrn94qqvg.mgg.grey at teleute.rpglink.com...
> On Fri, 29 Dec 2000 21:52:21 -0700, Bjorn Pettersen <pbjorn at uswest.net>
wrote:
> >henry_crun3583 at my-deja.com wrote:
> >> do :
> >>     R=Themistor.ReadResistance()
> >>     T=R_to_Temperature(R)
> >>   until (T >=25)
>
> >while 1:
> >   R=Themistor.ReadResistance()
> >   T=R_to_Temperature(R)
> >   if T >=25: break
>
>     Or, for those who don't like while 1:
>
> while T < 25:
>   R=Thermistor.ReadResistance()
>   T=R_to_Temerature()

T doesn't necessarily exist the first time 'T < 25' is evaluated above.
You'd need something like this:

T = 25
while T < 25:
  ...

At that point it's better to use 'while 1', which is at least idiomatic and
explicit in its inelegance.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From dnew at san.rr.com  Wed Dec 13 18:48:29 2000
From: dnew at san.rr.com (Darren New)
Date: Wed, 13 Dec 2000 23:48:29 GMT
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <29lZ5.138780$U46.4638547@news1.sttls1.wa.home.com> <uhf49kf8r.fsf@roundpoint.com> <qXFZ5.142740$U46.4719533@news1.sttls1.wa.home.com> <3A37B27A.B18E8497@uswest.net>
Message-ID: <3A380ACE.12BAFE1F@san.rr.com>

Bjorn Pettersen wrote:
> In these cases the compiler doesn't need a temporary variable. If you look at
> the (unoptimized) output for your favorite compiler you'll see that they're
> identical [I only tested it on msvc, but if it can do it, I'm assuming other
> compilers can also <wink> ]

One of the very popular C compilers for CP/M used to not be smart enough to
optimize a post-increment with no side-effects. (At least, not be default.
It's been a while.)  So while this may have been a valid concern back in the
days of 64K address spaces, I would be surprised if any compilers written in
the last 5 years were thus limited.

-- 
Darren New / Senior MTS & Free Radical / Invisible Worlds Inc.
San Diego, CA, USA (PST).  Cryptokeys on demand.
Personal malapropism generator free with purchase!
Steganography: The manual is hidden in the source code.
Gold mining is only profitable because of the gold found.


From ben.hutchings at roundpoint.com  Fri Dec  8 06:34:53 2000
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 08 Dec 2000 11:34:53 +0000
Subject: Music synthesizer in Python/C
References: <G588B3.6A5@world.std.com> <G588o8.DC7@world.std.com>
Message-ID: <uu28fjgde.fsf@roundpoint.com>

wware at world.std.com (Will Ware) writes:

> Will Ware (wware at world.std.com) wrote:
> > http://www.deja.com/getdoc.xp?AN=702341030&fmt=text
> 
> I posted the resulting tune as a uuencoded MP3 on
> alt.binaries.sounds.music.amateur, it hasn't appeared on deja.com
> yet
<snip>

Deja doesn't carry binaries.

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From pinard at iro.umontreal.ca  Wed Dec 13 13:54:56 2000
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=e7ois?= Pinard)
Date: 13 Dec 2000 13:54:56 -0500
Subject: How to REPL?
Message-ID: <oqk894ywvz.fsf@lin2.sram.qc.ca>

Hi, people.

I vaguely remember this has been discussed, a long while ago, but do not
handily find the discussion in my archives.

Would someone happen to have a modest recipe for a read-eval-print-loop,
that could be used to trigger interactive Python from within an otherwise
batch script?  I would like read-write access to the Python environment,
if possible.  Thanks for your reply, in advance.

If it matters, on _this_ machine, I'm using 1.5.2.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard



From israelt at optushome.com.au  Wed Dec 13 05:25:13 2000
From: israelt at optushome.com.au (israel raj thomas)
Date: Wed, 13 Dec 2000 10:25:13 GMT
Subject: Ruby, Python, Eiffel and Smalltalk
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de>
Message-ID: <mdne3t8q8670mbplnle1jrettaqnf8bssq@4ax.com>

>Do you think that it would be possible to use Python as a Smalltalk
>replacement if an IDE existed that hid some of Python's limitations?

Would it perhaps not be better to use Ruby ?
See below for information and for comparisons of Ruby, Python, Eiffel
and Smalltalk:

http://www.bagley.org/~doug/shootout/
http://www.pragmaticprogrammer.com/ruby/articles/index.html
http://www.pragmaticprogrammer.com/talks/introruby/introruby.htm
http://www.ddj.com/articles/2001/0101/0101b/0101b.htm
http://www.s-direktnet.de/homepages/neumann/rb/download_ruby.html
http://www.s-direktnet.de/homepages/neumann/ruby_en.html
http://www.s-direktnet.de/homepages/neumann/lang_cmp.en.htm
http://www-4.ibm.com/software/developer/library/ruby.html
http://www.sunworld.com/swol-02-1999/swol-02-regex.html



From howardbgolden at earthlink.net  Tue Dec 12 14:02:05 2000
From: howardbgolden at earthlink.net (Howard B. Golden)
Date: Tue, 12 Dec 2000 19:02:05 GMT
Subject: [ANNOUNCE] PyHTML 0.5 -- Indentation-structured HTML generator
References: <slrn92v6h6.6vl.jepler@potty.housenet> <slrn939lnr.6bm.jepler@potty.housenet>
Message-ID: <NCuZ5.15809$U4.1462831@newsread1.prod.itd.earthlink.net>

"Jeff Epler" <jepler at inetnebr.com> wrote:

> PyHTML is an extension of the Python grammar, plus a modified compiler
> which permits the use of HTML-style tags as Python blocks, and which makes
> these tags and expression-statements write their values as output, making
> it easy to write HTML document generators with the block-indentation
> style of Python.

This is somewhat off-topic, but I have been thinking of doing what you are
doing, but "inside-out."  I would like to make an XML version of Python's
syntax, similar to Greg Badros's JavaML (see
http://www.cs.washington.edu/homes/gjb/JavaML/index.html).  If this
"alternative" Python syntax were available, then it seems to me it would
lend itself even more to embedding XHTML inside it.

I'm planning to write up a PEP on this, hopefully by New Year's Day.
Comments welcome.

Howard




From mats at laplaza.org  Tue Dec 19 08:52:24 2000
From: mats at laplaza.org (Mats Wichmann)
Date: Tue, 19 Dec 2000 13:52:24 GMT
Subject: Programming Habits in Python
References: <cspU5.82459$U46.2617014@news1.sttls1.wa.home.com> <8vvb6p$ct2$1@news.wrc.xerox.com> <XiMU5.87123$U46.2803855@news1.sttls1.wa.home.com> <3aQU5.718476$8u4.11640842@news1.rdc1.bc.home.com> <901elo$3kv$1@news.chorus.net>
Message-ID: <3a329887.15154861@news.laplaza.org>

>This really is the crux of the problem. In most large companies the emphasis
>is on getting the work done NOW, not necessarily getting it done in the best
>possible way.


Brings back unpleasant memories.  The only employment situation I've
been in that I really didn't enjoy was working for a medium-sized
outfit recently acquired by one of the larger ones.  The best
description of what we did was system integration. We really needed
reusable components to plug into the various related projects, and to
a limited extent the things we had were reusable - barely.  But
basically a small amount of time developing a really solid framework
would have gone a LONG way towards delivering much better projects, on
time.  This could not be done because there was nobody to pay for it.
Projects were to be paid for by the client buying the project. There
was no budget item to pay for in-house frameworks or tools. Of course
a client couldn't be expected to pay for someone /else's/ project.....

I lasted nine months in that place, and looking back I'm surprised it
was that long.





From jsboyle at my-deja.com  Wed Dec 27 13:02:15 2000
From: jsboyle at my-deja.com (jsboyle at my-deja.com)
Date: Wed, 27 Dec 2000 18:02:15 GMT
Subject: Season's greetings (was Re: Some Python 2.1 ideas)
References: <mailman.977765348.13376.python-list@python.org> <928otj02l39@news1.newsguy.com>
Message-ID: <92dar2$juv$1@nnrp1.deja.com>

In article <928otj02l39 at news1.newsguy.com>,
  "Alex Martelli" <aleaxit at yahoo.com> wrote:

I also have a teenage son,. His experience is java, C, C++. I found
that the python project with the biggest zowie! factor was V-python,
formerly visual python:
http://sourceforge.net/projects/visualpython

This unfortunately requires a special python, for reasons I find
specious,  but allows some really neat 3D graphics (with
animation) using GL.
It runs under windows, Linux, and Mac. It is used at
Carnegie-Mellon in a physics course.

Jim
>
> Me, I had the best Christmas ever since I got my first electric
> train... got to show Python to my son (Pascal and Assembler
> his only previous programming experiences), and again and
> again he got the jaw-drops-"but-it-CAN'T-be-THAT-powerful!"
> effect which I well recall from my early Python days.  (And as
> a cherry on top he kibitzed me for an OKbridge session where
> I got close to 70%, but the Python part alone would have, by
> itself, been quite enough...!-).
>
> One thing I wasn't able to easily answer -- what's the
> *easiest* way for a Python beginner to do function-plots,
> with nicely labeled graphs, minor/major ticks, color,
> auto-scaled, &c &c?  On-screen display the major need,
> though nice printouts wouldn't hurt either.  Speed no
> issue, as his new PC is twice as fast as mine!-)
>
> I tried downloading DISLIN but it appears to embed its
> own Python, need its own setting of PYTHONPATH; and
> (from a superficial look at it) it doesn't seem to do auto
> scaling.  Any suggestions?  I'd like to keep amazing him
> with Python power and ease, as what dad wouldn't like
> for his late-teenage son...:-).
>
> Alex
>
>


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


From tim.one at home.com  Mon Dec 18 13:15:39 2000
From: tim.one at home.com (Tim Peters)
Date: Mon, 18 Dec 2000 13:15:39 -0500
Subject: while true: !!!
In-Reply-To: <20001218085229.A10917@westsidecnc.rh.rit.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEGKIEAA.tim.one@home.com>

[D-Man]
> I think that some people were confused with this statement.  He meant
> that 'true' and 'false' would be constants like in C++ and Java.  It
> would be illegal to try
>
> true = false
> ...

A non-empty string literal is true, and unalterably so, so everyone can be
happy writing:

while "forever, because stupid Guido won't define 'true'":
    wink




From wilrcool at hotmail.com  Fri Dec 22 18:55:07 2000
From: wilrcool at hotmail.com (Will R.)
Date: Fri, 22 Dec 2000 17:55:07 -0600
Subject: Courses
Message-ID: <E03875E946B542E3.0DB39C6CA5B6CBEF.061CFF9824AAA9D9@lp.airnews.net>

Does anyone know where I can take courses to learn Python in or around the
Houston/Spring TX area?

~Wil~




From grey at despair.rpglink.com  Tue Dec 26 11:03:01 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Tue, 26 Dec 2000 16:03:01 -0000
Subject: Python and Napster
References: <sx226.18970$RC1.511898@newsread1.prod.itd.earthlink.net>
Message-ID: <slrn94hg9l.5na.grey@teleute.rpglink.com>

On Tue, 26 Dec 2000 15:10:16 GMT, Doug Farrell <dfarrell01 at earthlink.net>
wrote:
>Is there a way to build a 'Napster' like client in Python? What I'm really

    Yes, Python can do networking.  The implementaion of the protocol, as with
most things, is up to you.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From Chris.Schaller at web.de  Fri Dec 22 15:56:04 2000
From: Chris.Schaller at web.de (Chris Schaller)
Date: 22 Dec 2000 20:56:04 GMT
Subject: namespaces & indentation
Message-ID: <920f54$m2f$1@news.mch.sbs.de>

Dear Pythoners...

  I always thought that I understood Python's scoping, but then I came across
that code:
> def t(x):
>   if x == 1:
>     c = 'yeah'
>   return c

If I execute t(1), I'll get 'yeah', but otherwise there'll be an error
message.  According to the indentation level c is only valid within the
if-clause, but not on the outside.  So I guess, the indentation level only
defines the lines of code that belong to the if-clause, but has nothing to do
with scopes.

  That's very confusing, I've already programmed a lot in Python, but only
due to a program error (I forgot to initialize c at the beginning of the
function - a common error) and an unsuccessful test this happened to a program
after months of error-free running.

  Any comments?

bye
  Chris...


From paul.moore at uk.origin-it.com  Thu Dec 21 11:40:37 2000
From: paul.moore at uk.origin-it.com (Paul Moore)
Date: Thu, 21 Dec 2000 17:40:37 +0100
Subject: Events in win32com
References: <Y1g=Op6IfWSoJn1ISFtOyuvqwBb3@4ax.com> <FY0=OlZFLy71DtnK62u5TEPlhOts@4ax.com> <3A40060B.10504@ActiveState.com> <cYxAOmxi9FgNERgBo29zuCl5c2h1@4ax.com> <ik724tc2v3cpcqfbt7i1ookvaona0mk5ql@4ax.com> <3A418D02.2040302@ActiveState.com>
Message-ID: <KzJCOorBIUxbuv3JYT4D29YCxCb2@4ax.com>

On Thu, 21 Dec 2000 05:59:32 +0100, Mark Hammond
<MarkH at ActiveState.com> wrote:
>Paul Moore wrote:
>> Can I rewrap it somehow, say as
>>   doc = DispatchWithEvents(ie.document, DocEvents)
>
>This is (should be) as simple as instantiating a sub-class of the makepy 
>generated event class, and passing the document.
>
>eg, something like:
>
>mod = EnsureModule(...)
>
>class MyEvents(mod.IDocumentEvents):
>   # your methods here....
>
>handler = MyEvents(ie.document)
>
># handler should start recieving events.

Wonderful! I'll try this later. I am continually impressed by how
powerful all of this stuff is. It's just a bit daunting knowing how to
find these things out... (But with the prompt and helpful responses
from you, that's not an issue, other than the fact that I start to
become a nuisance :-)

Happy Christmas,
Paul.



From not.this at seebelow.org  Fri Dec  8 16:41:20 2000
From: not.this at seebelow.org (Grant Griffin)
Date: Fri, 08 Dec 2000 21:41:20 +0000
Subject: Python: what all the hype is about?
References: <90locf$d80$1@nnrp1.deja.com> <9uus2tkehg7pr8gtvka4jh6q68kpvgfs3o@4ax.com>
Message-ID: <3A315580.A411C413@seebelow.org>

Dale Strickland-Clark wrote:
...
> Your choice of language is dependent on a number of factors - in no particular order:
> 
> 1. Suitability for the job
> 2. Knowledge of the language (the language you know is usually easier to code than one you don't)
> 3. Availability of tools (debuggers, source sharing, documenters)
> 4. Availability of support.
> 
> Python frequently scores well for 1 in Web/CGI, databases, middleware, file utilities, data
> processing and analysis but not so well for GUI, 3D games.
> 
> For item 3, Python is, at best, average.
> For 4, this group is a pretty reliable source of excellent information and help.
> 
> Someone will come along shortly and tell you the opposite, I'm sure.

No they won't.

the-argument-clinic-ly y'rs,

=g2
-- 
_____________________________________________________________________

Grant R. Griffin                                       g2 at dspguru.com
Publisher of dspGuru                           http://www.dspguru.com
Iowegian International Corporation	      http://www.iowegian.com


From ckrill at qvlinc.com  Fri Dec  8 14:24:32 2000
From: ckrill at qvlinc.com (Coy Krill)
Date: Fri, 08 Dec 2000 11:24:32 -0800
Subject: Pickles Cross-platform?
Message-ID: <3A313570.A71966A@qvlinc.com>

I seem to have to run 'dos2unix' past any pickles that I want to read
that originated on our Windows machines. Is this typical?  For some
reason I was under the impression that the unpickler handled the LF vs
CRLF contortions for you but it doesn't seem to be doing that.

Coy




From s.schwarzer at ndh.net  Mon Dec 11 20:02:21 2000
From: s.schwarzer at ndh.net (Stefan Schwarzer)
Date: Tue, 12 Dec 2000 02:02:21 +0100
Subject: Function points per lines of code
Message-ID: <3A35791C.5930A7FA@ndh.net>

Hello,

for an effort/schedule estimation I look for values of lines of code per
function point for Python. I have searched Deja.com and Google.com but to
no avail.

Can anyone help me? Thanks in advance.

Stefan


From sholden at holdenweb.com  Mon Dec 11 09:39:04 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 11 Dec 2000 09:39:04 -0500
Subject: easy question on input files in Python
References: <908jvp$dd1$1@nnrp1.deja.com> <908mmu02npb@news2.newsguy.com> <ewHRZOAK8CK6EwgW@ntlworld.com> <90c7jn$3bm$1@nnrp1.deja.com> <90d05f01tub@news1.newsguy.com> <90ite3$2j0$1@nnrp1.deja.com>
Message-ID: <0J5Z5.4082$Q54.142458@e420r-atl1.usenetserver.com>

<jerry_spicklemire at my-deja.com> wrote in message
news:90ite3$2j0$1 at nnrp1.deja.com...
> Alex Martelli  wrote:
>
> > What Ian wants is a 'overwritable buffer', so the _same_
> > memory area is reused at each read.  But Python file
> > objects don't support that.  Nor I believe do array
> > objects (their fromFile method appends to the array,
> > does not overwrite it).  I think it need to be done as a
> > C extension (maybe Numeric already offers it...?).
>
> Good point! Object arrays, not limited to numeric objects,
> are one of the coolest features of Numeric.
> I've no clue about over-writable-ness though.
>
> Thanks for the reply!
> Jerry S.
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
>
The augmented assignment operations can change objects in place, but of
course they are written in C as a part of the implementation :-(

Looks like this problem might need an extension module in C?

regards
 Steve






From scarblac at pino.selwerd.nl  Wed Dec 13 12:30:56 2000
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 13 Dec 2000 17:30:56 GMT
Subject: Moving from Perl to Python
References: <3a3a9552.76673185@news.tcn.net>
Message-ID: <slrn93fcdt.a8m.scarblac@pino.selwerd.nl>

Richard Zilavec <rzilavec at tcn.net> wrote in comp.lang.python: 
> I just got back from two book stores, nothing on Python, I was
> shocked!  So I went online and found a pile but I hate ordering
> online, I like to browse through them first.  Anyways, I really like
> O'Reilly books, is the Programming Python really advanced or is more
> along the lines of the Programming Perl where it covers the basics to
> advanced?

Personally I don't like Programming Python much. It's becoming old (there
are rumors of a new version soon) and not very well structured. But then,
it's a lot of book, there are many examples, and the important things are
probably all in there somewhere.

I started with Learning Python, and I like that book. It explains the
basics of the language well. But it doesn't go into loads of examples
on case studies or so, you will understand the language but you won't be
immediately familiar with the whole library. Maybe you can study the only
tutorial well and skip this.

Python Essential Reference is a nice reference book. It has a short
tutorial, then a good reference on the language and of the standard
library. One of the best books out there at the moment.

More books at http://www.python.org/psa/bookstore/

There is good online documentation at http://www.python.org/doc/current/ .

There are things out there on the web that explain Python to Perl programmers,
but I can't find them at the moment. 

There are some posts by Tom Christiansen to this group that you can find
on Deja about Perl v Python, those may be useful.

-- 
Remco Gerlich

> 
> man python on my machine doesn't work, I'm assuming something is wrong
> or maybe its a perldoc equivalent?  Where can I find the FAQs?
> 
> Thanks in advanced for any help or suggestions.
> 
> --
>  Richard Zilavec
>  rzilavec at tcn.net


From aleaxit at yahoo.com  Thu Dec 28 05:19:54 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 28 Dec 2000 11:19:54 +0100
Subject: classes (was Re: Same again please for OOP)
References: <mailman.977623149.8184.python-list@python.org> <9257mf0sl1@news1.newsguy.com> <roy-E5B925.13324224122000@news.panix.com> <925khi019hs@news1.newsguy.com> <kubsu1t81x.fsf@lasipalatsi.fi> <92708d0g4g@news1.newsguy.com> <mailman.977958972.13680.python-list@python.org>
Message-ID: <92f46d01d7n@news1.newsguy.com>

"D-Man" <dsh8290 at rit.edu> wrote in message
news:mailman.977958972.13680.python-list at python.org...
>
> (Sorry if this posts twice, but I don't see my earlier post)

I did, and I'm just trying to make sure you see the answer (which
is why I'm also emailing this!) since it's my fault...


> However, I can't get this sample code to work.  I keep getting
> AttributeError's or infinitely recursive __getattr__ function.  If I
> remove one of the leading underscores it works fine.  This is because
> of the current name mangling used to "enforce" private members.
>
> Am I missing something here, or should I only use 1 leading
> underscore?

My bad -- ONE leading underscore is right.  Sorry!


Alex





From matt at mondoinfo.com  Sat Dec  9 14:12:57 2000
From: matt at mondoinfo.com (Matthew Dixon Cowles)
Date: Sat, 09 Dec 2000 19:12:57 GMT
Subject: Dumb Newbie Question
References: <3A327D01.9841EF98@drew.edu>
Message-ID: <slrn93511o.t4.matt@happy-hour.mondoinfo.com>

On Sat, 09 Dec 2000 13:42:09 -0500, Scott Weisel <sweisel at drew.edu>
wrote:

>I thought I might be able to modify sys.path like this
>sys.path = sys.path + "c:\\python16\\programs 
>but I'm getting a TypeError stating "illegal argument type for
>built-in operation."

Scott,
sys.path is a list and the + operator won't append a string to a
list. You want either:

sys.path=sys.path+["c:\\python16\\programs"]

since + will concatenate two lists or else

sys.path.append("c:\\python16\\programs")

since a list's append method will take a string.

But I think it's more common to write:

sys.path.insert(0,"c:\\python16\\programs")

to put your own directory up front so that in case of a name conflict, it's
your own stuff that's picked up.

Regards,
Matt


From c.evans at clear.net.nz  Sat Dec  9 17:36:07 2000
From: c.evans at clear.net.nz (Carey Evans)
Date: 10 Dec 2000 11:36:07 +1300
Subject: Dijkstra's Shortest Path algorithm
References: <Pine.GSO.4.21.0012090042370.15110-100000@y.glue.umd.edu>
	<90t78p$5r1$03$1@news.t-online.com>
Message-ID: <87elzh8boo.fsf@psyche.dnsalias.org>

"J?rgen Hermann" <jh at web.de> writes:

> Maybe you should try to implement it in Python and NOT take 2-3 days. ;))
> 
> Tongue-in-cheek-ingly yours, J?rgen

Well...  I implemented it in Python last night and took 2-3 hours.  ;)

There's not much to it, so I'll post it here.  I'm putting this code
in the public domain, so you can do whatever you like with it.

----------------------------------------------------------------------
#!/usr/bin/python

# Andrew Snare's PQueue: http://www.pigpond.com/~earthpig/PQueue-0.1a.tar.bz2
import pqueue

# From Kingston, J. H., _Algorithms and Data Structures_, Addison Wesley, 1991.
data = { 'a': [('b', 13), ('c', 4), ('d', 7)],
         'b': [('a', 13), ('e', 1)],
         'c': [('a', 4), ('e', 7)],
         'd': [('a', 7), ('e', 2)],
         'e': [('b', 1), ('c', 7), ('d', 2)] }

# All shortest paths by Dijkstra's algorithm.
def all_shortest_paths(graph, start):
    # Previous node on current shortest path to start node.
    parents = {}

    # Current distance to start node.
    distances = {start: 0}

    # Which nodes we've seen - starts out with just the start node.
    visited = {}
    for v in graph.keys():
        visited[v] = 0
    visited[start] = 1

    # Set up priority queue with just the start node in it.
    pq = pqueue.PQueue()
    pq.insert(0, start)

    # Loop over closest nodes in queue until none left.
    while len(pq) != 0:
        vdist, v = pq.pop()

        # Loop over neighbouring nodes.
        for end, cost in graph[v]:
            # Work out the distance from the start, via this node's
            # shortest path, to the neighbouring node.
            newdist = vdist + cost
            if not visited[end]:
                # If we haven't seen the neighbour yet, queue it.
                pq.insert(newdist, end)
                parents[end] = v
                distances[end] = newdist
                visited[end] = 1
            elif newdist < distances[end]:
                # If we've seen it, and it's closer, decrease its distance.
                pq[end] = newdist
                parents[end] = v
                distances[end] = newdist

    return parents

# The shortest path from one node to another.
def shortest_path(graph, start, end):
    paths = all_shortest_paths(graph, start)
    nodes = [end]
    next = end
    while next != start:
        next = paths[next]
        nodes.append(next)
    nodes.reverse()
    return nodes

if __name__ == '__main__':
    # How to get from A to B.
    print shortest_path(data, 'a', 'b')
----------------------------------------------------------------------

-- 
	 Carey Evans  http://home.clear.net.nz/pages/c.evans/


From aleaxit at yahoo.com  Thu Dec 21 17:34:24 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 21 Dec 2000 23:34:24 +0100
Subject: A Suggestion for Python Colon Syntax
References: <91tc8h$2pp$1@hecate.umd.edu> <W2r06.1076$AH6.86249@newsc.telia.net> <91tfg1$282$1@hecate.umd.edu> <EEr06.842$Qb7.73987@newsb.telia.net> <91trir$52l$1@hecate.umd.edu>
Message-ID: <91u0k508r5@news2.newsguy.com>

"William Djaja Tjokroaminata" <billtj at z.glue.umd.edu> wrote in message
news:91trir$52l$1 at hecate.umd.edu...
    [snip]
> Therefore, my suggestion is really not to remove the colons, but to make
> them optional.  I want opinion of you who have programmed for a long time
> whether there will be some harms if the colons are made optional.  If not,

Speaking as somebody who has been programming for a quarter of a century,
and who's fallen wildly in love with Python since about 18 months ago (and
_deeply_ appreciates by far most of its features, syntactical, semantical,
and
pragmatical), I don't think there would be any harm in making the colons
optional.  Although they've become second nature for me after most of the
block-starting statements, I keep forgetting to put one after the mandatory
closing parens of each def-statement.  In a language so spare of useless
punctuation as Python most often is, there's something "against the grain"
in having to write as many as _three_ punctuation marks at the end of
a function definition, as in:

    def somefun():
        ...

Nothing major, mind you -- the compiler rapidly lets me know each and
every time I forget!-)


> probably someone can tell me how I can forward this suggestion to the core
> Python development team?

I _think_ feature requests can be placed on python.sourceforge.net,
although there may not be a specific place for them that's very
apparent -- but if you file a 'bug' I believe you can characterize it
as a feature-request.


Alex





From fredrik at effbot.org  Mon Dec  4 04:44:53 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Mon, 04 Dec 2000 09:44:53 GMT
Subject: hash() algorithm?
References: <nospam-36B66D.13423003122000@news> <ikzW5.1434$Kd1.226017@newsb.telia.net> <nospam-79EF46.15110903122000@news> <90fitt$q3h$1@news.udel.edu> <87bsus8sx1.fsf@psyche.dnsalias.org> <nospam-634BA6.02345204122000@news>
Message-ID: <pIJW5.1467$Kd1.231480@newsb.telia.net>

anand wrote:> Although I'm having trouble finding that file....
>
> Is the path diffent for the mac distribution?  Because I don't have a
> Objects directory, and stringobject.c doesn't exist anywhere.

no, it's not different -- stringobject.c is part of the inter-
preter's SOURCE code.  if you're looking at a binary release,
you won't find it.

source code is available here:

    http://www.python.org

and here's the relevant portion of the string hash
algorithm:

static long
string_hash(PyStringObject *a)
{
 register int len;
 register unsigned char *p;
 register long x;

 len = a->ob_size;
 p = (unsigned char *) a->ob_sval;
 x = *p << 7;
 while (--len >= 0)
  x = (1000003*x) ^ *p++;
 x ^= a->ob_size;
 if (x == -1)
  x = -2;
 return x;
}

as I mentioned before, other data types use different
hash algorithms.  check the sources for details.

</F>




From tcmay at got.net  Thu Dec 21 14:20:03 2000
From: tcmay at got.net (Tim May)
Date: Thu, 21 Dec 2000 11:20:03 -0800
Subject: teekay
References: <x%D%5.106612$IP1.3577747@news1.giganews.com> <F5N%5.108769$IP1.3655247@news1.giganews.com>
Message-ID: <tcmay-102827.11200321122000@news.got.net>

In article <F5N%5.108769$IP1.3655247 at news1.giganews.com>, 
nickb at fnord.io.com (Nick Bensema) wrote:

> In article <x%D%5.106612$IP1.3577747 at news1.giganews.com>,
> Nick Bensema <nickb at fnord.io.com> wrote:
> >Tk would make a great baby name.
> 
> It just occurred to me that if your baby was named Tk and you wanted
> to make him/her giggle, you could Tcl/Tk.


And if he's a bad baby, feed him to the Python.


--Tim May

-- 
Timothy C. May         tcmay at got.net        Corralitos, California
Political: Co-founder Cypherpunks/crypto anarchy/Cyphernomicon
Technical: physics/soft errors/Smalltalk/Squeak/agents/games/Go
Personal: b.1951/UCSB/Intel '74-'86/retired/investor/motorcycles/guns




From moshez at zadka.site.co.il  Sat Dec 23 23:08:45 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Sun, 24 Dec 2000 06:08:45 +0200 (IST)
Subject: Some Python 2.1 ideas
In-Reply-To: <005f01c06d01$d8114820$74eb0b18@C322162A>
References: <005f01c06d01$d8114820$74eb0b18@C322162A>
Message-ID: <20001224040845.5FE49A84F@darjeeling.zadka.site.co.il>

On Sat, 23 Dec 2000 09:00:25 -0800, "Bob Alexander" <bobalex at home.com> wrote:

> New operator: embedded assignment

This has already been discussed to death, many times. A PEP would be
useful, if only as a way to put a stop to the discussion <wink>

>     for i in 10:                # equivalent to "for i in xrange(10)"
>     for i in 5:10:             # equivalent to "for i in xrange(5, 10)"
>     for i in 20:10:-1:       # equivalent to "for i in xrange(20, 10, -1)"

This (actually the similar [i:j:k] notation) has been thought over for
2.0 and rejected. You can read the python-dev archives for the discussion.

> Since readline() and its relatives return strings with '\n' attached
> (sometimes), it would be convenient to have a built-in string method that
> removes one if it exists (seems like I'm often coding little functions to do
> this).
> 
>     <string>.chomp()    # sorry, borrowed the method name from Perl

Yep, this actually sounds useful.

-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From Daniel.Kinnaer at AdValvas.be  Mon Dec 18 12:06:11 2000
From: Daniel.Kinnaer at AdValvas.be (Daniel)
Date: Mon, 18 Dec 2000 17:06:11 GMT
Subject: Timers in Python?
Message-ID: <3a3e42ef.16144323@news.skynet.be>

Is there an equivalent of the Delphi Timer-component in Python as
well?  That is, I want to have two Timers : Timer1 needs to activate
itself every 2 minutes and Timer2 needs to activate itself every 7
minutes. 

In Delphi I can have 2 Timer.events  :

procedure TForm1.Timer1Timer(Sender: TObject);
begin
  //do a special task1
end;

procedure TForm1.Timer2Timer(Sender: TObject);
begin
  //do a special task2
end;

How is this done in Python?

Thanks for helping out.  Daniel


From glenfant at equod.com.nospam  Tue Dec  5 05:53:54 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Tue, 5 Dec 2000 11:53:54 +0100
Subject: SAX2 XML validating?
References: <3A2BF45A.246B757F@marge.hologic.com>
Message-ID: <90ih81$hj9$1@reader1.imaginet.fr>

Hi,

AFAIK, the only one XML validating parser for Python (100% python) is Lars
Marius Garshol's xmlproc
http://www.stud.ifi.uio.no/~lmariusg/download/python/xml/
If you're with Windows, you can use the MSXML3 parser as a COM object (needs
Win32com extensions from Mark Hammon).
Browse from http://msdn.microsoft.com/downloads/default.asp for MSXML3
Browse from http://www.activestate.com for Win32 extensions.

"Jeremy Scott" <jscott at marge.hologic.com> a ?crit dans le message news:
3A2BF45A.246B757F at marge.hologic.com...
> Does anybody know how to validate an XML file using the SAX2 on
> python2.0?
>



From jp at ulgo.koti.com.pl  Tue Dec 26 10:10:40 2000
From: jp at ulgo.koti.com.pl (Jacek =?iso-8859-2?Q?Pop=B3awski?=)
Date: 26 Dec 2000 15:10:40 GMT
Subject: ready for C or C++?
References: <20001225031852.04296.00003766@ng-fp1.aol.com>
Message-ID: <slrn94hd7g.1br.jp@localhost.localdomain>

Xmnemonic wrote:
>If I've learned Python and become proficient with it, am I ready to learn C or
>C++?

I know languages like C (even x86 asm), and I am learing Python now :) I think
programming is just programming, is some languages you must learn more,
or write more, and sometime you learn new technics (like structural/object),
but if you are programmer you can learn any language

-- 
No one sees an angel till it smashes to the ground
And then you run somewhere 
And leave it lying there
                                                   "Magica" - Ronnie James Dio


From emile at fenx.com  Tue Dec 19 10:33:53 2000
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 19 Dec 2000 07:33:53 -0800
Subject: tkinter event binding to mouse double click
References: <91n7l6$7ij$1@news.nuri.net> <91njgc$l4q$1@nnrp1.deja.com> <91nmqj$ls4$1@news.nuri.net>
Message-ID: <3A3F7FE1.2090009@fenx.com>

Both versions work for me on Win95, Win2k and linux.

Emile van Sebille
emile at fenx.com

June Kim wrote:

> "John Grayson" <johngrayson at home.com> wrote in message
> news:91njgc$l4q$1 at nnrp1.deja.com...
> 
>> In article <91n7l6$7ij$1 at news.nuri.net>,
>>   "June Kim" <junaftnoon at nospamplzyahoo.com> wrote:
>> 
>>> I'm using Python 2.0 on Win32.
>>> In the code following, the binding doesn't seem to work; it can't
>> 
>> detect
>> 
>>> mouse double-clicking at all. (Maybe a platform-specific problem, I
>> 
>> reckon)
>> 
>>> from Tkinter import *
>>> 
>>> root = Tk()
>>> 
>>> def callback(event):
>>>     print "clicked at", event.x, event.y
>>> 
>>> frame = Frame(root, width=100, height=100)
>>> frame.bind("<Double-Button-1>", callback)
>>> frame.pack()
>>> 
>>> root.mainloop()
>>> 
>>> 
>> 
>> The problem is that you have to click on the *frame* not the (empty)
>> space in the frame.
>> 
>> Try this:
>> 
>> frame = Frame(root, bg='red', width=100, height=100)
>> 
>> 
>> John Grayson
> 
> 
> Well, it doesn't work, either. Probably it will work on linux but not on
> win32.
> Thanks anyway.
> 
> Best Regards,
> 
> June



From ulf.engstrom at b2b-link.com  Tue Dec 12 08:06:00 2000
From: ulf.engstrom at b2b-link.com (=?iso-8859-1?Q?Ulf_Engstr=F6m?=)
Date: Tue, 12 Dec 2000 13:06:00 -0000
Subject: Print whithout \n
Message-ID: <02d501c0643c$4834a940$1b4b9fc3@no>

With a trailing ,
print 'a',
print 'b',

Ulf

Henrik:
> How do you print a string whithout a linefeed in the end.





From msoeima at netcabo.pt  Wed Dec 20 15:11:29 2000
From: msoeima at netcabo.pt (Marco Soeima)
Date: Wed, 20 Dec 2000 20:11:29 +0000
Subject: Python-list digest, Vol 1 #4292 - 14 msgs
In-Reply-To: <20001220195807.2229DEA5E@mail.python.org>; from python-list-request@python.org on Wed, Dec 20, 2000 at 02:58:07PM -0500
References: <20001220195807.2229DEA5E@mail.python.org>
Message-ID: <20001220201129.A8238@netcabo.pt>

* python-list-request at python.org (python-list-request at python.org) [20/12/00 20:11]:

Content-Description: Masthead (Python-list digest, Vol 1 #4292)
> Send Python-list mailing list submissions to
> 	python-list at python.org
> 
> To subscribe or unsubscribe via the World Wide Web, visit
> 	http://www.python.org/mailman/listinfo/python-list
> or, via email, send a message with subject or body 'help' to
> 	python-list-request at python.org
> 
> You can reach the person managing the list at
> 	python-list-admin at python.org
> 
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of Python-list digest..."
> 

Content-Description: Today's Topics (14 msgs)
> Today's Topics:
> 
>    1. ANNOUNCE: soundex.py (Skip Montanaro)
>    2. DCOracle-1.3.2 makefile problem .er. opportunity? (tim at 2wave.net)
>    3. Re: Problems compiling on Sunos5 (Andreas Jung)
>    4. RE: while true: !!! (Tim Peters)
>    5. Re: dynamic function call (Steve Holden)
>    6. Python in embedded systems? (Grant Edwards)
>    7. object deletion order question (howard at eegsoftware.com)
>    8. Running pdflatex from Python (Brian Elmegaard)
>    9. Re: Catching Database Exceptions (Sybase) (Thomas Gagne)
>   10. Import problem (David Allen)
>   11. Re: Q: dynamic function call (Robert Roy)
>   12. Threads - Newbie Questions (David Lees)
>   13. Re: Running pdflatex from Python (Fredrik Lundh)
>   14. Re: object deletion order question (Fredrik Lundh)

> From: Skip Montanaro <skip at mojam.com>
> Reply-To: skip at mojam.com (Skip Montanaro)
> To: python-list at python.org, python-announce-list at python.org
> Subject: ANNOUNCE: soundex.py
> Date: Wed, 20 Dec 2000 12:44:58 -0600 (CST)
> Message-ID: <14912.65066.958523.18862 at beluga.mojam.com>
> 
> Python 2.0 no longer ships with a soundex module.  Sometime ago, Tim Peters
> and Fred Drake each cooked up replacements written in Python.  I merged them
> together into a single module which is available from 
> 
>     http://musi-cal.mojam.com/~skip/python/
> 
> If you have any questions or comments on the module, please send them my
> way.
> 
> -- 
> Skip Montanaro (skip at mojam.com)
> Support the Mojam.com Affiliates Program: http://www.mojam.com/affl/
> (847)971-7098
> 

> From: tim at 2wave.net
> Newsgroups: comp.lang.python
> To: python-list at python.org
> Subject: DCOracle-1.3.2 makefile problem .er. opportunity?
> Date: Wed, 20 Dec 2000 18:22:14 GMT
> Message-ID: <91qtcd$dic$1 at nnrp1.deja.com>
> Organization: Deja.com
> 
> Hi
> 
> I'm no makefile expert, I'm afraid, and am trying to build DCOracle-
> 1.3.2 in Linux 7 (for Oracle 8.1.6), and not succeeding :-(. I've
> followed the makefile instructions provided, and get to the problem
> below. I haven't found any references to others having this problem on
> comp.lang.python, so I wonder if any of you who have been down this
> path before recognise the error - and point me in directions which
> might overcome it!
> 
> ------------------------------
> root> makefile -f Makefile.pre.in boot
> rm -f *.o *~
> rm -f *.a tags TAGS config.c Makefile.pre python sedscript
> rm -f *.so *.sl so_locations
> VERSION=`python -c "import sys; print sys.version[:3]"`;
> installdir=`python -c "import sys; print sys.prefix"`;
> exec_installdir=`python -c "import sys; print sys.exec_prefix"`;
> make -f ./Makefile.pre.in VPATH=. srcdir=.
>      VERSION=$VERSION
>      installdir=$installdir
>      exec_installdir=$exec_installdir
>      Makefile
> make[1]: Entering directory `/home/tim/rpms/DCOracle-1.3.2/src'
> make[1]: *** No rule to make target'/usr/lib/python1.5/config/Makefile'
> needed by 'sedscript'. Stop.
> make[1]: Leaving directory `/home/tim/rpms/DCOracle-1.3.2/src'
> 
> ------------------------------------------------
> 
> Thanks
> 
> Tim
> 
> 
> Sent via Deja.com
> http://www.deja.com/

> From: Andreas Jung <andreas at andreas-jung.com>
> Reply-To: Andreas Jung <andreas at andreas-jung.com>
> To: Venkatesh Prasad Ranganath <rvprasad at cis.ksu.edu>
> Cc: python-list at python.org
> Subject: Re: Problems compiling on Sunos5
> Date: Wed, 20 Dec 2000 20:02:34 +0100
> Message-ID: <20001220200234.A5081 at yetix.sz-sb.de>
> User-Agent: Mutt/1.3.12i
> 
> On Wed, Dec 20, 2000 at 11:34:15AM -0600, Venkatesh Prasad Ranganath wrote:
> > Hi,
> > 
> > I am trying to compile Python 2.0 on Sunos 5.  I am able to compile but many
> > functions are not being supported.  When I execute configure, the list of
> > functions checked by configure are reported as not supported but all those
> > functions are available.  For example, strftime() is reported as not available.
> > 
> > Is this a problem with the configure script or something I have to specify when
> > configuring.  My configure command was as follows
> > 
> > configure --with-threads=/usr/lib --with-cycle-gc --without-libdb
> 
> Usually you should not care about the functions available or not. When
> "make" works correctly  and you have a running Python interpreter
> after the termination of "make" ...be happy...and "make install".
> I have never seen any real problems with the installation of Python
> under Solaris.
> 
> Andreas
>  
> 

> From: "Tim Peters" <tim.one at home.com>
> To: <python-list at python.org>
> Subject: RE: while true: !!!
> Date: Wed, 20 Dec 2000 14:11:55 -0500
> Message-ID: <LNBBLJKPBEHFEDALKOLCCEMMIEAA.tim.one at home.com>
> 
> [gbreed at cix.compulink.co.uk]
> > ...
> > If you want to empty a sequence, but preserve the type, try
> > filter(lambda x:0, asequence).
> 
> Picking a nit:  that returns an empty list, regardless of type(asequence).
> To get an empty sequence of the same type, do
> 
>     asequence[:0]
> 
> or even
> 
>     asequence * 0
> 
> there's-only-one-way-to-do-it-ly y'rs  - tim
> 
> 

> From: "Steve Holden" <sholden at holdenweb.com>
> Newsgroups: comp.lang.python
> To: python-list at python.org
> Subject: Re: dynamic function call
> Date: Wed, 20 Dec 2000 14:10:39 -0500
> Message-ID: <Fw706.1805$WA2.28069 at e420r-atl2.usenetserver.com>
> Organization: WebUseNet Corp  http://www.usenetserver.com - Home of the fastest NNTP servers on the Net.
> 
> Hwanjo Yu <hwanjoyu at uiuc.edu> wrote in message
> news:Ko606.141$G66.2664 at vixen.cso.uiuc.edu...
> > Hi,
> >
> > Is there any way to bind a function name at runtime ?
> > For instance, we don't know the function name to call, but we know the
> > function name is saved in a variable, "fname".
> > How to call it in this case ?
> > Thanks in advance.
> >
> You just use the variable fname instead of the function name!  For example:
> 
> >>> def myfun(x):
> ...  print "x is", x
> ...
> >>> yourfun = myfun
> >>> yourfun('This is cool')
> x is This is cool
> 
> The "def" statement simply binds the function name to its compiled code
> body.  If you "assign" the functin to another variable, the same call
> mechanism can be used on that binding too.
> 
> regards
>  Steve
> 
> 
> 
> 

> From: grante at visi.com (Grant Edwards)
> Newsgroups: comp.lang.python
> To: python-list at python.org
> Subject: Python in embedded systems?
> Date: Wed, 20 Dec 2000 19:12:58 GMT
> Message-ID: <_w706.11237$TU6.1069571 at ptah.visi.com>
> User-Agent: slrn/0.9.6.2 (Linux)
> 
> Has anybody heard of a Python interpreter being used in an
> embedded system to provide user scripting capabilities?
> 
> At some point in the future, I might be looking for a small,
> safe, interpreted langauge to put into an embedded system. It
> needs to be something easy for a non EE/CS to grasp.
> Resource-wise, Forth would be a good choice except I think it's
> a difficult first language for non programmers.
> 
> I'd love to use Python, but I don't know if it will fit into
> the embedded system environment -- I probably can't allocate
> more than about 1MB of memory for the scripting feature.
> 
> I've seen some nice compact Scheme implimentations, but that
> too is probably not a good choice for something that will be
> used by somebody with no CS background.
> 
> I'm really hoping to avoid using some sort of BASIC, the Web
> server I'm running as a JavaScript subset interpreter, so maybe
> I can figure out how to use that for non-web stuff.
> 
> -- 
> Grant Edwards                   grante             Yow!  Are we on STRIKE yet?
>                                   at               
>                                visi.com            

> From: howard at eegsoftware.com
> Newsgroups: comp.lang.python
> To: python-list at python.org
> Subject: object deletion order question
> Date: Wed, 20 Dec 2000 19:13:40 GMT
> Message-ID: <3a41016d.3397213 at news-server.socal.rr.com>
> Organization: Road Runner
> 
> While locating a crash in a Windows Python extension I am working on
>     ( a Python wrap around a DirectX game library called CDX)
> I discovered my problem as being the order in which Python objects
> were being deleted.  
> 
> In this case, I had created a Sound object and a series of SoundBuffer
> objects.  INSIDE DirectX, deleting the Sound object implicitly
> releases the SoundBuffer objects.  When Python terminated (or at other
> deletions), it called the delete method of the SoundBuffer  objects
> after deleting the Sound objects, leaving DirectX calling into unknown
> data space (I know..it's a Microsoft tradition).
> 
> I patched around this problem by managing a list of buffer objects
> attached to a Sound object (explicitly specifying such objects) and
> deleting them explicitly.
> 
> However, that led my to question:
>         What ordering does Python use for object deletions? 
> 
> I somehow *assumed* that deletion was the reverse order of creation
> (in a global sense) but clearly my problem was a case of
> first-created, first-deleted.  Of course, I realize (now) I should not
> have *assumed* ANY deletion order.  
> 
> For a "good" fix, I will probably have to build some linkage between
> the two 'independent' classes in the C++ code rather than rely on
> working around the problem in Python code.
> Howard Lightstone
> EEGSoftware
> howard at eegsoftware.com

> From: Brian Elmegaard <be at et.dtu.dk>
> Newsgroups: comp.lang.python
> To: python-list at python.org
> Subject: Running pdflatex from Python
> Date: Wed, 20 Dec 2000 21:13:17 +0100
> Message-ID: <3A4112DD.E8B59F3F at et.dtu.dk>
> Organization: Department of Energy Engineering, Technical University of Denmark
> 
> This is posted to comp.lang.python and miktex at dsts.dk
> 
> When I try to run pdflatex (miktex 1.20e) from a Python script with
> popen("pdflatex "+filename) under win98 SE I get a fault in kernel32.dll. 
> 
> This does not happen for pure latex.
> 
> I would like to know if this also happens on other machines or only on mine.
> 
> 
> 
> -- 
> Brian Elmegaard (be at et.dtu.dk) http://www.et.dtu.dk/energysystems 
> Dept. of Energy Engineering, Technical University of Denmark, 
> Phone +45 4525 4169  Fax +45 4593 5215
> :) http://www.rk-speed.dk http://fiduso.dk http://sunsite.auc.dk/dk-tug

> From: Thomas Gagne <tgagne at ix.netcom.com>
> Newsgroups: comp.lang.python
> Reply-To: tgagne at eFinNet.com
> To: python-list at python.org
> CC: Dave Cole <djc at object-craft.com.au>
> Subject: Re: Catching Database Exceptions (Sybase)
> Date: Wed, 20 Dec 2000 13:21:28 -0500
> Message-ID: <3A40F8A8.8CD9A563 at ix.netcom.com>
> Organization: eFinNet, Corp.
> 
> Just to keep checking..
> 
> .. I deliberately used a procedure I know didn't exist.  This time it seems as
> though it tried to test the exceptions:
> 
> Traceback (innermost last):
>   File ".//floater.py", line 50, in ?
>     except Warning, env:
> NameError: Warning
> 
> The DB API 2.0 documentation suggested that all database warnings and errors
> should be subclasses of Error and Warning.  Is there something I'm supposed to
> be importing to make this work, or to define these objects?
> 
> --
> .tom
> 
> 
> 

> From: "David Allen" <s2mdalle at titan.vcu.edu>
> Newsgroups: comp.lang.python
> To: python-list at python.org
> Subject: Import problem
> Date: Wed, 20 Dec 2000 09:24:23 +0500
> Message-ID: <91r1j3$94h$1 at bob.news.rcn.net>
> Organization: Souls, Unlimited
> User-Agent:  Pan/0.9.1 (Unix)
> 
> I have a module that uses a particular object.  So
> at the top of the file, I import that object.  But
> for some reason, trying to create instances of that
> object always fails with a NameError.  Here's the
> error message:
> 
> Unhandled exception in thread:
> Traceback (innermost last):
>   File "./TkGui.py", line 285, in downloadThread
>     contentKey = self.createContentFrame(request)
>   File "./TkGui.py", line 350, in createContentFrame
>     resp = conn.getResource(resource, self.messageBar)
>   File "./GopherConnection.py", line 79, in getResource
>     resp = GopherResponse()
> NameError: GopherResponse
> 
> Now, to verify that I'm not smoking crack, here
> are the first few lines of GopherConnection.py:
> 
> from GopherResponse      import *
> from GopherResource      import *
> from ResourceInformation import *
> 
> And just in case you think I'm crazy and that
> GopherResponse isn't defined in GopherResponse.py,
> here are the first few lines of that file:
> 
> 
> from GopherObject        import *
> from GopherResource      import *
> from ResourceInformation import *
> 
> class GopherResponse(GopherObject):
> 	... code ...
> 
> So here's the issue - I'm importing a module, but
> things in its namespace aren't coming in.
> 
> As far as I can come up with, the only idea I have
> on why this might be happening is because many of
> the objects use each other, so you get a situation
> where many things import many other things, and you
> may be getting a chicken and egg situation with
> the importer.  But I thought python's importer was
> brigher than that.
> 
> What else can cause this?  How is it that you can
> from modulename import * and then NOT have access
> to the things in that namespace?  Alternatively,
> why does this case not work, when, say, I can use
> module X in module Y and it works fine?  
> 
> Any help on figuring this out would be much appreciated.
> -- 
> David Allen
> http://opop.nols.com/
> ----------------------------------------
> It's 5:50 AM...do you know where your stack pointer is?

> From: Robert Roy <rjroy at NO-SPAM.magma.ca>
> Newsgroups: comp.lang.python
> To: python-list at python.org
> Subject: Re: Q: dynamic function call
> Date: Wed, 20 Dec 2000 19:34:30 GMT
> Message-ID: <e1124tgv39ria0drqf3c76c5sr2g2ekg26 at 4ax.com>
> Organization: Excite at Home - The Leader in Broadband http://home.com/faster
> 
> On Wed, 20 Dec 2000 11:56:28 -0600, "Hwanjo Yu" <hwanjoyu at uiuc.edu>
> wrote:
> 
> >Hi,
> >
> >Is there any way to bind a function name at runtime ?
> >For instance, we don't know the function name to call, but we know the
> >function name is saved in a variable, "fname".
> >How to call it in this case ?
> >Thanks in advance.
> >
> Functions are first-class objects in python so it is no big deal to
> assign them to other names and then call them. There are also several
> ways to resolve names at runtime. Here are few quick examples of how
> this could be done.
> 
> Bob
> #########
> # this is the function we will play with
> def cb(name):
>     print 'callback:', name
>     
> # this gets passed the function
> def foo(name, callback):
>     callback(name)
> 
> #this gets passed the function name
> def bar(name, callbackname):
>     globals()[callbackname](name)
> 
> # call foo 
> foo('joe',cb)
> 
> #bind cb to c and call
> c=cb
> c('ralph')
> 
> # use eval to run from function name
> eval('cb')('called from eval')
> 
> # find the function in the locals dictionary
> locals()['cb']('called from locals')
> 
> #find the function in the globals dictionary
> globals()['cb']('called from globals')
> 
> # call bar
> bar('cb called from globals in function', 'cb')
> # note that we bound to c earlier so can call it as well
> bar('c called from globals in function', 'c')
> 
> 
> 

> From: David Lees <DavidL at NONOraqiaMAPS.com>
> Newsgroups: comp.lang.python
> To: python-list at python.org
> Subject: Threads - Newbie Questions
> Date: Wed, 20 Dec 2000 14:42:41 -0500
> Message-ID: <3A410BB1.AAB6BE3C at NONOraqiaMAPS.com>
> Organization: HarvardNET
> 
> I am now starting to play with Python threads.  Could someone refer me
> to documentation other than the Library reference for some simple
> examples?
> 
> My specific questions are:
> 1. I am passing the lock as an argument to each thread.  Is this the
> proper way to do it?
> 
> 2. What is the 'kwargs' - optional keyword arguments refered to in the
> Library Reference for start_new_thread?
> 
> Python is definitely cool.  I am starting to get the hang of it, with a
> lot of help from this group.
> 
> Thanks in advance.
> 
> david lees
> 
> Here is my initial code:
> 
> import time, thread
> 
> def TestProcess(x,n,lk):
>     print 'Process %d is starting' % n
>     #time.sleep(1)
>     lk.acquire()
>     print 'Acquired process %d ' % n
>     lk.release()
>     
> lx=thread.allocate_lock()
> lx.acquire()
> 
> for i in range(100):
>     thread.start_new(TestProcess,(0,i,lx))
> 
> time.sleep(3)
> 
> print 'Threads started, about to release lock'
> lx.release()

> From: "Fredrik Lundh" <fredrik at effbot.org>
> Newsgroups: comp.lang.python
> To: python-list at python.org
> Subject: Re: Running pdflatex from Python
> Date: Wed, 20 Dec 2000 19:45:20 GMT
> Message-ID: <k%706.701$AH6.70148 at newsc.telia.net>
> Organization: Telia Internet
> 
> Brian Elmegaard wrote:
> > This is posted to comp.lang.python and miktex at dsts.dk
> >
> > When I try to run pdflatex (miktex 1.20e) from a Python script with
> > popen("pdflatex "+filename) under win98 SE I get a fault in kernel32.dll.
> >
> > This does not happen for pure latex.
> 
> perhaps you're using an evil version of Norton Antivirus 2000?
> 
>     http://www.python.org/2.0/
> 
>     "Incompatibility warning: Norton Antivirus 2000 can cause blue
>     screen crashes on Windows 98 when a function in the os.popen*()
>     family is invoked. To prevent this problem, disable Norton Antivirus
>     when using Python. (Confirmed on Windows 98 Second Edition with
>     Norton Antivirus version 6.10.20. The same Norton Antivirus version
>     doesn't have this problem on Windows 2000. Norton Antivirus
>     version 5 on Windows 98SE doesn't have this problem either.)"
> 
> </F>
> 
> 

> From: "Fredrik Lundh" <fredrik at effbot.org>
> Newsgroups: comp.lang.python
> To: python-list at python.org
> Subject: Re: object deletion order question
> Date: Wed, 20 Dec 2000 19:45:19 GMT
> Message-ID: <j%706.700$AH6.70137 at newsc.telia.net>
> Organization: Telia Internet
> 
> howard at eegsoftware.com wrote:
> > However, that led my to question:
> >         What ordering does Python use for object deletions?
> 
> any order it likes.  it may remove an object as soon as
> possible, or never at all.
> 
> > When Python terminated (or at other deletions), it called the
> > delete method of the SoundBuffer  objects after deleting the
> > Sound objects, leaving DirectX calling into unknown data space
> > (I know..it's a Microsoft tradition).
> 
> Direct X uses reference counting just like Python.  if you mess
> up the reference count in Python, it'll crash as well.
> 
> avoiding this is easy, of course.  just make sure you keep the
> reference counts in sync:
> 
> -- use Python wrappers for all reference-counted Direct X
>    objects (e.g. SoundObject, SoundBufferObject)
> 
> -- only access Direct X objects via the Python wrappers
>    (e.g. through an LPDIRECTSOUND member in the Sound
>    object class)
> 
> -- add a PyObject (or better, SoundObject) pointer to the
>    SoundBuffer wrapper class, and make it point to the Sound
>    object when you create the buffer (don't forget to call
>    Py_INCREF on it)
> 
> -- only destroy Direct X objects in the Python wrapper's
>    dealloc method.  in the SoundBuffer wrapper, make sure
>    you decrement the pointer to the Sound object.
> 
> if you do this, the sound object will never be destroyed before
> any of its sound buffers.
> 
> > Of course, I realize (now) I should not have *assumed* ANY
> > deletion order.
> 
> why are you asking when you've already figured it out ;-)
> 
> </F>
> 
> 
> 
> 


Content-Description: Digest Footer
> -- 
> http://www.python.org/mailman/listinfo/python-list




From pbjorn at uswest.net  Mon Dec 25 00:23:03 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Sun, 24 Dec 2000 22:23:03 -0700
Subject: Why is Python slow? (was Re: One Python 2.1 idea)
References: <mailman.977590932.19335.python-list@python.org> <9237md$orm$1@nnrp1.deja.com> <924bjt0e3n@news1.newsguy.com> <m38zp6nk8k.fsf@localhost.localdomain> <925mpe$jjg$1@panix6.panix.com> <slrn94db6a.rc.neelk@alum.mit.edu>
Message-ID: <3A46D9B7.3F2D8C42@uswest.net>

Neelakantan Krishnaswami wrote:

> On 24 Dec 2000 12:37:02 -0800, Aahz Maruch <aahz at panix.com> wrote:
> [snip]

> > One of the reasons that gets brought up less often than it should is
> > that Guido is somewhat fanatical about also keeping the CPython
> > implementation clean and simple.  I'm pretty sure that -- Ghu forbid it
> > become necessary -- I could manage to maintain CPython if I had to.
> >
> > This leaves less room for optimization than many people think.
>
> Perhaps.
>
> I started to take a look at the Ocaml 3.00 bytecode interpreter the
> other day, and it is quite clean and readable. It is also right around
> 10-20 times faster than the Python interpreter, measuring just the
> basic busy-loop. The only filips in the Ocaml interpreter that jumped
> out at me were a) some macrology to enable the gcc computed labels
> extensions when it was available, and b) it uses tagging to avoid
> allocating on the heap when handling integers.
>
> [python code]
>
> When passed n=10000000, the median runtime for loop1 was 7.05 seconds,
> and for loop2 was 3.89 seconds.
>
> Some equivalent code in Ocaml:
>
> [ocaml code]

> With n=10000000, this ran in a median of 0.4 seconds. This is the
> performance of the interpreter, mind -- Ocaml also has a compiler
> which I didn't exercise. I think this suggests there is room for
> *huge* performance improvements in interpreted Python.

You're comparing apples and oranges here. Even though both languages were
interpreted in your tests, that's pretty much where the similarity ends.
Ocaml, or any other ML, is designed so that exact type inference can be done,
thus the interpreter knows exactly which types a variable can and cannot
take. This gives the interpreter many more possibilities for safe
optimizations.

A better comparison would be with LISP or Smalltalk, both of which are
usually at least an order of magnitude faster than Python, but at the cost of
significantly more complex interpreters/compilers.

-- bjorn




From loewis at informatik.hu-berlin.de  Sat Dec  9 05:42:00 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 09 Dec 2000 11:42:00 +0100
Subject: SSL how to?
References: <yZ3Y5.3016$9B.89007@news1.oke.nextra.no> <p6qu28eiqig.fsf@informatik.hu-berlin.de> <90s40v$f6d$1@troll.powertech.no>
Message-ID: <p6q1yvhc1vr.fsf@informatik.hu-berlin.de>

"Syver Enstad" <syver at NOSPAMcyberwatcher.com> writes:

> > Yes, httplib supports SSL in Python 2.
> 
> I am using Python 2 under Win 2000 Pro by the way.
> 
> If I understand it right you say that it's a part of the Be Open Source
> distribution but not on Windows?

The support is in the source code, both for Unix and Windows. It is
not included in any of the binary distributions. For Unix, it is
relatively simple to recompile the socket module so that it uses
OpenSSL; for Windows (including NT) people have reported problems
doing so. In any case, you need a working copy of OpenSSL for your
target system, first.

Regards,
Martin


From neelk at alum.mit.edu  Wed Dec 20 19:33:12 2000
From: neelk at alum.mit.edu (Neelakantan Krishnaswami)
Date: 21 Dec 2000 00:33:12 GMT
Subject: Python in "math through programming" curriculum
References: <2g2t3t06r2agrrjmfnr41eu4lc2u39a8qi@4ax.com> <3A3EFF47.67DE69CF@gte.net> <91n9i30pr7@news2.newsguy.com> <slrn9402bt.fa6.neelk@alum.mit.edu> <91qa3r01m78@news2.newsguy.com>
Message-ID: <slrn942nln.mjs.neelk@alum.mit.edu>

On Wed, 20 Dec 2000 13:52:53 +0100, Alex Martelli <aleaxit at yahoo.com> wrote:
>"Neelakantan Krishnaswami" <neelk at alum.mit.edu> wrote in message
>news:slrn9402bt.fa6.neelk at alum.mit.edu...
>    [snip]
>> I disagree with you here about Python's appropriateness. Conceptually,
>> the strongly typed FPLs are the best languages there are for learning
>> about the mathematics of computer programming. The combination of
>
> I wouldn't consider anything but Haskell if I were to teach "the
> mathematics of computer programming" (that may be because my
> exposure to Clean is too limited) stricto sensu [...].
>
> But "math through programming" is a completely different issue,
> at least as the original poster meant it (as he explained in a
> followup to this post of yours) and as I took it -- middle and
> high school levels, as he had said!  

Yes -- the reference to Knuth's _Concrete Mathematics_ made me think
that an introduction to CS was envisioned. For things like graphing
and statistics, there's no reason not to use Python, since it has good
graphing and statistical libraries. Static typechecking doesn't help
here, I don't think, and in Haskell's case it becomes a positive
nuisance (because of all the IO).

> I would *NOT* introduce Haskell in middle school, and even for high
> school, I have my doubts -- students would have to have strong
> *previous* exposure to mathematical concepts, AND to computer
> programming; few high schools, I suspect, are good enough for that.

I do think a typed FPL would work at the high school level, once the
kids have gotten exposed to algebra (for the idea of functions) and
geometry (for the idea of proofs). Then Haskell could be used as a way
of teaching them how to write proofs. It has the two great advantage
that proofs are programs, and hence executable and testable, and that
it has type-checking to help catch errors in them. Call it
experimental constructive mathematics. :)

You would definitely want to improve the quality of the error messages
these implementation emit though. An adult programmer would find them
feasible to work through, but something like

ERROR: Type error in application
*** Expression     : foldr (*) 1 take n (enumFrom 1)
*** Term           : take
*** Type           : Int -> [b] -> [b]
*** Does not match : [a]

will not help a ninth-grader figure out she needs to add some parens
and write "foldr (*) 1 (take n (enumFrom 1))" to compute the factorial.
And this is a very good error message, compared with most!

>> > Caml, IMHO, has rather cluttered syntax when compared to Haskell (or
>> > Python!); it only gets worse when the Object-oriented extensions are
>> > also considered.  Admittedly, a minor consideration, but, _for the
>> > specific issue of programming-newbies_, not a negligible one IMHO.
>>
>> This is the *big* weakness of the ML-family languages (including
>> Haskell and Clean): they all have gnarly syntaxes. Python is proof
>
> Haskell's syntax is *lovely*!  How can you class it in "the ML
> family"?  [Again, I don't know enough Clean to have a real opinion
> about it].
>
>What specific objections do you have to Haskell syntax...?

Honestly, Haskell looks very similar to ML and OCaml to my eyes. An ML
program translated to Haskell will read very similarly[*].

Some of the peculiarities of Haskell's syntax that I find annoying are
the use of case to distinguish the type and value namespaces, the use
of '\' to introduce lambdas (I find this ugly and twee), and the $
syntax to force strict evaluation. However, these are all fundamentally
nuisances, rather than mortal wounds.

However, a really huge lose is the do-notation, which is fantastically
ugly. I *love* the way that monads let me type effects in the language,
but I hate the way I have to type effects in the editor. :) IO is 
important enough that it shouldn't be painful to use.

[*] The usual caveats about tail recursion and lazy languages apply;
I'm assuming them away for the purposes of example. 


Neel



From aahz at panix.com  Wed Dec 27 20:05:58 2000
From: aahz at panix.com (Aahz Maruch)
Date: 27 Dec 2000 17:05:58 -0800
Subject: split hangs
References: <slrn94l1pu.4so.jp@localhost.localdomain>
Message-ID: <92e3lm$pni$1@panix3.panix.com>

In article <slrn94l1pu.4so.jp at localhost.localdomain>,
Jacek =?iso-8859-2?Q?Pop=B3awski?= <jp at ulgo.koti.com.pl> wrote:
>
>My program hangs when running that instruction:
>ourlist=ourregex.split(s)
>I can only kill it, or ctrl+z
>
>ourregex is of course result of re.compile()
>s is a long string
>it works good with other strings, why it hangs with that one?
>how can I fix it? how to find out what is bad?
>how it's possible that split() hangs?
>
>PS. python-1.5.2, Linux

What is the regex?
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

Ninth Virtual Anniversary: 4 days and counting


From grey at despair.rpglink.com  Sat Dec 30 16:55:08 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Sat, 30 Dec 2000 21:55:08 -0000
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <3A4D6A05.EEC97FA6@uswest.net> <slrn94qqvg.mgg.grey@teleute.rpglink.com> <ije36.96813$x6.43604873@news2.rdc2.tx.home.com> <slrn94qt4f.mgg.grey@teleute.rpglink.com> <CMo36.97857$x6.44474719@news2.rdc2.tx.home.com> <slrn94s9qg.134.grey@teleute.rpglink.com> <gdq36.2696$Qb7.302774@newsb.telia.net> <slrn94shbe.150.grey@teleute.rpglink.com> <X8s36.2731$Qb7.305339@newsb.telia.net>
Message-ID: <slrn94smds.150.grey@teleute.rpglink.com>

On Sat, 30 Dec 2000 21:08:07 GMT, Fredrik Lundh <fredrik at effbot.org> wrote:
>Who said "don't init variables?"

    You did.

>Of course you should initialize a variable when it has a meaning-
>ful initial value.

    It does have meaning.

>But in this case, the initial value is just something you made up
>to get into the loop.  

    That would be the meaning.  Otherwise it is uninitted and nothing, or
something unexpected, happens.  Sorry, *plonk*.  You want to argue for
argument's sake, fine.  But I don't think I need any of your poor advice.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From katz at Glue.umd.edu  Tue Dec 12 11:46:48 2000
From: katz at Glue.umd.edu (Roy Katz)
Date: Tue, 12 Dec 2000 11:46:48 -0500
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
In-Reply-To: <UksZ5.3364$O%.525994@newsc.telia.net>
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net>
Message-ID: <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu>

On Tue, 12 Dec 2000, Fredrik Lundh wrote:

> The real problem is that many Python tutorials fail to explain
> how things work (or hide it somewhere in the exercises), but
> that's not really Guido's fault...

Tough for the new Python programmer, then.  It's nt fair to them. 

> (Hint: Everything is a reference to an object.  Variables are
> named references, not actual objects.  Some objects can be
> modified in place.  Some cannot be modified.  That's all)

Okay! tell me, which are the objects which cannot be modified in place? I
assume these are int's, long's, float's, etc.  Strings, lists, objects and
dictionaries are call-by-reference.  Pardon my choice of diction, but I am
against inventing new terminiologies (modified-in-place??). 

I want to know where this is documented.  Specifically, where is the
dichotomy of ints/longs/floats vs. objects/strings/lists/dictionaries. 

While I'm brainstorming with Python grammar changes, what about a real &
operator.  No more of this hokey is-it-modifiable-in-place-or-not
business.  Explicit control is important. 


Thank you for enlightening me!


Roey





From michael at stroeder.com  Tue Dec 12 09:35:25 2000
From: michael at stroeder.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Tue, 12 Dec 2000 15:35:25 +0100
Subject: [Fwd: python-ldap-2.0]
Message-ID: <3A3637AD.4C5977FD@stroeder.com>

HI!

For everyone involved with Python and LDAP programming:

A discussion about the future python-ldap module has just started on
the mailing list python-ldap-dev at lists.sourceforge.net (see
http://python-ldap.sourceforge.net/lists.php for how to join the
mailing list).

Your feedback is appreciated.

Ciao, Michael.

-------- Original Message --------
Subject: python-ldap-2.0
Date: Sun, 10 Dec 2000 23:27:49 +0100 (CET)
From: David Leonard <david.leonard at dstc.edu.au>
To: Michael Str?der <michael at stroeder.com>
CC: python-ldap-dev <python-ldap-dev at lists.sourceforge.net>

On Sun, 10 Dec 2000, Michael Str?der typed thusly:
> I think we should leave the 1.x stuff as is and start with
> python-ldap-ext based on OpenLDAP 2.0.x...

Okay, to keep everyone on the list informed:

Michael, as a knowledgable Karlsruhe local, showed me all the local
restaurants that have closed down and then we discussed the
direction for
python-ldap. After a few drinks, and a discussion on cats, we agreed
on
the following plan:

	* python-ldap-1.10 will be tidied up, with just the memory-leak
	  patches applied, and a proper documentation build. (ie no v3
	  changes will be applied as it breaks v2 builds.) A final release
	  will be made.

	* A new python-ldap-2.0 will be started, with an API that is likely
	  to be somewhat incompatible with that of p-l-1.10. In particular,
	  the following will be the major compatibility targets:
		+ Python-2.0 support (ie unicode)
		+ OpenLDAP-2.0 support (ie LDAPv3/LDAP-EXT)

This means that people relying on v2 client/servers should be happy
enough
with a reasonably stable release.  And people wanting to live in the
3rd
millenium should be happy with a version (relatively) free of legacy
support.

Please, comment! Are there problems with the above plan?

In addition, I seek more comment on any more changes that the _ldap
library
may need to support LDAPv3 connections. (Remember that the purpose
of this
module is not to give an abstraction of X.500 directory services,
but rather
to provide an abstraction of the LDAP C API.)

d
-- 
David Leonard                           David.Leonard at dstc.edu.au
DSTC                                    Room:78-632  Ph:+61 7 336
58358
The University of Queensland            http://www.dstc.edu.au/
QLD 4072  AUSTRALIA                    
B73CD65FBEF4C089B79A8EBADF1A932F13EA0FC8

I put my chin on my knee, and looked for flaws in the soft grain of
my
beige plastic monitor casing. - Julian Assange


From junaftnoon at nospamplzyahoo.com  Tue Dec 19 06:44:10 2000
From: junaftnoon at nospamplzyahoo.com (June Kim)
Date: Tue, 19 Dec 2000 20:44:10 +0900
Subject: Using the fonts of different encodings in wxPython
Message-ID: <91nho6$hk2$1@news.nuri.net>

This problem is of wxPython 2.2.2 and Python 2.0 on Win98.
I'm using 8 bit character set (CJK) on windows. When I call the
wxFontEnumerator, what I get is just a partial list of the fonts on my
system -- only English fonts. Moreover, the wxPython doesn't seem to have
wxFONTENCODING_xxxx that fits my system's character set. What can I do to
use the fonts of different charater sets?





From aleaxit at yahoo.com  Fri Dec 22 17:59:33 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 22 Dec 2000 23:59:33 +0100
Subject: namespaces & indentation
References: <920f54$m2f$1@news.mch.sbs.de>
Message-ID: <920mi30125m@news2.newsguy.com>

"Chris Schaller" <Chris.Schaller at web.de> wrote in message
news:920f54$m2f$1 at news.mch.sbs.de...
> Dear Pythoners...
>
>   I always thought that I understood Python's scoping, but then I came
across
> that code:
> > def t(x):
> >   if x == 1:
> >     c = 'yeah'
> >   return c
>
> If I execute t(1), I'll get 'yeah', but otherwise there'll be an error
> message.  According to the indentation level c is only valid within the
> if-clause, but not on the outside.  So I guess, the indentation level only
> defines the lines of code that belong to the if-clause, but has nothing to
do
> with scopes.

Right!  Only modules, functions, and classes are/have scopes.


>   That's very confusing, I've already programmed a lot in Python, but only
> due to a program error (I forgot to initialize c at the beginning of the
> function - a common error) and an unsuccessful test this happened to a
program
> after months of error-free running.
>
>   Any comments?

Your unit tests apparently had scarce coverage if they never followed
this particular path.  When unit tests have insufficient coverage (a
common issue), any bug may lurk in the never-explored paths -- this
specific one is nothing special in this regard.

I don't recall -- which Python unit-test frameworks have coverage
indications...?


Alex





From g_will at cyberus.ca  Wed Dec 20 17:42:56 2000
From: g_will at cyberus.ca (Gordon Williams)
Date: Wed, 20 Dec 2000 17:42:56 -0500
Subject: Fatal Python error: PyThreadState_Get
Message-ID: <91rcjo$2dtl$1@news2.ottawa.cyberus.ca>

Using
W95
Python 2.0
Pythonwin build 135

I am now in the process of converting some python 1.5.2 code to python 2.0.
I now get the message below.


Fatal Python error: PyThreadState_Get: no current thread

abnormal program termination

I'm sending and receiving information in one thread via a serial port.  In
the 1.5.2 code I was using the Marshal serial module.  As that only works
with 1.5.2, I have now converted that to use windows native stuff using Mark
Hamond's work.

There was a thread earlier this month about this same problem.  I think that
they were also using W95.  It was unclear whether it was resolved or how it
was resolved.

Before I start tracking the problem down, is this a known problem?  What
causes it? Any pointers on where I should look first?

Regards,

Gordon Williams




From jhauser at ifm.uni-kiel.de  Tue Dec 19 14:53:01 2000
From: jhauser at ifm.uni-kiel.de (Janko Hauser)
Date: 19 Dec 2000 20:53:01 +0100
Subject: Converting a list to a string
References: <sRO%5.168$LU6.63588@typhoon.aracnet.com>
Message-ID: <87u280kx2a.fsf@lisboa.ifm.uni-kiel.de>

Use string.join(), which is much more efficent than your approach, which
needs to copy the string in every loop.

HTH,
__Janko
-- 
  Institut fuer Meereskunde             phone: 49-431-597 3989
  Dept. Theoretical Oceanography        fax  : 49-431-565876
  Duesternbrooker Weg 20                email: jhauser at ifm.uni-kiel.de
  24105 Kiel, Germany


From rturpin at my-deja.com  Thu Dec 28 10:02:15 2000
From: rturpin at my-deja.com (rturpin at my-deja.com)
Date: Thu, 28 Dec 2000 15:02:15 GMT
Subject: MySQL and __init__
References: <mailman.978012983.16484.python-list@python.org>
Message-ID: <92fkli$c6g$1@nnrp1.deja.com>

In article <mailman.978012983.16484.python-list at python.org>,
  Egbert Bouwman <egbert at bork.demon.nl> wrote:
> Creating an instance in a database-context means either
> filling an instance with data from an existing table-row,
> or collecting data and adding that as a new row to a table,
> or both. Is there any received wisdom or other advice how
> to do this ?

I haven't used it (yet), but take a look at SQLDict:

http://dustman.net/andy/python/SQLDict

Russell


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


From MarkH at ActiveState.com  Tue Dec  5 19:08:15 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Wed, 06 Dec 2000 00:08:15 GMT
Subject: win32evtlogutil: How does it works ?
References: <3A2D556F.6240C661@club-internet.fr>
Message-ID: <3A2D8223.4070408@ActiveState.com>

G P wrote:

> Any ideas ?

Not really - but you are not alone.  I have the same problem, and havent 
been able to track it down.  The PythonService support for the event log 
works fine, and I can't see the difference between the 2 :-(

I am sure it is just a registry setting, but I can't see what it is...

Mark.



From rodsenra at correionet.com.br  Wed Dec 13 13:15:00 2000
From: rodsenra at correionet.com.br (Rodrigo Senra)
Date: Wed, 13 Dec 2000 16:15:00 -0200
Subject: Problem with += (augmented assignment)
References: <9188of$q55$1@nnrp1.deja.com>
Message-ID: <3A37BCA4.4E2E9B@correionet.com.br>

mjjacob at my-deja.com wrote:
> 
> Hi folks!
> 
> I tried this under Python 2.0 running on Win 2k:
> 
> x = 0
> while x <= 42:
>         x += 1
> 
> Python always complains a Syntax Error: Invalid
> Syntax on this:
> 
>     x += 1
>        ^
> SyntaxError: invalid syntax
> 
> Augmented assignment only seems to work when used
> in first indentation level (script level).

It works for me( as expected ;o)

>>> x=0    
>>> x += 2
>>> while x<=42:
...     x+=1
...     print x
... 
3
4
.... and so it goes!

It is not a bug in the distribution! 
What environment are you using to edit the code ?
Maybe what you see is not what is being fed to the interpreter.
Strange nevertheless.
regards
Rod
-- 
Rodrigo Senra         
Computer Engineer   (GPr Sistemas Ltda)  rodsenra at correionet.com.br 
MSc Student of Reflection (IC- UNICAMP) Rodrigo.Senra at ic.unicamp.br
http://www.ic.unicamp.br/~921234 (see also   http://www.gpr.com.br)




From aleaxit at yahoo.com  Tue Dec 12 11:22:33 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 12 Dec 2000 17:22:33 +0100
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com>
Message-ID: <915jcf0c6o@news2.newsguy.com>

"Steve" <smnordby at yahoo.com> wrote in message
news:3A36368C.26AADDDE at yahoo.com...
> I dislike infinite while loops and use:
>
> for line in file.readlines()
>     ...do stuff...
>
> No need for a break since it will automagically when it runs out of
> lines.

Surely best, *IF* you're 100% sure that all the files you
ever read will always fit comfortably in memory.

If you must provide for reading huge files, this will just
not work.

Would you rather have a program with a construct you 'dislike',
but that will work; or a program you like a lot, that just does
not work?  I.e., are you a pragmatist, or an aesthete?

Python is first and foremost a highly pragmatical language --
although it seems to attract more than its share of aesthetes.

Python's beauty, like that of a sturdily-built brick outhouse,
is not the beauty of frills and decorations, and it's not the
stark minimalist beauty of less-is-more, either: it is, rather,
the beauty of simplicity, solidity, and fitness-for-purpose.

Firmitas, Utilitas, Venustas -- in *THIS* order, as Vitruvius
had it, NOT "comodit?, perpetuit?, bellezza" as Palladio did --
FAR too much software puts intended 'convenience' first, to the
detriment of solidity, but Python, like Unix, is in the GREAT
architectural tradition of *solidity FIRST*.  But that's another,
secondary flamethread (which has been running even since Leon
Battista Alberti swapped the firs two words in the list...) --
the Classical theorists and the Renaissance ones may differ on
the order of the first two priorities, but they do agree in
the placement of the third one... last among the three.

If 'venustas'/'bellezza'/'delight' is not the third of the
triad for you, but higher-up in the pecking-order, then you
may not be getting as much out of Python as it has potential
to give (it may still be a lot, and quite possibly still more
than other languages, but perhaps you might be happier with
languages which afford aesthetics-for-aesthetics'-sake a
higher importance than 'does it run' and 'does it run WELL').


"Infinite" loops are hardly ever appropriate -- loops normally
do terminate (although it may be appropriate to model, with
an "infinite" loop, one that is intended to run until the
machine itself is turned off, for example).  However, loops
that are meant to "cycle until told othewise" (are terminated
by an exception, a signal, or a condition best computed inside
the body of the loop itself) are extremely common.  If your
aesthetic sense interferes with their effective and idiomatic
usage, then it is most likely leading you astray.


Alex





From amk at mira.erols.com  Thu Dec 28 21:27:23 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 29 Dec 2000 02:27:23 GMT
Subject: Python core dumping
References: <92deoj$b2q$1@bob.news.rcn.net>
Message-ID: <slrn94nudk.1ho.amk@mira.erols.com>

On Wed, 27 Dec 2000 08:58:42 +0500, David Allen <s2mdalle at titan.vcu.edu> wrote:
>How can I get python to dump core so I can figure out
>what's wrong with my program, or alternatively, what's
>wrong with python?  Any help would be appreciated.

Is it possible that the script is doing an os.chdir() and the core
file is being written to another directory?  

Assuming that the program really isn't using any C extensions, the
most likely cause of a core dump is an infinite recursion that's
filling up the stack before hitting the recursion limit; try
sys.setrecursionlimit(100) to lower the limit, in the hopes of getting
an exception instead of a core dump.

--amk


From larkas at eden.rutgers.edu  Wed Dec 20 16:39:47 2000
From: larkas at eden.rutgers.edu (Larry Kasoff)
Date: Wed, 20 Dec 2000 16:39:47 -0500
Subject: python with VB/Delphi
References: <91opqf$nr7$1@nnrp1.deja.com>
Message-ID: <3A412723.4778E74F@eden.rutgers.edu>

If you are using Delphi, get Python for Delphi. It is a set of components
that makes it easy. A tutorial is included and the whole thing is
freeware. It has been updated recently so it works with Python 1.6 and 2.
You can get it at:

http://www.multimania.com/marat/delphi/python.htm

dsavitsk at my-deja.com wrote:

> the link form python.org to the tutorial on using vb/delphi interfaces
> to python programs seems to be no longer.  does anyone know of a good
> tutorial on doing this.
>
> do people have a preference for one of these (or neither)?
>
> thanks
> doug
>
> Sent via Deja.com
> http://www.deja.com/



From python9999 at my-deja.com  Mon Dec 18 05:13:19 2000
From: python9999 at my-deja.com (python9999 at my-deja.com)
Date: Mon, 18 Dec 2000 10:13:19 GMT
Subject: Enabling HTML links using xmhtml module.
Message-ID: <91knvt$bbs$1@nnrp1.deja.com>


Hi..
I am writing an application in pygtk, in which i need to display HTML
screens. i could display them using python's xmhtml module, but i am not
able to understand how to enable the links in the screens. i am
displaying the HTML screens on a gtk frame and i know that for enabling
link it needs a browser...

Can we use the module available, for enabling the links??

Can anyone give me a brief overview of what xmhtml module can do??

thanks for any help rendered.
regards,
Priya


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


From c.bertu at libero.it  Tue Dec  5 14:33:29 2000
From: c.bertu at libero.it (Carlo Bertuccini)
Date: Tue, 05 Dec 2000 19:33:29 GMT
Subject: win32security.AdjustTokenPrivileges
References: <3A2D378E.7050504@pixelsystems.com>
Message-ID: <dqbX5.38997$hk4.1468335@news.infostrada.it>

Hi Patrick,

> hProcess = GetCurrentProcess()
> processToken = OpenProcessToken(hProcess, TOKEN_ADJUST_PRIVILEGES |
> TOKEN_QUERY)

There is something strange in your code ... this is the declaration of
OpenProcessToken:

BOOL OpenProcessToken(
    HANDLE ProcessHandle, // handle to process
    DWORD DesiredAccess, // desired access to process
    PHANDLE TokenHandle  // pointer to handle of open access token
   );

The "ProcessToken" is not the Token Handle, but a boolean value: the result
of the function.
The AdjustTokenPrivileges needs the real Token Handle in order to set the
new privileges.

Regards,
--
- Carlo -
c.bertu at libero.it








From thomas at xs4all.net  Tue Dec 26 12:10:33 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Tue, 26 Dec 2000 18:10:33 +0100
Subject: ~Python ?
In-Reply-To: <slrn94hd3j.1br.jp@localhost.localdomain>; from jp@ulgo.koti.com.pl on Tue, Dec 26, 2000 at 03:08:35PM +0000
References: <3a4833f8.346848@news.t-online.de> <slrn94hd3j.1br.jp@localhost.localdomain>
Message-ID: <20001226181033.E25820@xs4all.nl>

On Tue, Dec 26, 2000 at 03:08:35PM +0000, Jacek Poplawski wrote:
> Gerson Kurz wrote:
> >You can have real fun if you try to deleting an
> >actual directory named ~ on Linux, boy. 
> 
> hmm... without -rf it's not dangerous... is it?
> anyway - you can use \~ or file manager (mc for example)

or rm ./~
or rm '~'
or rm "~"
or rm /full/path/to/~

There's more than one way to do it :)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From ullrich at math.okstate.edu  Thu Dec 28 14:57:31 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Thu, 28 Dec 2000 19:57:31 GMT
Subject: how do i check if a file exists with python?
References: <RlM26.98766$yR4.2639444@news1.rdc1.tx.home.com>
Message-ID: <92g5v8$r5k$1@nnrp1.deja.com>

In article <RlM26.98766$yR4.2639444 at news1.rdc1.tx.home.com>,
  "Scott Hathaway" <slhath at DownWithSpam.home.com> wrote:
> I have searched Deja, looked in two pyton books, and searched the
python.org
> site trying to find out this simple question.

Have you tried the _documentation_?

> Can someone please give me the call and the module it lives in?

I don't recall what it is exactly...[five minutes later]:

One way is with glob:

import glob

def FileExists(filename):
	return len(glob.glob(filename)) > 0

if FileExists(r'c:/windows/notepad.exe'):
	print 'yup'
if not FileExists(r'c:/windows/notepad.exk'):
	print "I didn't think so."

May or may not be the Right way.

> Thanks,
> Scott
>
>

--
Oh, dejanews lets you add a sig - that's useful...


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


From walterhnospam at gmx.de  Thu Dec 28 11:08:23 2000
From: walterhnospam at gmx.de (Walter Hofmann)
Date: Thu, 28 Dec 2000 17:08:23 +0100
Subject: cgi parameters
References: <3A4B4BD8.6A696387@wielun.dhs.org>
Message-ID: <slrn94mpbn.3pt.walterhnospam@frodo.uni-erlangen.de>

On Thu, 28 Dec 2000 15:19:04 +0100, QdlatY <qdlaty at wielun.dhs.org> wrote:
>
>Hello!
>
>I need to pass index parameter to script executed by post method (via
>form)
>
>I'm counting on Your help :)

You need hidden parameters:

<form method="POST" action="/cgi-bin/control.py">
...
<input name="hill" value="rctiny" type="hidden">
<input name="action" value="submit" type="hidden">
...
</form>

Walter


From amk at mira.erols.com  Wed Dec  6 21:42:19 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 7 Dec 2000 02:42:19 GMT
Subject: Programming Habits in Python
References: <cspU5.82459$U46.2617014@news1.sttls1.wa.home.com> <8vvb6p$ct2$1@news.wrc.xerox.com> <XiMU5.87123$U46.2803855@news1.sttls1.wa.home.com> <900cpl02ss1@news2.newsguy.com> <Ue3V5.91231$U46.2956112@news1.sttls1.wa.home.com>
Message-ID: <slrn92tv0u.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com>

[Coming to this thread two weeks late -- I've finally given up on
CNRI's news server and have started reading Usenet from home.
This may mean I have to do actual work.]

On Wed, 29 Nov 2000 08:37:08 GMT, Greg Jorgensen <gregj at pobox.com> wrote:
>I'm not contending that all junior programmers are dummies, or even that
>everyone should read Knuth, though it's hard for me to imagine how anyone
>claiming to be a programmer can sleep without Fundamental Algorithms under
>their pillow ;-). 

I sleep quite well, thank you. :) I've started reading Knuth's TAoCP
several times but never gotten very far, mostly because vol. 1 quickly
embroils you in MIX, making it very hard to distill anything of
practical value from the book.  I liken TAoCP to Newton's _Principia_:
monumental, quite important, but not too useful except for
specialists.  If a bright but unstudied person asked me how to learn
physics, I'd certainly point them at a text such as Halliday &
Resnick, not the _Principia_!  The algorithm textbook parallel to
Halliday/Resnick, for me at least, would be the
Cormen/Leiserson/Rivest book; clearly written, covers lots of topics,
fairly rigorous, but not so detailed as to be impossible to read.

--amk




From noahspurrier at my-deja.com  Wed Dec 27 22:54:24 2000
From: noahspurrier at my-deja.com (noahspurrier at my-deja.com)
Date: Thu, 28 Dec 2000 03:54:24 GMT
Subject: Does Python ftplib support APPEND?
Message-ID: <92edhe$g7f$1@nnrp1.deja.com>


As far as I can tell Python ftplib does not
support ftp append (from RFC959). Is that true? Is there
any way around this for me? We need to use append.

We're using Python 2.0

Yours,
Noah


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


From sholden at holdenweb.com  Wed Dec 13 18:38:02 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 13 Dec 2000 18:38:02 -0500
Subject: PgSQL and double quotes...
References: <914i7a$q0a$1@nnrp1.deja.com>
Message-ID: <5OTZ5.774$4y4.27551@e420r-atl1.usenetserver.com>

<karacim at my-deja.com> wrote in message news:914i7a$q0a$1 at nnrp1.deja.com...
> Hi,
> PgSQL doesn't seem to like the data that is
> sometimes in binary format and in somecases
> includes double/single quotes (ie "" or '' etc..)
> which causesng PgSQL to invoke parsing error...!
>
SQL data containing single quotes can be problematic, although you should
not experience any difficulty with double quotes.

> I wrote a simple class before the execute sintax
> as :
>   string.join(string.split(text, "'"), "''")
> but that still doesnt get rid of this "parsing
> error" from PgSQL... anyone has any ideas?
>
This seems to be a fairly convoluted way to write

    string.replace(text, "'", "''")

but appears to give the same result, at least in common cases (I didn't do
exhaustive testing).

This works for me using mxODBC to drive SQL Server 6.5 and 7.0, as well as
Access.  So I can only conclude that either it's the fact you're working
with bonary data or some implementation deficiency of PgSQL (not, I must
say, a system I am familiar with).

> thx
>
> Aus
>

A small code example might allow the group to exercise its collective mind a
little more creatively on solving your problem, if you can isolate the
problem to a ten- or twenty-line test case.

regards
 Steve






From jochen at pc1.uni-duesseldorf.de  Fri Dec  8 05:49:52 2000
From: jochen at pc1.uni-duesseldorf.de (Jochen =?iso-8859-1?q?K=FCpper?=)
Date: 08 Dec 2000 11:49:52 +0100
Subject: parallel port access?
References: <3A268378.4DA1B831@sympatico.ca> <3A272881.6F444750@engcorp.com> <3A2D6761.814ABE8E@bioeng.ucsd.edu> <slrn92su1q.12g.ssthapa@localhost.localdomain> <udhf4gk079.fsf@box.home.de> <3A2FC158.B0AD2CD2@engcorp.com>
Message-ID: <ud66kvmblb.fsf@box.home.de>

>>>>> Peter Hansen wrote on Thu, 07 Dec 2000 11:56:56 -0500:

Peter> Jochen K?pper wrote:

>> So I would say you'll need a generic wrapper function for sending data
>> to the OS, that's it. On Unix I'd say all the problem reduces to find
>> out wether there is a printer-file and what it is called.

Peter> I'm guessing that many of the people interested in this kind of thing
Peter> (direct control of parallel port) want to use it to directly twiddle the
Peter> bits, in order to gain quick-n-dirty control of some external
Peter> equipment.
[...]

Oh, sure. But in that situation (q-a-d, hack, ...) there needn't be a
generic nor clean solution, IMHO:) And "hacking" a parallel port isn't
really that hard, I'd say it's more than easy compared to some unusual
equipment you're going to access.
Moreover such stuff usually doesn't need to be widely portable...

I'd actually suggest the original poster to write a driver for that (a
kernel module on Linux, e.g.). This should provide a abstraction-layer
at the highest possible layer and as such simplify the application!

To my mind low-level hardware-support shouldn't go into a "language"!
Abstraction from the different philosophies of generic parport access
would make sense, though.

Greetings,
Jochen
-- 
Einigkeit und Recht und Freiheit                http://www.Jochen-Kuepper.de
    Libert?, ?galit?, Fraternit?                GnuPG key: 44BCCD8E
        Sex, drugs and rock-n-roll


From clgonsal at keeshah.penguinpowered.com  Sat Dec 30 22:42:44 2000
From: clgonsal at keeshah.penguinpowered.com (C.Laurence Gonsalves)
Date: Sun, 31 Dec 2000 03:42:44 GMT
Subject: HTML Parser
References: <mailman.978224102.28797.python-list@python.org>
Message-ID: <slrn94tb2q.7b4.clgonsal@keeshah.penguinpowered.com>

On Fri, 29 Dec 2000 10:26:31 -0500, Voitenko, Denis <dvoitenko at qode.com>
wrote:
>I am trying to write an HTML parser. I am starting off with a simple
>one like so:
>
...
>newline=re.compile('\n')
...
>input_file = file.read()
...
>jsp_content = newline.split(input_file)

Two things, neither of which answer your question (other have already
done that...):

First, you don't need to use re to split a file into lines. You could've
just said:

jsp_content = file.readlines()

(note that this, like your existing code, reads the entire file into
memory, which might not be a good idea if your file is huge)

Second, (this isn't Python related) you probably don't want to split
your file into lines in any case. HTML is *not* a line based language.
The following is a perfectly valid HTML tag:

    <img
        src="foo.jpg"
        width="640"
        height="480"
        alt="Howdy!"
    >

Your code wouldn't work with such tags, since it works line-by-line.

-- 
  C. Laurence Gonsalves                "Any sufficiently advanced
  clgonsal at kami.com                     technology is indistinguishable
  http://cryogen.com/clgonsal/          from magic." -- Arthur C. Clarke


From pehr at linfoo.eink.com  Thu Dec  7 03:12:52 2000
From: pehr at linfoo.eink.com (Pehr Anderson)
Date: Thu, 7 Dec 2000 03:12:52 -0500
Subject: XML-RPC was [pygtk] Python and Orbit Examples
In-Reply-To: <v04220803b654c258a639@[192.168.11.33]>; from Blair.Lowe@compeng.net on Wed, Dec 06, 2000 at 09:36:42PM -0700
References: <Pine.LNX.4.21.0012050828550.24722-100000@tupelo.daa.com.au> <200012060632.eB66WIK27614@ZOOstation.cc> <20001206165355.B12869@linfoo.eink.com> <v04220802b654744d4fc6@[192.168.11.33]> <20001206194117.D12869@linfoo.eink.com> <v04220803b654c258a639@[192.168.11.33]>
Message-ID: <20001207031252.C9749@linfoo.eink.com>

Dear Blair,

I'm not sure what the best reference is since I haven't set up
the apache + mod_ssl server end before. I assume you could just
use a python cgi script that parses the arguments from a web form.
Then you send it the web form with your XML tucked into a 
form variable.

I don't know how much this changes in the SSL wolrd, but it seems
like you could use the simplifying facade of filling in a secure
web-form by hand, pasting a block of XML into the textfield.

For reference on HTML forms and such, I really like the specs
at http://www.w3.org/MarkUp/

	-pehr


On Wed, Dec 06, 2000 at 09:36:42PM -0700, Blair Lowe wrote:
> Thanks Pehr,
> 
> At 19:41 -0500 2000/12/06, Pehr Anderson wrote:
> >Dear Blair,
> >
> >I'd recommend doing XML over a simple socket, without relying on
> >HTTP or SSL as a transport protocol, but that would require you
> >to include a security model within your XML specification.
> 
> Not my choice. The vendor decides this. I just have to connect and 
> pass XML stuff back and forth. Probably good advice for a vendor, 
> though.
> 
> >
> >It's not a bad idea to handle security separately from transport
> >unless you are set on the SSL security model.  It would be
> >perfectly reasonable to do a HTTP POST with your XML querry
> >and wait for the reply as an XML document, then all you need
> is the SSL interface for python.
> 
> >
> >Python 2.0 includes support for SSL in its web libraries,
> >or you could try to do something with SSL by calling
> >lynx using os.popen("lynx --source https://...").read()
> 
> I shall look into python 2.0. We are using 1.5 right now.
> 
> Pardon my ignorance, but would I connect to the ssl site with a POST 
> context of XML/SSL, or something? How do I get the other site ready 
> to accept XML? Headers? Where can I find more information on this 
> stuff (ie. what book(s), where are docs on line).
> 
> Thanks,
> Blair.
> 
> >
> >         -pehr
> >
> >On Wed, Dec 06, 2000 at 04:05:03PM -0700, Blair Lowe wrote:
> >>  Sounds cool.
> >>
> >>  I found pyXML. at http://download.sourceforge.net/pyxml/.
> >>
> >>  Can one send XML to an SSL site with pyXML as well? If so, where is
> >>  some good docs for this. The version I got (PyXML-0.6.2-2.0.i386.rpm)
> >>  has not docs, and the overview at 4suite say I need to download
> >>  something else.
> >>
> >>  Are there some better tools for doing this. I started to look at
> >>  Zope, but it was rather cumbersome and knowing where to start with
> >>  the documentation was not easy to find either.
> >>
> >>  TTYL,
> >  > Blair.
> >>
> >>  At 16:53 -0500 2000/12/06, Pehr Anderson wrote:
> >>  >Dear Folks,
> >>  >
> >>  >I would highly recommend using XML-RPC as a minimally complicated
> >>  >way to communicate data between your apps.
> >>  >
> >>  >This might sound complicated but it really only means
> >>  >	1. specify XML encodings for your data, use these throught your code
> >  > >	2. open a socket on the target
> >>  >	3. send an XML query and wait for an XML response
> >>  >	4. parse the response
> >>  >
> >>  >The beautiful thing about XML-RPC (remote proceedure call) is
> >>  >that it is just dirt simple. You open a socket, dump in
> >>  >human readable text, and wait for a human-readable response.
> >>  >Everything gets completely defined somewhere and servers and
> >>  >clients could be coded in *anything*!
> >>  >No complicated tools are required to implement any parts of
> >>  >this system though you might want to look at how others
> >>  >parse XML rather than doing it all by hand.
> >>  >
> >>  >And the best part is you'll have an app that anyone can understand.
> >>  >No secret church of CORBA, COM or other "special tools".
> >>  >XML is directly comprehensible by HUMAN BEINGS and that makes it
> >>  >a sustainable encoding with simple tcp sockets for transport.
> >>  >
> >>  >	-pehr
> >>  >
> >
> >  > >
> [snip]
> Computer Engineering Inc. http://www.compeng.net
> Phone: 780 499 5687 (9 - 5 MST) Fax:   780 435 0693 (24 Hours)



From rcameszREMOVETHIS at dds.removethistoo.nl  Sat Dec 23 05:43:07 2000
From: rcameszREMOVETHIS at dds.removethistoo.nl (Robert Amesz)
Date: Sat, 23 Dec 2000 10:43:07 GMT
Subject: namespaces & indentation
References: <mailman.977538666.16764.python-list@python.org>
Message-ID: <901371F49rcamesz@127.0.0.1>

Tim Peters wrote:

>1. Strive to test all paths thru each function (good advice
>regardless of language).
>
>2. It's likely that a framework for warning messages will be added
>to Python 2.1.  In the case above, the compiler could do some flow
>analysis and generate a compile-time warning that "c" may not be
>bound at the time the "return" is executed.  Java's rules make that
>kind of code a compile-time error, but some people play such extreme
>tricks with Python's dynamicism that we'll probably have to settle
>for a nag msg in Python.


It seems to me you're making a good case for a Python-Lint program. 
There are many examples like these floating around, for instance when 
you should write self.SomeVar but accidentally drop the .self in a 
member function it won't be considered an error, even if self.SomeVar 
appears somewhere else in the class. All these technically valid but 
probably erroneous pieces of code could be flagged by a Lint program 
without burdening Python's bytecode-compiler.


>if-thine-eye-offends-thee-pluck-it-out-ly y'rs  - tim

As the saying goes: don't cut off your nose to spite your face.


Robert Amesz


From aleaxit at yahoo.com  Tue Dec 12 08:09:25 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 12 Dec 2000 14:09:25 +0100
Subject: passing list from C to Python
References: <9154s2$2sb@micky.ibh-dd.de>
Message-ID: <9158la030jp@news2.newsguy.com>

"Mike M?ller" <mmueller at dgfz.de> wrote in message
news:9154s2$2sb at micky.ibh-dd.de...
    [snip]
> PyListObject *list = Py_BuildValue("[ff]?, *arg1, *arg2);
> PyErr_Clear();
>
> // the following does not work, need something that calls python function
> with list
>     PyListObject *newList = PyObject_CallMethod(module_object,
> "pythonModul", list);

There's just a small error in the last call (also, this is C++, or
the-very-latest-C-standard -- in classic ISO C89, you have to declare
everything at the START of the block, *and* comments are /* ... */,
NOT //-to-end-of-line).

PyListObject *newList = PyObject_CallMethod(module_object,
    "pythonModul", "(O)", list);

Note that PyObject_CallMethod, much like Py_BuildValue, takes
a format-string describing the following arguments, and builds
the tuple according to it.  I think you do need the parenthesis
in the case of a format-string describing just one object, to
make sure it's still wrapped into a (1-element) tuple, which
is how Python callables expect their arguments (but maybe the
PyObject_CallMethod is smart enough to do it implicitly; the
docs don't make it clear and I haven't checked the sources).

> If there is any other simple way to do it, please let me know.

I think that it would be even simpler if the Python code you
call just modified its list-argument, but it's a minor issue.


Alex





From skip at mojam.com  Thu Dec 14 12:42:09 2000
From: skip at mojam.com (Skip Montanaro)
Date: Thu, 14 Dec 2000 11:42:09 -0600 (CST)
Subject: null bytes in re pattern - difference between 1.5.2 and 2.0?
In-Reply-To: <LNBBLJKPBEHFEDALKOLCMENJIDAA.tim.one@home.com>
References: <14904.14997.494451.840804@beluga.mojam.com>
	<LNBBLJKPBEHFEDALKOLCMENJIDAA.tim.one@home.com>
Message-ID: <14905.1649.544998.48498@beluga.mojam.com>

    me> Accordingly, using raw strings in 2.0 fails.

    Tim> Eh?  Prove it.  That is, submit a bug report with a specific
    Tim> failing example if that's true.

Done.  Further investigation reveals some interaction between re.sub and my
PYTHONSTARTUP, which replaces __builtin__.str.  The bug report is at

    http://sourceforge.net/bugs/?func=detailbug&bug_id=125808&group_id=5470

Skip



From gregj at pobox.com  Wed Dec 20 01:16:27 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Wed, 20 Dec 2000 06:16:27 GMT
Subject: while true: !!!
References: <mailman.977127721.4398.python-list@python.org> <DnD%5.166633$U46.5342944@news1.sttls1.wa.home.com> <slrn93vcml.nh7.qrczak@qrnik.zagroda>
Message-ID: <%8Y%5.172260$U46.5443571@news1.sttls1.wa.home.com>

"Marcin 'Qrczak' Kowalczyk" <qrczak at knm.org.pl> wrote in message
news:slrn93vcml.nh7.qrczak at qrnik.zagroda...

> bool, false and true are keywords in C++ and don't require #inclusion
> of any header. The mentioned compiler is not conforming to ISO/ANSI C++
> (perhaps because it's old).
>
> In C99 _Bool is a keyword, and bool, false and true are macros defined
> in <stdbool.h>.

It was an old compiler. I haven't used C++ in over a year, but I believe
that the addition of bool caused some problems: sizeof(bool) is not
standardized, and bool is defined as 0 is false, anything else is true,
which means a bool expression may not be equal to "true." Please forgive me
if my understanding of this isn't correct: the last C++ compiler I used
didn't have bool support built-in.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From michael at stroeder.com  Thu Dec 28 06:02:01 2000
From: michael at stroeder.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Thu, 28 Dec 2000 12:02:01 +0100
Subject: PEP 222 draft
References: <slrn947bj1.1rs.amk@207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com> <3A43AFEE.AC8F7BD@stroeder.com> <slrn94c7ti.m4.amk@207-172-57-128.s128.tnt2.ann.va.dialup.rcn.com>
Message-ID: <3A4B1DA9.FC201651@stroeder.com>

"A.M. Kuchling" wrote:
> 
> On Fri, 22 Dec 2000 20:47:58 +0100,
>         Michael Str?der <michael at stroeder.com> wrote:
> >This method should either take a parameter or use an object
> >attribute to check the currently used cipher used against a list of
> >accepted ciphers to enforce a local policy or be renamed to
> >something like isSslConn() which does not mean much...
> 
> I'd prefer the renaming;

Me too.

Ciao, Michael.


From db3l at fitlinxx.com  Fri Dec 29 21:40:00 2000
From: db3l at fitlinxx.com (David Bolen)
Date: 29 Dec 2000 21:40:00 -0500
Subject: msxml3 and IStream interface
References: <QV_26.228$Io4.17256@news.siol.net> <3A4D3814.6090201@ActiveState.com>
Message-ID: <uvgs2aayn.fsf@ctwd0143.fitlinxx.com>

Mark Hammond <MarkH at ActiveState.com> writes:

> You are certainly on the right track.  Note that IStream is somewhat 
> special, in that the win32com package knows about this explicitly, so 
> there is no need for makepy support for this interface.  You _do_ need 
> makepy support for the XML lib, as you have done.  (you probably knew 
> all that - just being clear)

In a similar vein (already known interfaces versus not), I'm beginning
to look/work more with the server side of things, and was wondering
how one might go about using win32com to produce a COM server using a
non-IDispatch interface, where that interface wasn't one of the
predefined ones in pythoncom.

Your book doesn't seem to really touch the construction of
non-IDispatch interfaces - is the most practical approach for me to
build a typelib via standard means (compiling an IDL), then use makepy
to define it for win32com, which then would let me declare it in the
_com_interfaces entry in my class?

And if I'm intending to use these over DCOM and might involve
non-standard marshaling code, is there a way to make that code
available to win32com?

Book and/or code pointers are fine (I've saved various COM stuff from
you and Alex in the past which I'm still also going through).

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From rodsenra at correionet.com.br  Sat Dec  9 20:53:23 2000
From: rodsenra at correionet.com.br (Rodrigo Senra)
Date: Sat, 09 Dec 2000 23:53:23 -0200
Subject: python 2.0
References: <3A3255E2.DC54424B@coretec-inc.com>
Message-ID: <3A32E213.CB797E6@correionet.com.br>

James Amundson wrote:
> 
> Does anyone know where I can download a binary
> version of python 2.0 for HP-UX 10.20.

AFAIK, there is only version 1.6 available (compatible
with swinstall). However, I had success in compiling
Python 2.0 src to HP10.20 and HP11. Just download gcc
and it will be a breeze! (Just a hint, I disabled thread
support during compilation. Though I'm not saying it
doesn't work on HP-UX. )
HTH
Rod
-- 
Rodrigo Senra         
Computer Engineer   (GPr Sistemas Ltda)  rodsenra at correionet.com.br 
MSc Student of Reflection (IC- UNICAMP) Rodrigo.Senra at ic.unicamp.br
http://www.ic.unicamp.br/~921234 (see also   http://www.gpr.com.br)




From emile at fenx.com  Sun Dec 31 00:31:25 2000
From: emile at fenx.com (Emile van Sebille)
Date: Sat, 30 Dec 2000 21:31:25 -0800
Subject: What happened in this class
Message-ID: <92mgca$7kktg$1@ID-11957.news.dfncis.de>

I saw Pearu's lazy extension and built and installed it on
my linux system(recent cvs).  I thought I'd see how it
played in a class vs getattr and got this result.  Thinking
something broke along the way, I then switched over to the
win95 system(2.0 #8), and got the same thing.  This doesn't
happen under 1.52.


class test:
    def __getattr__(self, name):
        return "egs"

t = test()
print dir(t)[1:5]

['e', 'g', 'g', 's']


Now I know this is contrived, but I still didn't expect
these results.  Should I have?

--

Emile van Sebille
emile at fenx.com
-------------------






From barry at digicool.com  Mon Dec  4 12:14:51 2000
From: barry at digicool.com (Barry A. Warsaw)
Date: Mon, 4 Dec 2000 12:14:51 -0500
Subject: Mailing lists are moving
Message-ID: <14891.53515.250965.697739@anthem.concentric.net>

Hello folks,

Digital Creations has reserved a new machine to handle the Mailman
mailing lists for the Python.Org domain.  The machine that currently
runs the mailing lists has been having resource problems for a long
time now.  I hope that with the boost in hardware and disk space, the
lists will run smooth for a long time to come.

Over the next several days, I'll be migrating the lists one-by-one to
the new machine.  This should be almost entirely transparent to you.
You don't need to change the way you access the lists, either via
email or via the web.

However during the transition for each list, I will be shutting off
mail and web for short periods of time, making Mailman inaccessible
either via mail or web for a few minutes.  Mail will be less of a
problem because it'll just queue up until the server is restarted.
Web access will fail temporarily, but a reload a few minutes later
should be just fine.

This will /not/ affect access to the rest of the www.python.org
material.

You can send questions or problems to me directly.
-Barry



From othello at javanet.com  Sun Dec 31 01:09:45 2000
From: othello at javanet.com (Raymond Hettinger)
Date: Sun, 31 Dec 2000 01:09:45 -0500
Subject: 2001 Enchancement Wishlist
References: <3A4CEACB.220998A0@javanet.com> <92iv76$flh$1@slb2.atl.mindspring.net>
Message-ID: <3A4ECDA9.E2796F78@javanet.com>

Andrew Dalke wrote:

> Raymond Hettinger wrote:
> [snip]
> >I recommend that we extend the same courtesy to
> >   the % string formatting operator.  For example:
> >
> >             print "Spam with %s, eggs, and spam" % aUserDictionary
>
> Time machine beat you to it.  I wrote a docstring formatting
> package with templates based on % interpolation using a
> specialized dictionary-like object.  You could do something like:
>
>    "%(module.name|htmlescape|h1)s<p>%(module.docstring|htmlescape)s"
>
> It would parse up the components, get the first term (including
> the "." parts) then apply the results through the set of
> "|" transformations.
>
> >>> class MyDict:
> ...     def __getitem__(self, name):
> ...             print "Asking for", repr(name)
> ...             return "*"
> ...
> >>> d = MyDict()
> >>> "%(module.name|htmlescape|h1)s<p>%(module.docstring|htmlescape)s"%\
> ...     d
> Asking for 'module.name|htmlescape|h1'
> Asking for 'module.docstring|htmlescape'
> '*<p>*'
> >>>
>

Thank you time machine ;) The question remaining is how to hurry along the
existing PEP for extending eval() to accept all mapping objects, as in:

class SpreadSheet:
    _cells = {}
    def __setitem__( self, key, formula ):
        self._cells[key] = formula
    def __getitem__( self, key ):
        return eval( self._cells[key], self )

ss = SpreadSheet()
ss['a1'] = '5'
ss['a2'] = 'a1*5'
print ss['a2']


>
> >   I think [__init__] should continue to return the new object
> > by default but also allow another return item to be substituted.
> > For example:
> >
> >            def __init__( self ):
> >                   if MyClass.aSingleton == None:
> >                         MyClass.aSingleton = self
> >                   return MyClass.aSingleton
>
> Use a factory function for the general case.  For example
> if you need this singleton behaviour you can do:
>
> class MyClass:
>   pass
> _MyClass_singleton = MyClass()
> def MyClass():
>   return _MyClass_singleton
>
>

I was hoping to avoid a factory method or global variable or using
modules to create single instances.  Compared to enriching __init__,
these approaches seem kludgy, break the encapsulation of the class,
and are hard to implement after clients are already making calls to
the constructor.  The decision to make a object a Singleton or to
revoke that decision should be invisible to clients of the class.

Only one of the posts so far reacted positively to the idea, but none
indicated that it would be hard to implement.  It would seem to add
flexibility without breaking anything currently written.

>
> Oh, and __del__ also does not allow a return value.

While I can imagine potentially useful return values for __del__,
that method is unique in that it can be called by a statement rather
than a function.

>
> >4.  Allow dictionaries to respond to sequence operations
> > as if .keys() were in the list object.  Examples:
>
> What if the keys in the dictionary are integers?
>
> data = {1: "one", 3: "three"}
> print data[1]  # This prints "one"
> print data[0]  # Do you want this to be valid or an error?
>

Thank you for ferretting out a design conflict.
Here is a revised wish (borrowed from AWK):

Override the IN keyword for dictionaries and access the keys directly:

knights = { 'lancalot':'good', 'gallahad':'brave', 'robin':'coward' }
for k in knights:
        print k, knights[k]
if 'arthur' in knights:
        print 'Not a king'


Thank you for your insightful comments,

Raymond Hettinger
othello at javanet.com



From ahnzzang at dreamwiz.com  Tue Dec 12 23:21:06 2000
From: ahnzzang at dreamwiz.com (=?utf-8?B?7JWI7JiB7Jqw?=)
Date: Wed, 13 Dec 2000 13:21:06 +0900
Subject: Python-list digest, Vol 1 #4135 - 3 msgs
References: <20001204170136.623E41CE9D@dinsdale.python.org>
Message-ID: <003e01c064bc$3ca5c520$c5782bd3@www>

I'M SORRY
I WISH DELETTE MY MEALE ADDRESS
 
  ----- Original Message ----- 
  From: python-list-request at python.org 
  To: python-list at python.org 
  Sent: Tuesday, December 05, 2000 2:01 AM
  Subject: Python-list digest, Vol 1 #4135 - 3 msgs


  Send Python-list mailing list submissions to
  python-list at python.org

  To subscribe or unsubscribe via the World Wide Web, visit
  http://www.python.org/mailman/listinfo/python-list
  or, via email, send a message with subject or body 'help' to
  python-list-request at python.org

  You can reach the person managing the list at
  python-list-admin at python.org

  When replying, please edit your Subject line so it is more specific
  than "Re: Contents of Python-list digest..."




------------------------------------------------------------------------------


  Today's Topics:

     1. Re: [OT] Apple IIe help please (Brian McGroarty)
     2. Re: Python mailing list/usenet bridge. (Simon B.)
     3. (no subject) (Eric Stechmann)



------------------------------------------------------------------------------


  -- 
  http://www.python.org/mailman/listinfo/python-list



------------------------------------------------------------------------------



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20001213/49d01991/attachment.html>

From aleaxit at yahoo.com  Fri Dec 29 03:30:38 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 29 Dec 2000 09:30:38 +0100
Subject: how do i check if a file exists with python?
References: <RlM26.98766$yR4.2639444@news1.rdc1.tx.home.com> <phU26.31$KD3.4767@typhoon.aracnet.com>
Message-ID: <92hi3h01f0d@news2.newsguy.com>

"Daniel Klein" <DanielK at aracnet.com> wrote in message
news:phU26.31$KD3.4767 at typhoon.aracnet.com...
> How about something like
>
> def fileExists(f):
>     try:
>         file = open(f)
>     except IOError:
>         exists = 0
>     else:
>         exists = 1
>     return exists
>
> Although this is not as elegant as the person who submitted the 'glob'
> solution.

This is one way to answer "did the file named by string f exist *IN A
FORM THAT COULD BE OPENED FOR READING BY THIS PROCESS* _at the time the
question was posed_", which is rather different from "does the file so
named exist at all right now".  Note that os.access with os.R_OK is
answering a subtly *different* question on systems where setgid/setuid
is a possibility (this often trips people up, but it's more of an issue
for non-scripting languages, since setuid scripts aren't popular, and
rightly so:-).

When somebody asks how to determine whether a file exists, it IS quite
likely that what they mean is something different, such as, whether it
could be read.  os.access with os.F_OK, or os.path.exists, or
os.path.isfile (maybe in conjunction with 'not os.path.islink', but
that depends on what position one takes on 'existence' of symbolic
links...), all answer existence-questions of some form or other.

It's unlikely, although of course possible, that 'actual existence'
is exactly what one needs to determine, and ALL one needs (and, in
general, if this IS indeed the question, it is not answerable: a file
COULD exist, but in a way that remains totally hidden from the current
process).  try/except *right in the code that DOES something with the
file* (typically opening it, if possible) is far more likely to be the
right approach -- it's just about the only way to address the issue
of 'right now' as opposed to 'back when I asked', although not totally
(on some systems, under certain situation, a file COULD be deleted
"from under you" even if you have it open... but, it's less likely to
happen than if you _didn't_ keep it open after asking:-).


Alex





From grante at visi.com  Thu Dec  7 11:14:34 2000
From: grante at visi.com (Grant Edwards)
Date: Thu, 07 Dec 2000 16:14:34 GMT
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <3A2DD2CA.14A5D7E2@engcorp.com> <90l3j9$4bn$1@kapsel.intranet> <udzoi9cedt.fsf@box.home.de>
Message-ID: <KHOX5.3213$TU6.94362@ptah.visi.com>

In article <udzoi9cedt.fsf at box.home.de>, Jochen K?pper wrote:
>>>>>> Piotr Kuchta wrote on Wed, 6 Dec 2000 11:15:05 +0100:
>
>Piotr> that's true that Dutchmen are very talented: common Dutchman
>Piotr> speaks well two foreign languages, English and German.
>
>Isn't that simply what Dutch is all about? 
>Where's the foreign language;-))
>
>Greetings,
>Jochen
>
>PS: Yes, most Dutch people are much better at foreign languages than I am!

If a person who speaks two languages is "bilingual" and a
person who speaks three languages is "trilingual", then what do
you call a person who only speaks one langage?



...





...







American!

-- 
Grant Edwards                   grante             Yow!  I was making donuts
                                  at               and now I'm on a bus!
                               visi.com            


From root at [127.0.0.1]  Mon Dec 11 08:19:02 2000
From: root at [127.0.0.1] (deadmeat)
Date: Mon, 11 Dec 2000 13:19:02 GMT
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com>
Message-ID: <av4Z5.7633$KY1.20581@news1.rivrw1.nsw.optushome.com.au>

> >     count = 0
> >     while 1:
> >         lines = file.readlines(8192)
> >         if not lines: break
> >         count = count + len(lines)

I'm new to Python, but I use 'break' and 'continue' regularly in others.
originally i considered the practise evil, and not much better than GOTOs,
however as my programs started getting bigger and more reliable (ie. i
started checking for all error conditions :), if..then..else, while..do and
for blocks just became a enourmous mess. as irritating as GOTOisms are, it's
a lot neater to simply short circut the block when something occurs (eg. an
error) than intent an entire block due to one single (if, while, for, etc)
statement.

i suspect the problem will be even worse in python because indentation is
mandatory. then again, it uses try..except to catch errors, which helps cut
the if..then trees down to just one pair.





From emile at fenx.com  Mon Dec 25 14:09:06 2000
From: emile at fenx.com (Emile van Sebille)
Date: Mon, 25 Dec 2000 11:09:06 -0800
Subject: __init__ question
References: <EvM16.270$LU6.118037@typhoon.aracnet.com>
Message-ID: <92860t$6950p$1@ID-11957.news.dfncis.de>

*args (and **kwargs) allow for additional parameters to be
passed to a function.
*args in the def becomes a tuple, **kwargs a dictionary.

>>> def init(a, b, c, *args, **kwargs):
 print a,b,c
 print args
 print kwargs


>>> init(1,2,3,4,5,6,d=7,e=8,f=9)
1 2 3
(4, 5, 6)
{'f': 9, 'd': 7, 'e': 8}
>>>

--

Emile van Sebille
emile at fenx.com
-------------------


"Daniel Klein" <DanielK at aracnet.com> wrote in message
news:EvM16.270$LU6.118037 at typhoon.aracnet.com...
> I'm trying to decipher some Python code and came across
the line:
>
>     def __init__(self, *args):
>
> What is the meaning of the '*args' expression?
>
> Thanks,
> Dan:
>
>




From noahspurrier at my-deja.com  Tue Dec 19 16:17:27 2000
From: noahspurrier at my-deja.com (noahspurrier at my-deja.com)
Date: Tue, 19 Dec 2000 21:17:27 GMT
Subject: Wanted: Python White Papers
Message-ID: <91oj8v$i0r$1@nnrp1.deja.com>

Is anyone aware of any white papers to explain the use of Python
at any big, high-profile companies?
I'm trying to use a "look-how-these-cool-companies-use-python"
argument to defend the use of Python for a project. That won't be my
only argument, of course :-) But currently Python is not winning many
management converts here based solely on its technical merits.
Many still say, "What's Python? Never heard of it...". It helps
a lot when I say "Google uses it!", but it would help more if I
knew HOW it is used and WHY.

Actually, the only high-profile company that uses Python that I
can think of is Google. Is anyone familiar with their architecture?
How does Python fit into their design?

Is there anyone out there from Google that can represent Google's use
of Python? What are Google's plans for the future of Python in their
architecture? I heard that they might be switching to Java.

Yours,
Noah Spurrier


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


From jcm at bigskytel.com  Thu Dec 28 07:14:06 2000
From: jcm at bigskytel.com (David Porter)
Date: Thu, 28 Dec 2000 05:14:06 -0700
Subject: Is this a BUG?
In-Reply-To: <3A4B0C5E.1E7976D2@wielun.dhs.org>; from qdlaty@wielun.dhs.org on Thu, Dec 28, 2000 at 10:48:14AM +0100
References: <3A4B0C5E.1E7976D2@wielun.dhs.org>
Message-ID: <20001228051406.A9310@bigskytel.com>

* QdlatY <qdlaty at wielun.dhs.org>:

> I have script:
> ----------------
> import cgi
> 
> form = cgi.FieldStorage()
> 
> print form
> print "!",form["index"].value,"!"
> ------------------
> I launched it with parameter index=1
> And this is an output:
> ----------------
> FieldStorage(None, None, [MiniFieldStorage('index', '1')])
> ! 1 !
> ----------------
>  ^ ^
> Why the hell there are spaces? index property has value '1'!
> not ' 1 '?!

Because you asked for them:

>>> print 'a', 'b'
a b
>>> print 'a'+'b'
ab
>>> print 'a' 'b'
ab

see: http://www.python.org/doc/current/ref/print.html

Also, there are several tutorials on this page which will prime you on the
basics of Python: http://www.python.org/doc/Intros.html


David



From paul.moore at uk.origin-it.com  Tue Dec 19 11:37:22 2000
From: paul.moore at uk.origin-it.com (Paul Moore)
Date: Tue, 19 Dec 2000 17:37:22 +0100
Subject: Events in win32com
References: <Y1g=Op6IfWSoJn1ISFtOyuvqwBb3@4ax.com>
Message-ID: <FY0=OlZFLy71DtnK62u5TEPlhOts@4ax.com>

On Tue, 19 Dec 2000 13:49:23 +0100, Paul Moore
<paul.moore at uk.origin-it.com> wrote:

>If I create an object using
>
>obj = win32com.client.Dispatch(progid)
>
>can I see events from the object?

After scanning Deja, I found an article describing DispatchWithEvents.
The example in the documentation (OnVisible) works fine. But the
OnQuit event doesn't seem to work :-( Basically, the event routine is
never fired. Is this because the event routine has a ByRef "Cancel"
parameter? Or is it something else?

Thanks,
Paul



From swanky_mode at my-deja.com  Sun Dec 10 14:01:19 2000
From: swanky_mode at my-deja.com (swanky_mode at my-deja.com)
Date: Sun, 10 Dec 2000 19:01:19 GMT
Subject: Problems With socket Under NT & PythonWin 2.0
Message-ID: <910jtt$nnh$1@nnrp1.deja.com>

Greetings;

I'm trying to learn a bit more about Python programming under Windows
NT, and came across a problem with the Socket EchoServer.py example in
section 7.2.2 inside the Python reference pages.

As soon as I run EchoServer inside the PythonWin IDE with Python 2.0,
the entire IDE "hangs"; e.g., the mouse pointer icon is set to an hour
glass, and there is no change in its settings, as well as not being
able to click on any of the toolbar buttons.  The only way I can stop
the application is by closing PythonWin from the NT Task Manager.

With this in mind, I have the following questions:

Are the socket examples inside the Python reference Unix/Linux only?

Do I need any additional modules or libraries to run socket programs
inside PythonWin?

Do I need to change any of my environment settings?

I'm running Windows NT 4.0 with Service Pack 6a applied.  I do have the
TCP/IP protocol stack installed and configured.  I have no problem
running any other socket based applications I've done in C++ or Java,
and I can run the Peer Web Server Services (IIS) without any problems.
I also have Cygwin 1.0 installed as well.

Any help or insights on this problem would be greatly appreciated.

Thank you in advance for your time and help...



Sent via Deja.com http://www.deja.com/
Before you buy.


From robin at jessikat.fsnet.co.uk  Sat Dec 30 19:34:49 2000
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sun, 31 Dec 2000 00:34:49 +0000
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <05C6FUhLDNUU-pn2-tuTnzOCokxem@bill>
Message-ID: <lgwFRKAp8nT6Ew$P@jessikat.fsnet.co.uk>

In article <05C6FUhLDNUU-pn2-tuTnzOCokxem at bill>, William Sonna
<wsonna at attgloball.net> writes
>On Sun, 30 Dec 3900 03:00:05, henry_crun3583 at my-deja.com wrote:
>
>> Can someone point me at an explanation of the logic for not having
>> do:
>> until
>
>Its a design flaw in Python.
>
>There is no rational reason not to have it. Anyone who has worked with
>Pascal or Rexx (or even regular expressions) knows that "one iteration
>plus zero or more iterations" is NOT as good as "one or more 
>iterations", because the resulting code is sloppier.
>
>So while there are workarounds for this bit of illogic, they are, 
>nonetheless workarounds.
>
in this case I have to agree with the pascalers there's no reason not to
have these extra loop constructs other than style/convention. On the
other hand there are other loop constructs so if we add one why not the
others eg

        T=0
        until T>=25:
                stuff

or 
        loop:
                stuff
                if T>=25: break

of course the the above/repeat seems to be redundant as

        while 1:
                T=....
                if T>=25: break

Python is actually a bit broken as continue inside try blocks is not
allowed.

The malevolence/stupidity of the data set used by Guido will forbid any
such minor modifications of the language as his original tutees
presumably chose just those constructs which, in his benevolence, he now
allows.
-- 
Robin Becker


From aleaxit at yahoo.com  Tue Dec 12 12:26:38 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 12 Dec 2000 18:26:38 +0100
Subject: Inheriting the @ sign from Ruby
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu>
Message-ID: <915n4k0gbp@news2.newsguy.com>

"Roy Katz" <katz at Glue.umd.edu> wrote in message
news:Pine.GSO.4.21.0012121027001.14542-100000 at y.glue.umd.edu...
> On Tue, 12 Dec 2000, Alex Martelli wrote:
>
> > 'self' is not "pasted in front of a name" -- 'self' is an object,
> > and it is used with perfectly regular syntax for object access.
>
> Yes, we understand this.  Passing 'self' (or whatever people choose to
> call it) to an instance method is automatic upon invocation, and manual
> for anything else.  So we have:

I'm not sure what you mean in this context by 'invocation'.  Consider:

    foo = FooClass()

    f = foo.method
    g = FooClass.method

    f('pap')
    g(foo, 'pap')

The passing of the first-argument object is automatic "upon
invocation" of a BOUND-method-object; it's NON-automatic
("manual", if you wish) "upon invocation" of an UNBOUND-
method-object.  (Where in each case I am taking 'invocation'
as a synonym of 'call').

So, I'm not clear what exact distinction you're trying to
draw here by saying 'upon invocation' as against "anything
else".  Aren't the calls of f and g 'invocations' both?
Or do you distinguish as 'invocations' some of the calls
to a callable object, depending on said object's nature?
If the latter, then I've never heard of such a distinction.


> To me, grammatically and practically, self *is* pasted in as the first
> argument.  I'm bringing up a grammatical question.  My question was
> whether or not the shorthand was practical.

When you call a bound-method, the first argument can be,
I guess, said to be 'pasted' (strange terminology!).  But
that has nothing to do with the USE of the first (or
other) arguments in the body of functions that receive
them.  Such arguments are used with uniform syntax, no
matter whether they're the first, or not -- just the
same, identical syntax as that available for ANY use
of ANY other object.

Your 'shorthand' singles out ONE subcase (attribute
access via dot-notation) out of all, for ONE object
(the one received by a method as its first argument)
out of all, introducing an alternative way of
expressing that single, specific combination out of
all the huge variety; in other words, it introduces
a special-case, ad-hoc irregularity in a perfectly
simple and regular scheme.  This is one of the reasons
I find it a terrible idea (and from other responses
I've read on this thread, it seems to be a widespread
opinion).


> > So, you would like '@identifier' to be a special-case
> > syntax meaning exactly the same as 'self.identifier' (if
> > the first argument, as per normal convention, is named
> > 'self').
>
> I would like the @identifier to denote <self-object>.identifier, where
> <self-object> is the first parameter to a method. Regardless of whether it
> is 'self' or 'myself';  the 'def bind(myself, x)' was just an example.

If the singled-out argument is not named 'self', as it normally
is, readability will suffer anyway, I keep asserting.  The
'shorthand' will do nothing substantial to help, and if it
encourages such breaches of convention it will positively
hinder, the current uniformity of usage.


> > such a violation of
> > normal expectations will make your code harder to read
> > and understand by any maintainers.
>
> Why should there be any violation if all instance identifiers are labelled
> with a @? To be honest, I believe that the @identifier convention would
> decrease the number of variant programming styles.  We all want that

How can it possibly DECREASE them?  It would introduce ONE MORE way
to express the SAME, IDENTICAL thing, and thus it can only INCREASE
variability of expression, as some code says '@goo' while other
code keeps saying 'self.goo' instead.

> 'self' be used (instead of any other name).  Fine! the @identifier
> scheme would at least reduce this, then. You wouldn't see 'myself' (or
> some other deviant) pasted in front of every instance variable; you'd
> see a common @ sign. So for reading someone else's code, that's a good
> thing.

Absolutely NOT.  If lulled by the @-signs in most places, I'm LESS
likely to notice that the code's author has wantonly called the
self-object 'myself', while 'self', if it exists at all, may refer
to some other object (say a global).  So when the time comes that
the self-object is used, or should be used, e.g. in calling a base
class method, or getattr, etc, etc, it becomes MORE likely, not
less, than the confusion will prove damaging.

*EXPLICIT IS BETTER THAN IMPLICIT* -- a very important guiding
principle of Python programming.
    localvar=self.goo
is more explicit than
    localvar=@goo
in that, if, for example, I need to change this for maintenance
to
    localvar=getattr(self,'goo',None)
(a reasonably frequent occurrence), I don't have to unravel the
implicitness -- I *KNOW*, because I read it *right there*, the
name of the object that I need to pass to getattr (etc etc).


> > > What faults or strengths do you see with this?
> >
> > I see no strength in this supplementary syntax form
> > offering a second, different way to express what can
> > be perfectly well expressed explicitly today.
>
> shorthand??

Is saving 4 characters, '@' vs 'self.', so important as
to warrant the 'fork' in coding styles, the redundancy
itself, etc?  Not to me.  Most Pythonistas seem to be
even more minimalist, begrudging such new constructs as
list-comprehensions (which save a lot more than 4 chars,
and even have some 'conceptual' advantages).


> > I see it as a fault that all this does is offer a
> > second, different way to express what can be perfectly
> > well expressed explicitly today.
>
> Save for being able to call the first parameter of a method whatever you
> like, wouldn't this reduce the number of programming conventions?

Maybe there's an implication I find it hard to swallow: would
you want to make 'self.foo' ILLEGAL, breaking _all_ existing
Python code?  Because, if not, it appears to be obvious that
some code will keep saying 'self.foo' while other says '@foo',
and I really don't understand how you can possibly claim the
obviously-false 'reduction' in 'number of conventions'...

> > It would also be a (small) step along the road of
>
> I disagree; Like I mentioned above, this would enforce
> standardization.

...or the 'enforcement of standardization', either.


> > turning the currently crystal-clear syntax of Python
> > into that morass of special-purpose, ad-hoc, 'convenient'
> > kludges that make certain other languages' typical
> > programs look so peculiarly similar to line noise.
>
> Python's language is not crystal-clear, imho.  If we're so tough on

I was (and am) claiming crystal-clear *SYNTAX*.

> perl for having a kludgy interface, why do we have {} for
> defining dictionaries instead of [ key:value ]?

Because the Python object denoted by the syntax:
    [ whatever_you_want ]
is *ALWAYS* a list, and a dictionary is not a list.  Breaking
this 'always' would hardly enhance language-clarity!


> Why do we have to
> explain to new Python programmers how the label-value system Python uses
> is different than traditional call-by-reference and
> call-by-value?

Sez who?  People I've seen with Python as their first
language, and people I've seen coming from Java (which
has exactly the same argument-passing mechanism), seem
to have no trouble (there are many other languages which
work this way, but I have not observed Python users who
started out in any of those other languages).  I'm not
sure what you mean by 'label-value', but the semantics
of argument-passing, in Python, is exactly that of the
'assignment' operator, which simplifies the exposition
of both (the great uniformity of the Python treatment of
references is quite beneficial).

When passing any argument, the receiving function always
gets a reference to that argument, usable exactly as any
other reference to the same object would be (e.g., any
variable bound to that object, any list-item bound to
that object, any dictionary-entry bound to that object),
and so on.  If the object thus 'passed' is immutable, it
cannot be changed through the reference the function has;
if it offers methods by which it can be changed, those
methods may be called -- exactly as would be the case if
the reference was a list-item, dictionary-entry, whatever.

Re-binding the reference only affects the reference, not
the previously-bound object, nor any other reference, for
any of these kinds of references (arguments just as well
as any other).  You can't get any simpler, more uniform,
or clearer, than this.  (Calling re-binding 'assignment'
*MAY* be slightly-confusing terminology to some people).

People whose only language exposure, previous to Python
(or Java, etc) was to languages whose argument-passing
mechanisms differed greatly may be confused, I guess. But
the fact that Python (or Java) differ on this from other
languages hardly touches on their _clarity_.  Actually,
the issue is more likely to be with '=', SO-CALLED
'assignment' in Java or Python, ACTUAL assignment in
Fortran, C, C++.  The '=', in Fortran, C, or C++, will
normally "copy bits from one place to another", i.e.,
'assign' stuff.  In Java or Python, '=' _rebinds the
reference on its left-hand side so it now refers to
the result of the expression on its right-hand side_
(if any bits are copied to perform this, it will be the
few bits the language internally uses to keep track
for each reference of the issue 'to what object does
this reference refer right now' -- language-internal
issues of small import to its conceptual-model).


> (I've been programming in Python for three years now
> and I understand only that passing an int or float is call-by-value,
> otherwise it is call-by-reference.

Wrong, in terms of those languages where call-by-value
and call-by-reference have meaning (and, not coincidentally,
'=' or equivalent similar syntax DOES perform assignment).

Take C++, for example:

Foo foo(1);
Foo bar(2);

void afun(Foo& a, Foo& b)
{
    Foo temp(0);
    temp = a;
    a = b;
    b = temp;
}

afun(foo, bar);


Now THIS is call-by-reference: the assignment operators
which afun calls on its formal arguments a and b affect
the actual arguments passed, foo and bar.  Unless class
Foo redefines the assignment-operator operator= to
something VERY peculiar, this sequence will leave the
global variables 'foo' and 'bar' ``swapped'' in terms
of contents after the call to afun.

In Python, on the other hand:

foo = Foo(1)
bar = Foo(2)

def afun(a, b):
    temp = a
    a = b
    b = temp

afun(foo, bar)

all the *re-bindings* (SO-CALLED 'assignments', but
not really assignments!) that THIS afun performs on
its LOCAL VARIABLES temp, a, and b, have ABSOLUTELY
NO EFFECT WHATSOEVER on the actual-arguments foo
and bar: they are absolutely unchanged and pristine
after the call to afun, referring to exactly the
same objects to which they referred before it.

So, modeling Python's argument-passing as "by reference"
in this case (although it's an "otherwise") is simply
wrong, in terms of (for example) C++'s references (but
Visual Basic, Fortran, and other languages which DO
have pass-by-reference and actual assignment work much
like C++ does here).


> At the risk of naivete, I would
> like a simple & referencing operator so that I can explicitly,
> finally, have control over this.)

You have full control, and clarity, NOW -- if you take
the trouble to understand a little bit better how Python
really works, rather than proposing changes to it.

> When would I ever use print>> ?

In my humble opinion, never; I consider it a hateful
kludge, and its insertion in the Python language the
worst single thing that ever happened to Python in
its (almost) 10-years history.

OK, so, I guess even I (and I know few people prouder
than me) can allow the BDFL *one* terrible mistake
per 10 years or so...:-).

> Isn't there a less-ugly way of generating a raw string, instead of placing
> an 'r' in front of it?

Not that I know, though other languages have tried all
sorts of slight syntax-sugar variations (using " to
delimit non-raw, and ' to delimit raw-strings, for
example).  The r-decoration seems just fine to me.


> Good that you mentioned that the language is deteriorating. IMHO, from

I have said no such thing.  I said Python WOULD deteriorate
IF the @ special-case syntax was added to it (the only
deterioration I ever saw happen to Python was the recent
addition of the >> special-case syntax for print).


> the above, it's seemed like this from the start. (It's still my favorite
> language).

I'm really surprised that you have not taken the trouble
to grasp the references/objects semantics of Python, and
how it works uniformly and smoothly (in argument-passing
just as everywhere else), in 3 years, for what you say
it's your favourite language.

It's really a very, VERY simple idea -- though I hardly
ever see it explained as explicitly as it deserves to be,
and in fact the common idea of calling '=' "assignment"
(used in the Python docs, too) does its own little part
to obscure it.


Alex





From prestonlanders at my-deja.com  Tue Dec 26 13:48:49 2000
From: prestonlanders at my-deja.com (Preston Landers)
Date: Tue, 26 Dec 2000 18:48:49 GMT
Subject: mail processing module?
References: <91rcig$rgu$1@nnrp1.deja.com> <m266kd45m0.fsf@freak.kaiserty.com> <91vumo$di3$1@nnrp1.deja.com>
Message-ID: <92ap6g$o7n$1@nnrp1.deja.com>

In article <91vumo$di3$1 at nnrp1.deja.com>,
  Preston Landers <prestonlanders at my-deja.com> wrote:

I just want to see if people have had a chance to glance at this, the
current interface to the maillib draft and tell me a good reason why
it's horribly broken, or need some tweaking, or is just fine.  I've
gotten tenative permission to post code, but I need to sort out a few
more details.  I want to see if I can get some preliminary feedback on
the module interface.  I will be making some tweaks to make it fully
agnostic of transport mechanisms, but my basic plan is still to unify
all the lower level modules into one maillib module (with several
internal helper classes...)

thanks
Preston

> Like I said, I am trying to:
>
> a) reimplement as little as possible
>
> b) make the interface for using the module as simple as possible
>
> Here's an example of use (sorry if there are formatting problems):
>
>  import maillib
>
>  # make mail exception availible in this namespace
>  MailError = maillib.MailError
>
>  # set up a configuration dictionary with the recommended options
>
>  config_dict = {
>
>      'sendmail_from': 'Your Name <your at address.here>',
>      'sendmail_smtphost': 'smtp.yourdomain.com',
>      'sendmail_user_agent':  'Catchy Slogan Here',
>
>      'getmail_method': 'imap',  # or 'pop'
>      'getmail_user': 'uname',
>      'getmail_passwd': 'pwd',
>      'getmail_server': 'imap-server.yourdomain.com',
>      'getmail_server_port': 143, # standard imap port, 110 is POP port
>
>      }
>
>  Debug = None  # set to 1 for more verbose output to stderr
>  my_mail = maillib.maillib(config_dict, Debug)  # you now have a
maillib
> object
>
>  ### SENDING MAIL
>
>  mailto_list = ['destination at address.com']
>  subject = 'Hello, maillib!'
>  mail_body = 'Hello, world.'
>
>  attachment_list = [] # no attachments, empty list or None
>
>  ### ATTACHMENT 1
>
>  description_1 = 'Picture of Tux the Penguin'
>  filename_1 = 'tux.jpg'
>  mimetype_1 = None # let the module guess from the filename, only
works
> with common types
>  item_1 = open_file_handle.read()
>  attachment_1 = (description_1, mimetype_1, filename_1, item_1)
>
>  attachment_list.append(attachment_1)
>
>  try:
>     my_mail.send_mail(mailto_list, subject, mail_body,
attachment_list)
>  except MailError, msg:
>     print 'There was a mail sending problem'
>     print msg
>     raise
>
>  print 'Mail sent ok.'
>
>  # RECIEVE EMAIL
>
>  try:
>     my_mail.getmail_login()
>     Delete = 1 # delete the message after you fetch it
>     raw_message = my_mail.fetch_first_msg(Delete)
>     my_mail.do_cleanup()  # logs off mail server!
>  except MailError, msg:
>     print 'There was a problem retrieving the email.'
>     print msg
>
>  ### The raw unprocessed message is now in raw_message
>
>  ### DECODING MESSAGES
>
>  try:
>    (header_dict, message_body, attachments_list) =
> my_mail.decode_email(raw_message)
>  except MailError, msg:
>     print 'There was a problem decoding the email.'
>     print msg
>
>   print 'From: ', header_dict.get('From', 'unknown')
>   print 'To: ', header_dict.get('To', 'unknown')
>   print
>   print message_body
>
> --- Preston <pibble at yahoo dot com>
>
> Sent via Deja.com
> http://www.deja.com/
>

--
|| Preston Landers <prestonlanders at my-deja.com> ||


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


From slhath at home.com  Sat Dec  9 15:01:10 2000
From: slhath at home.com (Scott Hathaway)
Date: Sat, 09 Dec 2000 20:01:10 GMT
Subject: corba, soap, com, etc...
References: <pKsY5.58762$yR4.1496678@news1.rdc1.tx.home.com> <90tph1$o21$1@nnrp1.deja.com>
Message-ID: <acwY5.58954$yR4.1507182@news1.rdc1.tx.home.com>

Wow!  Thanks for the links.  I have been trying to decide between them.
Does anyone know of a comparison between them.  For instance, xml-rpc looks
very easy to use and clean.  Is it being replaced by SOAP?

Has anyone used more than one of these methods who could let me know which
they like better, etc?

Thanks,
Scott

<jerry_spicklemire at my-deja.com> wrote in message
news:90tph1$o21$1 at nnrp1.deja.com...
> Scott Hathaway wrote:
>
> > If there is a corba or soap interface to python, can you please point
> > me to it?
>
>  http://www.uk.research.att.com/omniORB/omniORBpy/
>
>  http://theopenlab.uml.edu/pyorbit/
>
>  http://casbah.org/Scarab/
>
>  http://www.pythonware.com/products/xmlrpc/
>
> All of these, and more,  are listed at "the Vaults":
>
>  http://www.vex.net/parnassus/apyllo.py/126307487.549099265
>
> Have fun!
> Jerry S.
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.




From abrahams at mediaone.net  Sat Dec 23 10:38:10 2000
From: abrahams at mediaone.net (David Abrahams)
Date: Sat, 23 Dec 2000 15:38:10 GMT
Subject: A Suggestion for Python Colon Syntax
References: <mailman.977544967.21073.python-list@python.org>
Message-ID: <CF316.26760$1t.879107@typhoon.ne.mediaone.net>

"Tim Peters" <tim.one at home.com> wrote in message
news:mailman.977544967.21073.python-list at python.org...

> 2. The NumPy people had a real problem in that
>
>     x = x + y
>
> could end up generating multimegabyte temp arrays while
>
>     x += y
>
> need not.  That is, for them it wasn't "just another way" to write the
> former; it offers crucially different semantics.

Yes! Any chance of getting this capability for extension classes?

-Dave




From aahz at panix.com  Mon Dec 25 11:34:38 2000
From: aahz at panix.com (Aahz Maruch)
Date: 25 Dec 2000 08:34:38 -0800
Subject: Season's greetings (was Re: Some Python 2.1 ideas)
References: <925tl301ub0@news1.newsguy.com> <mailman.977688488.25438.python-list@python.org> <20001225150250.6FA06A84F@darjeeling.zadka.site.co.il> <mailman.977760789.10224.python-list@python.org>
Message-ID: <927suu$r5l$1@panix2.panix.com>

In article <mailman.977760789.10224.python-list at python.org>,
Bob Alexander <bobalex at home.com> wrote:
>
>P.S. Merry Christmas!

Bah, humbug.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

'"Crisp" is a good quality for crackers; less so for pot roast.'  --pnh


From peter at engcorp.com  Fri Dec  8 02:06:41 2000
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 08 Dec 2000 02:06:41 -0500
Subject: Python legalities
References: <3A2FC63D.652101A@accessone.com>
Message-ID: <3A308881.7FBEBAD4@engcorp.com>

Helen (or maybe it was really Bruce) Dawson wrote:
> 
> If we license software from a third party, then part of what we
> are paying them for is for them to promise that there aren't any
> copyright or patent issues, and for them to take the legal heat if
> there are any issues.
> 
> With free software in general, and Python in particular, there is
> nobody to indemnify us.
> 
> Bruce Dawson
> comments at cygnus-software.com

Somehow I doubt the distinction between "free" and commercial software
is really that clear.  Is it definitely, legally, contractually the case
that you are completely safe licensing software for money from a third
party?  Isn't it possible they have inadvertently included copyrighted
material from other sources without realizing it (or deliberately), and
the copyright owner could come after you?

I'm not saying you are definitely wrong, but perhaps the legal eagles
are reading too much benefit into the difference between licensing with
money exchanged and licensing without money exchanged.

IANAL, but some of my best friends know people who are.


From sholden at holdenweb.com  Fri Dec 22 12:26:08 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 22 Dec 2000 12:26:08 -0500
Subject: __init__ like in java
References: <91vosi$19h$1@reader1.imaginet.fr>
Message-ID: <BaM06.7866$3O1.86831@e420r-atl1.usenetserver.com>

Gilles Lenfant <glenfant at equod.com.nospam> wrote in message
news:91vosi$19h$1 at reader1.imaginet.fr...
> Hi,
>
> Is there an *elegant pythonic*  way to have several constructors for one
> class ?
>
> TIA
>
>
If you mean constructor function overloading according to the call
signature, as is possible in Java and C++, then I'm afraid the answer has to
be "no".  This would have to be programmed inside the __init__() function,
as the Python language does not allow such overloading: only one definition
of a method with a given name can exist in a class definition.

regards
 Steve






From tim.one at home.com  Fri Dec 22 22:57:50 2000
From: tim.one at home.com (Tim Peters)
Date: Fri, 22 Dec 2000 22:57:50 -0500
Subject: A Suggestion for Python Colon Syntax
In-Reply-To: <9207a3$av6$1@hecate.umd.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEDDIFAA.tim.one@home.com>

[William Djaja Tjokroaminata]
> Exactly.

Good.  Let's stop there <wink>.

> I just regretted the initial study that used newbies and they
> preferred colons.  At the present time, correct me if I am wrong, I
> think it is still rare that Python is one's first programming
> language.  People who start to use Python may already use some other
> popular programming languages.  Probably if we do another survey now,
> the result may be different.  I don't think survey is a bad thing.
> Richard Stallman sometimes asked opinions on the emacs newsgroup
> before he decided whether to go one way or the other.

So does Guido, but it's rare, and in this case you're ignoring that he
*already decided*.

A survey is not a usability study, and Guido only asks people when he
doesn't care how it turns out.  For example, when complex numbers were
introduced, he asked the numeric people to pick one of "i" or "j" to
indicate the imaginary part.  He couldn't care less, and they couldn't stop
arguing about it.  They picked "i".  People complain about that a *lot* more
than they complain about the trailing colon (although both complaints add up
to "not much"), but they're not getting a choice of "j" at this stage
either.

> ...
> It seems to me that the colon is redundant, because the next-level
> indentation is required afterwards.

Yes; I agreed to that in my first msg.  It doesn't matter, though, in the
sense that Python is not about minimizing keystrokes (not even so much as
Perl is, let alone extreme one-liner languages like APL or J).  It's much
more about maximizing readability.  It may be that you're in a minority who
does not find that the trailing colon increases readability?  It increases
readability for me, and that matches the testimony of most people who've
chimed in on this issue over the years.

> Isn't it the basic premise of Python to take the best features of
> other programming languages?

It has never seemed so to me, and I'm not aware of Guido ever having said
so.  As a design principle (sometimes honored in the breach), yes, borrowing
from other languages is very apparent in Python, as it is in virtually all
language design efforts.  In this particular case, Guido borrowed the rule
from ABC because he *loved* that part of ABC; most of the rest of ABC he
dropped.  It's never been a popularity contest, though (if popularity or
"surveys" had any role to play, every language would use curly braces <0.3
wink>).

> Probably a decade ago ABC was around, but now I think many more
> people have encountered Tcl rather than ABC.

ABC was never popular.  And I'm not sure what more people being familiar
with Tcl has to do with this.  That *may* influence a decision to add
semantic features (like, say, file events) someday, but is unlikely to have
any influence on future syntax -- the languages are worlds apart
syntactically.

> I can see clearly the big difference (as I have experienced) between
> using indentation and curly braces; it makes really consistent coding
> style.  However, I don't see the colons as falling into the same
> category.

Guido does.  Now what?

> Yes, Guido wants everybody to use colons at the end of the clauses,
> but right now what prevents someone to put semicolon at the end
> of every statement?

Some people do.  When they do it public, we ridicule them.  Then they
usually stop <wink>.

>  Just compare the two or three formats:
>
>     xxxx xxxx xxxx:
>         xxxx xxxx xxxx
>         xxxx xxxx xxxx
>
>     xxxx xxxx xxxx:
>         xxxx xxxx xxxx;
>         xxxx xxxx xxxx;
>
>     xxxx xxxx xxxx
>         xxxx xxxx xxxx
>         xxxx xxxx xxxx
>
> Which one do you think is the most consistent in layout?  To me, it
> is not the first one.

The most consistent would be this:

    xxxx xxxx xxxx
    xxxx xxxx xxxx
    xxxx xxxx xxxx

That is, consistency is a red herring.

> ...
> But the semicolon at the end of the statement break the Pythonic
> rule.

If you put them there, yes.  Happily, nobody does; for example, you won't
find a semicolon at the end of any code stmt in any .py file in the standard
distribution.  When conformity is voluntarily and universally achieved,
there's no need to legislate it.

> ...
> I am not too familiar with the parsing stuff.  However, in my
> simplistic opinion, Python is not a free-form language like C
> or Perl.  Therefore, probably it is reasonable for any parser
> to breaks a Python code first into lines, even in backward
> parsing, instead of parsing it token by token first.  In parsing
> backwards, can then it just detect first that the line is at
> different indentation level rather than try to detect the colon?

When you're writing code and hit the ENTER key, good editing environments
try to *suggest* sensible indentation for the new line.  In part, that
requires guessing whether the statement just ended opens a block (in which
case the new line should be indented more) or not (in *most* of which cases
the indentation should be duplicated from the statement that just ended).

Regular expressions don't suffice for this determination, so it's a lot of
painful character-at-a-time parsing.  The trailing colon is a great hint:
if the stmt that just ended doesn't end with a colon, there's no need to
endure the expense of further analysis (the line that just ended can't
possibly open a new block without that colon).  Speed is important here,
because the user expects the response to the ENTER key to appear
instantaneous, and the parsing code is usually written in an interpreted
language.  The Emacs Python mode is greatly helped by some parsing
primitives supplied by elisp and coded in C.  IDLE (and by inheritance, also
PythonWin, which shares IDLE's auto-indent code) has a much rougher time of
it, being coded in pure Python, and having the added speed burden of needing
to talk to a Tk text widget thru the Tkinter interface layer to find out
anything about what's in the buffer (even worse, that's indirected in Python
code too, so that PythonWin can talk to the Scintilla text widget instead).

Every quirk of the syntax is exploited mercilessly to reduce processing
time, and that's a hard job (I wrote both of those parsers, so I'm not just
guessing about that); the use of colons to open blocks is one of the quirks
that can be exploited a lot.  The IDLE code would need to be redone from
scratch without it (IDLE doesn't preserve any of the alphanumeric characters
now:  it squashes all runs of alphanumerics into a single "x" character,
because it can do that quickly, and then chew over far fewer total
characters at Python speed).


At this point, if you want to keep pushing this the way to do it is to open
a PEP:

    http://python.sourceforge.net/peps/

While a PEP needs an implementation before it can become final, the PEP
author need not write the implementation, and the PEP can be accepted before
an implementation is even started.  What you would need to do is make a
compelling (to Guido) case, and identify all the consequences and how
they'll be dealt with.  The process is covered in more detail in

    http://python.sourceforge.net/peps/pep-0001.html

I expect the PEP will be rejected, but that doesn't mean it will be.  Guido
did surprise me once, in about 1995 <wink>.

or-you-could-just-resolve-to-use-python-for-a-year-before-
    improving-it-ly y'rs  - tim




From ckrill at qvlinc.com  Fri Dec  8 19:06:46 2000
From: ckrill at qvlinc.com (Coy Krill)
Date: Fri, 08 Dec 2000 16:06:46 -0800
Subject: Pickles Cross-platform?
References: <3A313570.A71966A@qvlinc.com> <p6qofymiqc7.fsf@informatik.hu-berlin.de>
Message-ID: <3A317796.D6E21AA@qvlinc.com>

Martin von Loewis wrote:

> Coy Krill <ckrill at qvlinc.com> writes:
>
> > I seem to have to run 'dos2unix' past any pickles that I want to read
> > that originated on our Windows machines. Is this typical?  For some
> > reason I was under the impression that the unpickler handled the LF vs
> > CRLF contortions for you but it doesn't seem to be doing that.
>
> There shouldn't be any CR in a pickled file; just make sure you always
> open files in binary ('b') mode.
>
> Regards,
> Martin

I either didn't see that recommendation before or I had forgotten it. I'll
change to program right now. Been so long since I had to write a program
for DOS/Windows I had forgotten about opening the thing in binary mode!  I
believe that's a good thing.

Thanks,
Coy




From amos at digicool.com  Wed Dec 27 20:23:42 2000
From: amos at digicool.com (Amos Latteier)
Date: Wed, 27 Dec 2000 17:23:42 -0800
Subject: PEP 222 draft
Message-ID: <3A4A961E.3A3054F8@digicool.com>

Thanks Andrew for getting the ball rolling on CGI improvements!

I think that having request and response classes is a good idea.
Everyone else is doing it, why shouldn't we. Here are my comments:

Your request class looks reasonable. 

I have a couple issues with your response class. The response class has
no provisions for streaming. Also it probably won't work well with very
large responses since it seems to require that the entire response body
is set and retrieved as a string (via the setBody and __str__) methods.

I'd prefer something like

  def write(self, data):
      """
      Write data to the body of the request. This method can be called
multiple times
      to stream output (assuming your web server supports streaming).
Call finish()
      when you've finished writing the body of the response. Note that
you must set
      the status and all headers and cookies before calling this method.
      """

  def finish(self):
      "Call this method to signal that you've finished with the
response."

We even might want to have a way to hand the response a file object to
represent the body. We also might want to let the user control buffering
of the response in some way, though I kind of doubt it.

Also I don't see the use of the '__str__' method on the response object.
In Zope this is used by ZPublisher to get the value of the response and
send it back to the web server. I'd rather have this handled internally
by the response. It would be triggered by a call to the 'finish' method.
If you need to see the response guts you already have the 'dump' method.
Also as I mentioned above the '__str__' method does not scale well to
very large responses.

Finally a nit, we probably want an 'addHeader' as well as a 'setHeader'
method since some headers can have multiple values.

-Amos

--
Amos Latteier         mailto:amos at digicool.com
Digital Creations     http://www.digicool.com



From dsh8290 at rit.edu  Fri Dec 22 14:19:55 2000
From: dsh8290 at rit.edu (D-Man)
Date: Fri, 22 Dec 2000 14:19:55 -0500
Subject: A Suggestion for Python Colon Syntax
In-Reply-To: <9207a3$av6$1@hecate.umd.edu>; from billtj@y.glue.umd.edu on Fri, Dec 22, 2000 at 06:42:11PM +0000
References: <mailman.977446446.16022.python-list@python.org> <9207a3$av6$1@hecate.umd.edu>
Message-ID: <20001222141955.A25146@westsidecnc.rh.rit.edu>

On Fri, Dec 22, 2000 at 06:42:11PM +0000, William Djaja Tjokroaminata wrote:

<snip>

> Exactly.  I just regretted the initial study that used newbies and they
> preferred colons.  At the present time, correct me if I am wrong, I think
> it is still rare that Python is one's first programming language.  

I think you are right,  most people have never heard of Python.  It's
really too bad because python is such a nice language and much easier
to learn as a first language than many others.


> I am not too familiar with the parsing stuff.  However, in my simplistic
> opinion, Python is not a free-form language like C or Perl.  Therefore,
> probably it is reasonable for any parser to breaks a Python code first
> into lines, even in backward parsing, instead of parsing it token by
> token first.  In parsing backwards, can then it just detect first that the
> line is at different indentation level rather than try to detect the
> colon?

If the parser (I believe you are referring to someone's emacs mode
thing) used the indentation level to infer the structure, it would
totally defeat the purpose.  The purpose of that parser is with the
following input:

if 1 :<enter>print "hello"

To give the following result in the editor:

if 1:
	print "hello"


I use vim myself and rather like this feature in C/C++/Java modes.  It
also handles the unindent when a } is typed.


I for one rather like having the colon and I have more experience
using C++, C, and Java than Python.

-D

BTW, A python mode for vim would be a very cool thing.  I also have a
habit of using a comment to identify the end of blocks (so it is
obvious what the brace is for). 

Ex in C I would write :

	if ( cond )
	{
	} /* if */

In C++/Java I use:
	if ( cond )
	{
	} // if

In Python I write:
	if ( cond) :
		pass
	# end if


It would be cool if the editor would recognize the "# end" like it
currently does for "}"



From S.Tian at shu.ac.uk  Wed Dec  6 16:24:08 2000
From: S.Tian at shu.ac.uk (Shicheng)
Date: Wed, 06 Dec 2000 21:24:08 GMT
Subject: sendmail: under Windows NT/2000
Message-ID: <90mapg$u7m$1@nnrp1.deja.com>

Hello all,
I test a python cgi sendmail program; in the code
there are the following lines:

...
sendmail_cmd = "/usr/lib/sendmail -oi "
...
try: mfd = os.popen(sendmail_cmd + real_to, "w")
...

I think the string "/usr/lib/sendmail -oi " suggests
that the code should be used under the UNIX platform.

My query is: under Windows NT/2000, how could I change the
above line of code so that I could make use of the NT/2000
mailing facility in the code?

Thanks,

Shicheng


Sent via Deja.com http://www.deja.com/
Before you buy.


From franzja at anubis.network.com  Mon Dec  4 15:22:42 2000
From: franzja at anubis.network.com (Joseph A. Franzmeier)
Date: Mon, 04 Dec 2000 14:22:42 -0600
Subject: Programming with Python
References: <90fq39$jgq$1@nnrp1.deja.com> <9iOW5.55847$L26.16633720@typhoon.southeast.rr.com> <90gj77$7a0$1@nnrp1.deja.com>
Message-ID: <3A2BFD12.7AB3A628@anubis.network.com>


A book talked of in Linux Journal is "The Quick Python Book"  by Daryl
Harms
ISBN#  1884777740
GREAT BOOK TO LEARN FROM,
 and maybe  THE best on the market today.

The other to recommend is  "The Python Reference"  by  Guido Van Rousson
the creater of Python.

"Simon B." wrote:

> In article <9iOW5.55847$L26.16633720 at typhoon.southeast.rr.com>,
>   "Don Tuttle" <tuttledon at hotmail.com> wrote:
> > >
> > > Any good?
> >
> > IMO, this is the only bad Python book on the market.  Don't waste your
> > money.
>

snip

--

|    StorageTek    |
| cout << C_ya << later << Joe;


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20001204/445bfd54/attachment.html>

From dealfaro at eecs.berkeley.edu  Sat Dec 16 06:05:00 2000
From: dealfaro at eecs.berkeley.edu (Luca de Alfaro)
Date: Sat, 16 Dec 2000 03:05:00 -0800
Subject: Alarms in threads - how to get/mimic?
Message-ID: <Pine.LNX.4.21.0012160248520.7483-100000@traal.eecs.berkeley.edu>

I am trying to write a multi-threaded program to read web pages.  The
program spawns (say) 10 threads, assigning to each thread a file to
read with the ftp or http protocol; each thread reads it, and returns
the info.

The problem is that the threads sometimes hang for a very long time
waiting for a connection (httplib seems to have very long timeouts??),
whereas instead I don't care for the answer if it comes later than
(say) a few seconds.

The obvious solution would be to have each thread set an alarm before
trying the url read operation.  However, alarms cannot be used in
threads in Python (they can only be used in the main thread). 

The next solution would be to set an alarm in the main thread, and if
the alarm goes off while waiting for some subthreads to be join-ed,
kill the laggard threads that can't be joined.  This is fine, except
that (a) I cannot kill subthreads in Python, and (b) I don't want to
risk killing a thread at an arbitrary point (it can cause
inconsistencies), but only while it's waiting for a connection.

Have I reached the limits of Python?  Should I be coding in Java
instead?  Please, tell me it's not true, and that there's an obvious
solution...  I like Python much better than Java otherwise...

Thanks, 

Luca de Alfaro




From aleaxit at yahoo.com  Thu Dec 21 07:07:04 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 21 Dec 2000 13:07:04 +0100
Subject: Python's scope rules
References: <3A40813B.C04D21C@et.dtu.dk> <91s38r$sjd$1@news.udel.edu> <slrn943bf3.q3c.qrczak@qrnik.zagroda>
Message-ID: <91srpu0262q@news2.newsguy.com>

"Marcin 'Qrczak' Kowalczyk" <qrczak at knm.org.pl> wrote in message
news:slrn943bf3.q3c.qrczak at qrnik.zagroda...
> Wed, 20 Dec 2000 12:47:09 -0500, Terry Reedy <tjreedy at udel.edu> pisze:
>
> > The only things I find 'odd' (and definitely confusing at first)
> > is the use of 'global' for the module namespace (and I suppose
> > there is some history here) and the sometimes claim that there are
> > two rather than three possible scopes.
>
> The fact that scopes don't nest is a practical problem. Yesterday I was
> doing some customizations to viewcvs and wanted to factor out a common
> pattern of code of 5 lines which was repeated 5 times by moving it to
> a local function, but it referred to so many free local variables that
> I finally did not do it, because I would have to pass them explicitly
> and the result would not be much less ugly than the current state.

When I needed to do a similar refactoring I came up with a simple
solution based on 'explicit is better than implicit'.

Original code, say (just typing it in more or less at random!-):

def original(anargument, another, yetanother, onemore):
    alocal=whatever(anargument,onemore)
    if onemore>alocal:
        somethingelse=acleverfun
    else:
        somethingelse=asimplerfun
    for i, j in zip(another,yetanother):
        if i>alocal or j<anargument:
            onemore.peep(j,i)
        else:
            somethingelse(onemore,j,i)
    another.reverse()
    for i, j in zip(another,yetanother):
        if i>alocal or j<anargument:
            onemore.peep(j,i)
        else:
            somethingelse(onemore,j,i)
    another.reverse()
    yetanother.reverse()
    for i, j in zip(another,yetanother):
        if i>alocal or j<anargument:
            onemore.peep(j,i)
        else:
            somethingelse(onemore,j,i)
    another.reverse()
    for i, j in zip(another,yetanother):
        if i>alocal or j<anargument:
            onemore.peep(j,i)
        else:
            somethingelse(onemore,j,i)

How do I factor out the loop into a local function without
having to pass it half a dozen argument, is the issue.

A simple and reasonably explicit answer might then be:

def factored(anargument, another, yetanother, onemore):
    def local_loop():
        for i, j in zip(another,yetanother):
            if i>alocal or j<anargument:
                onemore.peep(j,i)
            else:
                somethingelse(onemore,j,i)
    def runit(func, dict):
        exec func.func_code in dict

    alocal=whatever(anargument,onemore)
    if onemore>alocal:
        somethingelse=acleverfun
    else:
        somethingelse=asimplerfun
    dict = locals()

    runit(local_loop, dict)
    another.reverse()
    runit(local_loop, dict)
    another.reverse()
    yetanother.reverse()
    runit(local_loop, dict)
    another.reverse()
    runit(local_loop, dict)


This kind of approach only works for a "piece of code
to be factored out" that doesn't rebind local variables,
of course; local variables can only be rebound by
assignment statements in local scope.  For a more
general case, I think it might be best to introduce
an instance, holding, as its attributes, the relevant
'variables', finessing scope issues entirely through
more explicitness.

The issue of rebinding non-local variables may well
persist even when local scopes are allowed to nest,
depending on difficult design choices regarding how
one differentiates between (re-)binding a variable in
an 'outer' scope, and causing a variable thus named
to 'spring into existence' in _this_ (local) scope.

http://python.sourceforge.net/peps/pep-0227.html
seems to be strongly oriented to simply disallowing
rebinding of variables in outer nested scopes, and
it seems that's what we'll see in Python 2.1.  If so,
then factoring-out will still need the introduction
of a state-holding instance (or dirtier 'container'
tricks as PEP 227 suggests) when the factored-out
code needs to rebind (what used to be) locals; but
things will be simpler when no such rebinding occurs.


> I don't want dynamic scoping like in old Lisp and probably nobody wants
> it. But nested lexical scoping is used in other good languages, it's
> so natural and convenient that it's odd that Python does not use it.
> IMHO Python would be a better language with true lexical scoping.

I think you can say 'will' rather than 'would', as I see no
reason why PEP 227 should not make it into Python 2.1.

I do hope that Python keeps the simplicity of avoiding the
distinction between 'let' and 'set!' (creation of a new
binding in this scope, vs modification of existing binding
in some nested scope), variable-declarations, or other
complexities to allow fancy stuff such as rebinding of
lexically-nested-scoped variables.  I know of no language
that allows that without constraints -- say that your
scope is currently nested 5 levels deep and there's a
binding for 'x' in each nested scope, how do you specify
'the x that's bound exactly 3 levels up from this one'?

Java specifically disallows homonym variables in nested
scopes, and maybe that should be considered as a part of
PEP 227 -- such homonimy frequently causes confusion in
languages which allow it, I think, and apparently Java's
designers agree.


Alex





From emile at fenx.com  Wed Dec  6 09:04:35 2000
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 06 Dec 2000 06:04:35 -0800
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <3A2DD2CA.14A5D7E2@engcorp.com> <90l3j9$4bn$1@kapsel.intranet>
Message-ID: <3A2E4773.9010605@fenx.com>


Piotr Kuchta wrote:

> Peter Hansen <peter at engcorp.com> wrote:
<snip> 
>> The headquarters of the World Esperanto Association is headquartered in
>> Rotterdam, Netherlands.  As it happens Python's creator, Guido van
>> Rossum, is also from the Netherlands, where his exposure to Esperanto
>> (still widely spoken in that country) led him to attempt to bring the
> 
> 
> I didn't noticed people speaking Esperanto in Amsterdam ;), but that's true

Well, there's a reason for that!  Dig into the history of Feyenoord and 
Ajax and I'm sure you'll see it.

;-)

Emile van Sebille
emile at fenx.com



From loewis at informatik.hu-berlin.de  Fri Dec  8 15:51:55 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 08 Dec 2000 21:51:55 +0100
Subject: xml.dom
References: <90q292$k9k$1@merlin.rembrandtstr.bocholt.de> <3A3100FD.D6C09C1D@ogbuji.net> <90raqe$2f6$1@merlin.rembrandtstr.bocholt.de>
Message-ID: <p6qwvdaiql0.fsf@informatik.hu-berlin.de>

Magnus Heino <magnus.heino at rivermen.se> writes:

> Maybe you could tell me the differences and relations between the xml 
> included in Python2.0, PyXML and Fourthoughs distro?

Python 2.0 supports minidom; Uche's example was a 4DOM application.
4DOM is included both in PyXML and 4Suite.

Please have a look at

http://python.sourceforge.net/devel-docs/lib/markup.html

for documentation on minidom.

> Seems to me that the Python2.0 is a subset of PyXML (because it wasnt 
> finished in time for the release maybe?) and that PyXML includes things 
> from Fourthught, but not of all of it.. for some reason.

It's not because it was not finished, it was not included because it
was considered too large.

> Are there any good books to recommend in this area? 

If you can read German, I can recommend the "Python 2" book that I
just published at Addison-Wesley :-)

Regards,
Martin


From darrell at dorb.com  Tue Dec 12 13:18:58 2000
From: darrell at dorb.com (Darrell Gallion)
Date: Tue, 12 Dec 2000 13:18:58 -0500
Subject: Fault trees structure
References: <D_IY5.49$17.2785@m2newsread.uni2.es> <9128e902uo3@news2.newsguy.com> <z8eZ5.249$YZ4.8868@m2newsread.uni2.es> <yfhZ5.23310$Ia1.4003773@typhoon.nyroc.rr.com> <91521d02r3g@news2.newsguy.com>
Message-ID: <915pu2$6gf$1@news.kodak.com>

Alex Martelli wrote:
> def connectInput(self, name):
>     self.__dict__.setdefault('_input',{})[name] = None
>
> is a possible alternative (also in other cases), but
> not very readable.  It would seem best to just set
> empty dictionaries in __init__ for _input and _output,
> so as to avoid having test for them at every connect,
> and possibly having errors lurking elsewhere, too:
>
Sounds good.
This wasn't required once I moved the __init__ into NodeBase by your
suggestion.

[snip]
>
> Of course, giving NodeBase an __init__ means having
> to call it in lieu, or from, specific subclasses'
> ones, but that seems OK; right now OrNode and
> AndNode sport identical __init__ methods, so why
> not factor them out of the subclasses and up into
> the NodeBase common baseclass.
>
Done.

>
>
> >     def logic(self, collection):
> >         cc= reduce(lambda x,y: x and y, self._input.values())
> >         self.propagate(collection, cc)
>
> If the values are constrained to "boolean" (0 and 1), then
> using operator.and_ as the first argument to reduce might
> well be faster, and perhaps more readable (similarly for
> or-nodes).
>
Not sure that values will be constrained to boolean.

When I tried this something odd happened.
None isn't valid for and_ while it is for "and"

>>> and_(1, 1)
1
>>> and_(None, 1)
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: bad operand type(s) for &
>>> None and 1
>>>

Made some other changes as well.
Thanks for taking the time to look this over.
Maybe I'll have find a use for this code :)

--Darrell


class NodeBase:
    def __init__(self, name, logicFunc):
        self._name=name
        self._previousOutputValue=None
        self._logicFunc=logicFunc
        self._output={}
        self._input={}

    def name(self):
        return self._name

    def connectInput(self, name):
        self._input[name] = None

    def connectOutput(self, otherNode, name):
        self._output[otherNode.name()] = name
        otherNode.connectInput(name)

    def signal(self, collection, name, value=1):
        self._input[name]=value
        self.logic(collection)

    def propagate(self, collection, value):
        if value==self._previousOutputValue:
            # No point in propagating the same value
            return
        for nodeName, inputName in self._output.items():
            try:
                node=getattr(collection, nodeName)
                node.signal(collection, inputName, value)
            except:
                print self._name
                print nodeName
                print collection
                raise

    def logic(self, collection):
        cc= self._logicFunc(self, collection)
        self.propagate(collection, cc)

def andFunc(self, collection):
        return reduce(lambda x,y: x and y, self._input.values())

def announceFunc(self, collection):
        cc= reduce(lambda x,y: x or y, self._input.values())
        if cc:
            print self._name, self._input
        else:
            # Noise on the input, but nothing to report
            print '.',
        return cc

def sourceFunc(self, collection):
    assert(0)

class SourceNode(NodeBase):
    def __init__(self, name):
        NodeBase.__init__(self, name, sourceFunc)

    def __call__(self, collection, value):
        self.propagate(collection, value)


class FaultTree:#(UserDict.UserDict): Not sure about this
    def __init__(self, *nodes):
        for n in nodes:
            setattr(self, n.name(), n)

ft=FaultTree(NodeBase("and1", andFunc), NodeBase("and2", andFunc),
NodeBase("and3", andFunc),
            NodeBase("fail", announceFunc))

ft.and1.connectOutput(ft.and3,"one")
ft.and2.connectOutput(ft.and3,"two")
ft.and3.connectOutput(ft.fail,"who cares")

    # Setup some signal sources
drivers=[]
for x in range(3):
    s1=SourceNode(x)
    drivers.append(s1)
    s1.connectOutput(ft.and1, x)

    # Setup some signal sources
for x in range(3):
    s1=SourceNode(x)
    drivers.append(s1)
    s1.connectOutput(ft.and2, x)

    # Make it go
for i in drivers:
    i(ft, 1)






From michael at stroeder.com  Fri Dec 29 07:55:07 2000
From: michael at stroeder.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Fri, 29 Dec 2000 13:55:07 +0100
Subject: Python vs. PHP (& Java?)
References: <69BC4C6AE83ED311BC9400805FA7B14AF84D2C@il93exp01.css.mot.com> <mailman.978059571.14515.python-list@python.org>
Message-ID: <3A4C89AA.7E741D91@stroeder.com>

Moshe Zadka wrote:
> 
> On Thu, 28 Dec 2000, Kemp Randy-W18971 <Randy.L.Kemp at motorola.com> wrote:
> 
> > 2.  Could we argue that an artificial language, such as Esperanto, is better
> >     to use around the world then a natural language, like English?
> 
> Yes. I think this would be wonderful, except political constraints dictate
> otherwise.

<OFFTOPIC>
Esperanto is pretty much derived from romanian languages. Rather
hard to learn for many people. How about http://www.lojban.org ?
</OFFTOPIC>

Ciao, Michael.


From dale at out-think.NOSPAMco.uk  Mon Dec  4 08:09:32 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Mon, 04 Dec 2000 13:09:32 +0000
Subject: SAP R/3 automation with win32com
References: <90flpu$gpu$1@nnrp1.deja.com>
Message-ID: <kl5n2tsfk3hud1c37gjo56290sueccuegk@4ax.com>

obou at my-deja.com wrote:

>Hello!
>
>I'm trying to do SAP R/3 automation with win32com using the SAP DCOM
>Connector (a tool that presents SAP Business objects as COM Components).
>The business objects (BAPI) are kind of objects with parameters and
>methods.
>Everything works fine, except the function call (the most important
>part of course), and it could be because of the way I have to give the
>different parameters to the BAPI method. I have a VB example and don't
>know if the way I convert the VB function call to python function call
>is correct.
>Maybe someone who knows both VB and Python could help me.
>
>Here's the Python code:
>
>from win32com.client import Dispatch
>
># Create a connection object
>sapConn = Dispatch('SAP.CCSession.1')
>
># Insert connection informations
>syst = raw_input('System :')
>uname = raw_input('Username: ')
>pwd = raw_input('Password: ')
>client = raw_input('Client: ')
>sapConn.PutSessionInfo(syst, uname, pwd, 'EN', client)
>
># Log in
>sapConn.Logon()
>
># Create a sales order business object
>salesOrd = sapConn.CreateInstance('SAP.SalesOrder.1')
>
># Create the parameters for calling the method CreateFromData
>header = salesOrd.DimAs('BapiCreateFromData', 'OrderHeaderIn')
>itemsIn = salesOrd.DimAs('BapiCreateFromData', 'OrderItemsIn')
>partners = salesOrd.DimAs('BapiCreateFromData', 'OrderPartners')
>ret = salesOrd.DimAs('BapiCreateFromData', 'Return')
>
># Initialize structure with data.
>partners.AddNew()
>partners.Fields("Partn_Role").Value = 'AG'
>partners.Fields("Partn_Numb").Value = '0000001234'
>partners.Update()
>
># Method call. Everything works fine up to this call!!!
>salesOrd.BapiCreateFromData(OrderHeaderIn = header,
>                            OrderItemsIn = itemsIn,
>                            OrderPartners = partners,
>                            Return = ret)
>
>In VB this function call looks like this:
>salesOrd.BapiCreateFromData OrderHeaderIn:=header,_
>                          OrderItemsIn:=itemsIn,_
>                          OrderPartners:=partners,_
>                          Return:=ret
>
>I tried it with or without the parameter names, with or without the
>return value in the parameters list, I always get this exception:
>
>Traceback (most recent call last):
>  File "d:\python20\pythonwin\pywin\framework\scriptutils.py", line
>301, in RunScript
>    exec codeObject in __main__.__dict__
>  File "d:\sapconn.py", line 35, in ?
>    Return = ret)
>  File "<COMObject CreateInstance>", line 2, in BapiCreateFromData
>  File "d:\python20\win32com\client\dynamic.py", line 234, in
>_ApplyTypes_
>    result = apply(self._oleobj_.InvokeTypes, (dispid, LCID, wFlags,
>retType, argTypes) + args)
>com_error: (-2147352567, 'Exception occurred.', (0, None, "Table
>doesn't contain 1 record\015\012(COM_PARAMETER_NO_SINGLE_RECORD): error
>code 16388", None, 0, -2147467260), None)
>
>Thank you for your assistance,
>
>Olivier Boudry
>
>
>
>Sent via Deja.com http://www.deja.com/
>Before you buy.

I can't see anything obviously wrong so there's a chance that the error might be in the preceding
code. Have you tried running through this interactively and checking that the objects you are
creating all look OK?

Also, do you know to what table the error refers? 
--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants




From tbryan at python.net  Tue Dec 26 02:39:37 2000
From: tbryan at python.net (Thomas A. Bryan)
Date: Tue, 26 Dec 2000 02:39:37 -0500
Subject: ~Python ?
References: <3a4833f8.346848@news.t-online.de>
Message-ID: <3A484B39.A16C0B99@python.net>

Gerson Kurz wrote:
> 
> On Linux, ~ usually refers to the users homedir, but not so in Python
> (at least, in ActivePython 2.0). I found out the hard way when using a
> script that issued os.makedirs() relative to ~ - It created an actual
> directory named ~. You can have real fun if you try to deleting an
> actual directory named ~ on Linux, boy.
> 
> I understand that ~ is probably handled by the shell, but if Python
> doesn't deal with it, its hard to use Python in some kinds of scripts.
> (I think Perl, Pythons ugly twin-brother, does)

Wouldn't os.environ['HOME'] be an acceptable way of handling this problem?
I'm not sure whether such a variable is available under other OSs, such as 
Windows, but at least environment variables are available under Windows.  
I think that '~' is a foreign concept to non-UNIX users.  (ActivePython is 
a Windows distribution of Python, right?)

---Tom


From aleaxit at yahoo.com  Mon Dec 11 16:52:46 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 11 Dec 2000 22:52:46 +0100
Subject: winreg
References: <sh9Z5.6204$bw.481106@news.flash.net>
Message-ID: <913ifd018tv@news2.newsguy.com>

"Scott Hathaway" <slhath at home.com.nospam> wrote in message
news:sh9Z5.6204$bw.481106 at news.flash.net...
> Hello Everyone,
>
> I was looking at the online documentation on the winreg module that allows
> editing of the windows registry.  I am too new to python to understand how
> to code it.  Can someone get me started in the right direction?
>
> I need to add a key and then add 5 values to that key.

With Python 2, for example:

import _winreg as w

parentKey = w.OpenKey(w.HKEY_CLASSES_ROOT,r'file\shell',0,w.KEY_WRITE)
newKey = w.CreateKey(parentKey,'SubkeyName')
w.SetValueEx(newKey,'AValueName1',0,w.REG_SZ,'Something here')
w.SetValueEx(newKey,'AValueName2',0,w.REG_SZ,'Something else')

will create a key 'file\shell\SubkeyName' (assuming 'file' and 'file\shell'
already exist!) under HKEY_CLASSES_ROOT, with no default (unnamed)
value, and two named values under it.


Alex





From Daniel.Kinnaer at AdValvas.be  Wed Dec 27 05:54:25 2000
From: Daniel.Kinnaer at AdValvas.be (Daniel)
Date: Wed, 27 Dec 2000 10:54:25 GMT
Subject: Just starting out!
References: <92buip$kcp$1@nnrp1.deja.com> <u5i26.194888$U46.5994443@news1.sttls1.wa.home.com>
Message-ID: <3a49c756.8528401@news.skynet.be>

On Wed, 27 Dec 2000 08:52:42 GMT, "Greg Jorgensen" <gregj at pobox.com>
wrote:

>Learning Python assumes some programming experience. Check out the free
>tutorials at:
>
>http://www.phrantic.com/scoop/tocpyth.htm
>
>(Java and XML tutorials also available on the same site).
>
>Look at the book "Python Essential Reference" by David Beazley (New Riders).
>It contains an excellent reference (for Python 1.5.2) and a terse but good
>tutorial. The other Python books I've seen are no better than
>freely-available web resources.
>
>Reading and studying code until you understand it, and experimenting with
>your own code, teaches programming better than books and tutorials. This
>newsgroup makes a good study hall.
>
>--
>Greg Jorgensen
>Deschooling Society
>Portland, Oregon, USA
>gregj at pobox.com
>
>
Where can one find a tutorial with lots and lots of example code?
Reading about theory is fine when one already is proficient at a
language... I learn fastest when examples are given, and that is what
I'm looking for. Is there such a book/tutorial available?

As a newbie to Python, I always compare with Delphi and how I would go
about solving a problem.  Very often, I can't find an answer in the
ref-guide nor in the tutorials (or I am looking at the wrong places
:-)  E.g.  I was trying to convert a number to a string and add
another string to it. In Delphi i would write s:= IntToStr(myNumber) +
' is my number';  How is this done in Python? 

regards,
Daniel


From tanya at ct.marconimed.com  Mon Dec 11 16:51:14 2000
From: tanya at ct.marconimed.com (tanya)
Date: Mon, 11 Dec 2000 21:51:14 GMT
Subject: Python support for MIDL RPC
Message-ID: <3A354BEE.7D5C9895@ct.marconimed.com>

I need to implement DCE RPC Server/Client in Python using TCP/IP to
communicate with NT peers.
What Python has to support it? Do I have to write the
interface/marshaling stuff myself? (Hope not...???)

Thans.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: tanya.vcf
Type: text/x-vcard
Size: 241 bytes
Desc: Card for tanya
URL: <http://mail.python.org/pipermail/python-list/attachments/20001211/d34321c8/attachment.vcf>

From rhamorim at my-deja.com  Thu Dec  7 08:47:11 2000
From: rhamorim at my-deja.com (rhamorim at my-deja.com)
Date: Thu, 07 Dec 2000 13:47:11 GMT
Subject: Pmw.CounterDialog
References: <90g351$prk$1@nnrp1.deja.com> <anbX5.16289$R77.1305391@nnrp4.clara.net>
Message-ID: <90o4ct$b66$1@nnrp1.deja.com>

I'm glad it worked. If you have any further problems, just tell me.

Regards,

Roberto Amorim

In article <anbX5.16289$R77.1305391 at nnrp4.clara.net>,
  "Richard" <richard at NOstarfighterSPAM.freeuk.com> wrote:
> Many thanks for your solution to this problem, I'd given up hope of
getting
> a reply :-)
>
> Richard


Sent via Deja.com http://www.deja.com/
Before you buy.


From aleaxit at yahoo.com  Fri Dec 15 07:16:13 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 15 Dec 2000 13:16:13 +0100
Subject: Solution: Direct access to Printer I/O lines
References: <917k9b$9i9$1@nnrp1.deja.com> <slrn93hrjn.2vp.ssthapa@localhost.localdomain> <91cnp3$g59$1@nnrp1.deja.com>
Message-ID: <91d23f0jk7@news1.newsguy.com>

<jkndeja at my-deja.com> wrote in message news:91cnp3$g59$1 at nnrp1.deja.com...
    [snip]
> I have the following thoughts about the interface, and would be
> interested in any comments:

Sure!

> - The module implements a class, say directlpt
> - The module only allows three 'Singleton' instances of the class,
> corresponding to LPT1--3. This is set up in the constructor:
>
> import directlpt
>
> l1 = directlpt.directlpt(1) # gives access to LPT1:
>
> anotherl1 = directlpt.directlpt(1) # will raise an exception - only one
> instance allowed

This gives me a good chance for one of my favourite rants,
'"Singleton" design-pattern considered harmful'!-)  But
'Singleton' is not being ideally applied, so I'll skip that.

Why given an error when the object mapping the same LPTx: is
asked-for twice?  Just return what is, functionally, "the same"
object -- whether the id() is the same or not is quite a
secondary issue, so, if you want to keep directlpt as a
class, that's OK too -- the id() will differ [you can't
change what's returned by "calling the class", so a new
object will come up each time], but that's OK as long as
'different' object for the same printer-port share all
state among them.  Or else, make directlpt.directlpt into
a factory function, returning an object of class (or type)
'_directlpt', say, in which case you may be able to also
get the same identities.

Either of these will work better than constraining different
client-code modules to coordinate their generation and
release of directlpt objects.  I think the pattern based
on different-identities/same-state has advantages; for
example, it makes it possible (though it may still be
tricky) for client-code to _inherit_ from directlpt:

class PrettyLpt(directlpt):
    def __init__(self, n, doodad):
        directlpt.__init__(self, n)
        self.doodad = doodad
    # etc, etc

If you do want to allow that, more or less unrestrainedly,
you'll have to take some care to avoid name clashes (but
starting directlpt's attribute names with '__' will do:-).


Alex





From brent at rcfile.org  Tue Dec 12 10:12:01 2000
From: brent at rcfile.org (Brent Verner)
Date: Tue, 12 Dec 2000 10:12:01 -0500
Subject: Preliminary python2 packages for Debian woody
In-Reply-To: <20001212153400.A24099@mediasupervision.de>; from gregor@mediasupervision.de on Tue, Dec 12, 2000 at 03:34:00PM +0100
References: <20001212153400.A24099@mediasupervision.de>
Message-ID: <20001212101201.A15620@rcfile.org>

On 12 Dec 2000 at 15:34 (+0100), Gregor Hoffleit wrote:
| Hi,
| 
| I completed a set of preliminary packages of python2 for Debian woody. A big
| warning: They are nearly completely untested, and they aren't yet suitable
| for release. Please provide me with feedback so that the final packages will
| be more stable.

cool.

[ apt line snipped ]

| Noteworthy things:
| 
| - The packages are called python2-base, python2-dev etc.pp.
| 
| - They won't replace the existing python packages (Python 1.5.2), they can
|   be installed alongside each other.

$ apt-get update
$ apt-get install python2-base

hmm, that wanted to remove most every other python (1.5.2-10) package on
my freshly updated unstable box. Any suggestions?

thanks.
  Brent




From aleaxit at yahoo.com  Fri Dec 15 06:59:54 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 15 Dec 2000 12:59:54 +0100
Subject: Getting a code obj's function?
References: <91co73$fpm$1@news.rz.uni-karlsruhe.de>
Message-ID: <91d14u0ilb@news1.newsguy.com>

"Bjoern Giesler" <un4e at rzstud1.rz.uni-karlsruhe.de> wrote in message
news:91co73$fpm$1 at news.rz.uni-karlsruhe.de...
> Hi,
>
> another inane question...
>
> ...is there a standard way to get the function object that a code object
is
> associated with? On exception, I only get the code object from the
> traceback.

Right.  It might not even belong to any function...


> If there is no standard way, could somebody please comment on the
following:
> I've modified my Python interpreter so that a code object has one more
> member, co_func. This gets set to None initially, and PyFunc_New() sets it
> to the new function object that is created. What are the potential
problems
> with this approach, if any? I see all kinds of memory leaks / double frees
> coming, but haven't really enough understanding of Python to know what
might

Circular reference -- if object F holds a reference to object C and
viceversa, neither is ever going to go away unless/until this is
explicitly broken (Python 2's GC may ameliorate this problem, though).

> happen. My assumption was that a code object is assigned to a maximum of
> exactly one function. Is this correct?

No; the func_code attribute is not read-only:

D:\PySym>python
Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
>>> def f():
...   print 'f'
...
>>> def g():
...   print 'g'
...
>>> f()
f
>>> g()
g
>>> f.func_code = g.func_code
>>> f()
g
>>> g()
g
>>> ^Z

so a given code-object may belong to any number of functions -- from
0 on up.  (The 'new' module also lets you build a function using any
existing code-object).  Of course, this is not 'typical'!-)


Whenever you want to follow some existing reference 'backwards', you
are faced with several possible choices.  One is the one you've taken,
introducing a backwards-link, which, however, creates circularity
(which may or may not be a problem in a given case).  Another is
to use a 'weak' reference of some kind for the backwards-link (only
advantage, avoiding the unwontedly-keep-alive circularity).  These
choices share a weakness if the 'existing reference' are many-to-1,
as, 'in theory', is the F->C one in Python.

Depending on 'how _often_' you need to follow the back-way, a
practicable choice may be just to make sure you can always enumerate
all objects which 'might' be the one you're looking for -- e.g.,
all function-objects, in this case; the 'back-link' is then an
exhaustive search over all candidate objects, noting the ones
whose (forward) links satisfy your condition (a set with 0 to
whatever elements).  If enumerating _all_ candidates is not an
acceptable performance-hit, this strategy may sometimes be
tweaked to make performance satisfactory, e.g. by keeping the
objects in suitable 'bins' (you'd have to catch assignments to
func_code to change a function's bin -- not sure how easy or
hard that would be).


Alex





From thomas at xs4all.net  Thu Dec 28 06:57:37 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Thu, 28 Dec 2000 12:57:37 +0100
Subject: Is this a BUG?
In-Reply-To: <3A4B0C5E.1E7976D2@wielun.dhs.org>; from qdlaty@wielun.dhs.org on Thu, Dec 28, 2000 at 10:48:14AM +0100
References: <3A4B0C5E.1E7976D2@wielun.dhs.org>
Message-ID: <20001228125736.A1811@xs4all.nl>

On Thu, Dec 28, 2000 at 10:48:14AM +0100, QdlatY wrote:
> Hello!
> 
> I have script:
> ----------------
> import cgi
> 
> form = cgi.FieldStorage()
> 
> print form
> print "!",form["index"].value,"!"
> ------------------
> I launched it with parameter index=1
> And this is an output:
> ----------------
> FieldStorage(None, None, [MiniFieldStorage('index', '1')])
> ! 1 !
> ----------------
>  ^ ^
> Why the hell there are spaces? index property has value '1'!
> not ' 1 '?!

>>> print "!",1,"!"
! 1 !

It's the space inherit in the system. (See the documentation for 'print' --
it's supposed to do that :) You have to forcibly oppress it to avoid the
space:

>>> print "!%s!"%1
!1!

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From jepler at inetnebr.com  Thu Dec  7 09:09:43 2000
From: jepler at inetnebr.com (Jeff Epler)
Date: Thu, 07 Dec 2000 14:09:43 GMT
Subject: PyHTML -- Indentation-structured HTML generator
Message-ID: <slrn92v6h6.6vl.jepler@potty.housenet>

PyHTML -- Indentation-structured HTML generator with all the power of Python
					    Jeff Epler <jepler at inetnebr.com>
							     6 December 2000

PyHTML is an extension of the Python grammar, plus a modified compiler
which permits the use of HTML-style tags as Python blocks, making it easy
to write HTML document generators with the block-indentation style of Python.

This is a rough-cut version of pyhtml, intended to get a reaction from
the Python community.  Please let me know your thoughts, either in the
comp.lang.python newsgroup or via direct mail.  See INSTALL for instructions
on getting PyHTML running.

Thanks to the authors of Quixote: A.M. Kuchling, Neil Schemenauer,
and Greg Ward for both the inspiration of an augmented Python language
for HTML generation and for some of the code in that fine package.

Thanks to everyone who has been involved with the development of Python.

You can download pyhtml or browse its source at
	http://incolor.inetnebr.com/jepler/pyhtml/

Example:
#---- cut here for example1.pyh ----
# This program writes out the lines in README in preformatted HTML,
# calling the htmlesc() function to prevent interpretation of any
# special characters, such as <, >, and &.
from htmlesc import htmlesc

<HTML>
  <HEAD>
    <TITLE> "PyHTML Makes It Easy"
  <BODY>
    <PRE>
      for i in open("README").readlines():
        htmlesc(i)
print str(__output) # Current cruft necessary to see output
#---- example1.pyh ends ----

Syntax:
compound_stmt: ... | tag_stmt
tag_stmt: '<' NAME [tag_args] '>' suite
tag_args: NAME '=' expr (',' NAME '=' expr)* [',']

How it works:
The C parser for Python is modified to include a "tag statement" which is
valid in all the same places as a for statement (compound_stmt).  The builtin
modules symbol and ast are extended to be able to understand these nodes.
The compiler is extended to emit code both for the tags (my work) and for
expression-statements (Taken from Quixote) to write output to a special object.
The following code:
	<UL>
		for i in range(10):
		<LI> str(i)
would output something like
<UL><LI>0</LI><LI>2</LI>....<LI>9</LI></UL>




From fredrik at effbot.org  Tue Dec 19 14:47:06 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 19 Dec 2000 19:47:06 GMT
Subject: Tkinter text widget dump method ?
References: <mailman.977253243.31585.python-list@python.org>
Message-ID: <_WO%5.515$AH6.44674@newsc.telia.net>

Joe Murray wrote:
> In versions Python v 1.5.2 and > and Tkinter v 1.129 and >, there are no
> implementations of the Tkinter.Text widget dump methods.

from the eff-bot archives:

http://www.deja.com/getdoc.xp?AN=405964453

From: "Fredrik Lundh" <fredrik at pythonware.com>
Subject: Re: Tkinter.Text dump not implemented ?
Date: 28 Oct 1998

:::

#
# dump the contents of text widget
#
# fredrik lundh, october 1998
#
# fredrik at pythonware.com
# http://www.pythonware.com
#

from Tkinter import *
import string, random

def dump(text):
    dump = text.tk.splitlist(text.tk.call(text, "dump", 1.0, "end"))
    data = []
    for i in range(0, len(dump), 3):
        data.append(tuple(dump[i:i+3]))
    return data # a list of (tag, str, pos) tuples

# --------------------------------------------------------------------
# sample code

root = Tk()

text = Text(root, font="courier", width=65, wrap=WORD)
text.pack()

text.insert(END, (9*"-"+"!")*6 + "-----\n")

for i in range(200):
    text.insert(END, random.choice(("spam ", "and ", "egg ")))

for tag, str, pos in dump(text):
    print pos, tag, "=", repr(str)

root.mainloop()

:::

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From sandj.williams at gte.net  Fri Dec 15 10:31:06 2000
From: sandj.williams at gte.net (Steve Williams)
Date: Fri, 15 Dec 2000 15:31:06 GMT
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <tcmay-FCA22F.14073713122000@news.got.net> <918v7a02scn@news1.newsguy.com> <tcmay-04E80C.16181313122000@news.got.net> <91a1ed02a4j@news2.newsguy.com> <3A389B1A.4F33FFD4@cfmu.eurocontrol.be> <91ae8s02llf@news2.newsguy.com> <3A397778.2E6D84D3@cosc.canterbury.ac.nz> <91c7r4$7a0$1@slb1.atl.mindspring.net>
Message-ID: <3A3A3A58.3917FA57@gte.net>

Andrew Dalke wrote:

> Greg Ewing wrote:
> >the TADS (Text Adventure Development System)
> >lets you declare objects with multiple
> >classes, e.g.
> >
> >   bed: Furniture, Surface, Scenery
> >      ...
> ...
> >This is really just syntactic sugar,
> ...
> >To mention something vaguely Python-relevant:
> >if Python had a convenient syntax for declaring
> >singletons, it could be a pretty good language for
> >Interactive Fiction...
>
> One packet of syntactic sugar, coming up!
>
> def make_a(*classes):
>   class make_a:
>     pass
>   make_a.__bases__ = tuple(classes)
>   return make_a()
>
> class Furniture:
>      def comfy(self): return 1
>
> class Surface:
>     def genus(self): return 0  # Imagine a spherical cow...
>
> class Scenery:
>     def description(self): return "boring"
>
> >>> bed = make_a(Furniture, Surface, Scenery)
> >>> bed
> <__main__.make_a instance at 0x120330628>
> >>> bed.description()
> 'boring'
> >>> bed.genus()
> 0
> >>> bed.comfy()
> 1
> >>>
>
>                     Andrew
>                     dalke at acm.org

A cow is genus 1



From dale at out-think.NOSPAMco.uk  Wed Dec  6 12:59:09 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Wed, 06 Dec 2000 17:59:09 +0000
Subject: Python: what all the hype is about?
References: <90locf$d80$1@nnrp1.deja.com>
Message-ID: <9uus2tkehg7pr8gtvka4jh6q68kpvgfs3o@4ax.com>

aivkov at my-deja.com wrote:

>Dear all,
>
>Excuse my ingnorance in the subject area. Can someone explaing me very
>briefly, why one should choose Python over Java or is it better to
>compare it with Javascript/ECMA?
>
>I hope I am not trying to compare apples with oranges.
>
>Thanks,
>Anatoliy.
>
>
>Sent via Deja.com http://www.deja.com/
>Before you buy.

I haven't noticed much hype. Indeed, Python still seems like a pretty well kept secret to me.

Comparing languages is somewhat subjective unless you have a particular application in mind. Even
then, it's not really a science.

Your choice of language is dependent on a number of factors - in no particular order:

1. Suitability for the job
2. Knowledge of the language (the language you know is usually easier to code than one you don't)
3. Availability of tools (debuggers, source sharing, documenters)
4. Availability of support.

Python frequently scores well for 1 in Web/CGI, databases, middleware, file utilities, data
processing and analysis but not so well for GUI, 3D games.

For item 3, Python is, at best, average.
For 4, this group is a pretty reliable source of excellent information and help.

Someone will come along shortly and tell you the opposite, I'm sure.

--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants


From rhamorim at my-deja.com  Thu Dec  7 07:01:18 2000
From: rhamorim at my-deja.com (rhamorim at my-deja.com)
Date: Thu, 07 Dec 2000 12:01:18 GMT
Subject: Finding values on sequences and .index()
References: <90lvnm$k2a$1@nnrp1.deja.com> <3A2F0531.EBFD0C72@home.com>
Message-ID: <90nu6c$6gt$1@nnrp1.deja.com>

Thanks, it does help. I was so worried about not using for loops that I
couldn't see those options. :-)

Best regards,

Roberto Amorim

In article <3A2F0531.EBFD0C72 at home.com>,
  Donald O'Donnell <donod at home.com> wrote:
> rhamorim at my-deja.com wrote:
> >
> > Hi.
> >
> > I have a list that may have many ocurrences of the same data. How
do I
> > find the indexes for all ocurrences of it? list.index(value) only
works
> > for the first ocurrence.
> ...
>
> How about using a list comprehension (ver 2.0) like this:
>
> >>> a = [1, 2, 3, 4, 1, 5, 6, 7, 1, 8, 9, 1]
> >>> n = 1
> >>> x = [i for i in range(len(a)) if a[i]==n]
> >>> x
> [0, 4, 8, 11]
>
> Or doing it the old fashioned way (ver 1.5):
>
> >>> x = []
> >>> for i in range(len(a)):
> ... 	if a[i] == n: x.append(i)
> ...
> >>> x
> [0, 4, 8, 11]
>
> Hope this helps.
>
> Don
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From tbryan at python.net  Tue Dec  5 19:58:04 2000
From: tbryan at python.net (Thomas A. Bryan)
Date: Tue, 05 Dec 2000 19:58:04 -0500
Subject: resource module??
References: <8v1p0d$7cl$1@nnrp1.deja.com> <p6qy9yh780r.fsf@informatik.hu-berlin.de> <90jiet$ltn$1@nnrp1.deja.com>
Message-ID: <3A2D8F1C.E3355BD9@python.net>

etsang at my-deja.com wrote:
> 
> Hi, I uncomment out the line #resource resource.c in Modules/Setup, amd
> then rebuild theo whole thing again, but it is still not importing.
> Before rebuilding I have use make clean.

You sometimes need to run make distclean before rebuilding.  I think that 
modifying the Setup file is one of those cases. 

By the way, I strongly recommend that you create a Modules/Setup.local 
instead of modifying Setup directly.  That way, you can easily keep 
track of your local options and reuse your Setup.local when building 
future releases.

---Tom


From DavidL at NONOraqiaMAPS.com  Wed Dec 20 14:42:41 2000
From: DavidL at NONOraqiaMAPS.com (David Lees)
Date: Wed, 20 Dec 2000 14:42:41 -0500
Subject: Threads - Newbie Questions
Message-ID: <3A410BB1.AAB6BE3C@NONOraqiaMAPS.com>

I am now starting to play with Python threads.  Could someone refer me
to documentation other than the Library reference for some simple
examples?

My specific questions are:
1. I am passing the lock as an argument to each thread.  Is this the
proper way to do it?

2. What is the 'kwargs' - optional keyword arguments refered to in the
Library Reference for start_new_thread?

Python is definitely cool.  I am starting to get the hang of it, with a
lot of help from this group.

Thanks in advance.

david lees

Here is my initial code:

import time, thread

def TestProcess(x,n,lk):
    print 'Process %d is starting' % n
    #time.sleep(1)
    lk.acquire()
    print 'Acquired process %d ' % n
    lk.release()
    
lx=thread.allocate_lock()
lx.acquire()

for i in range(100):
    thread.start_new(TestProcess,(0,i,lx))

time.sleep(3)

print 'Threads started, about to release lock'
lx.release()


From kragen at dnaco.net  Sat Dec 30 20:59:46 2000
From: kragen at dnaco.net (Kragen Sitaker)
Date: Sun, 31 Dec 2000 01:59:46 GMT
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <slrn94smds.150.grey@teleute.rpglink.com> <FOt36.17$vd1.2224@e420r-atl2.usenetserver.com> <slrn94t1vj.150.grey@teleute.rpglink.com>
Message-ID: <mqw36.50$9s1.10042@e420r-atl2.usenetserver.com>

In article <slrn94t1vj.150.grey at teleute.rpglink.com>,
Steve Lamb <morpheus at here.not.there> wrote:
>On Sat, 30 Dec 2000 23:00:53 GMT, Kragen Sitaker <kragen at dnaco.net> wrote:
>>Hahaha!  You just plonked Fredrik Lundh in comp.lang.python.  He was
>>being helpful and polite, and you plonked him because you disagreed
>>with him.  I can guess how much help you'll get next time you have a
>>Python question.  :)
>
>    I can tell you I'll get far better without him than with him when he is
>giving advice that runs contrary to about 10+ years of programming that I've
>been taught from various different sources.  I don't care if he is God
>himself, advocating not initing a variable is stupid, plain and simple.

Some of us continue to gain new insights about programming by listening
to those who disagree with us, despite having 20 years of experience
programming.  Our humility removes limits to our knowledge and achievement.

On the other hand, some of us believe that programming for a year, or
five years, or 10+ years, means that we know everything there is to
know about the subject, and that if someone more experienced disagrees
with us, it is more likely because the more experienced person is
stupid, plain and simple, than because we can't understand their
reasoning.  This class of people generally stops learning after their
first year or so programming.

It's sad to watch, but funny, too, in a sad way.
-- 
<kragen at pobox.com>       Kragen Sitaker     <http://www.pobox.com/~kragen/>
Perilous to all of us are the devices of an art deeper than we possess
ourselves.
       -- Gandalf the White [J.R.R. Tolkien, "The Two Towers", Bk 3, Ch. XI]




From johngrayson at home.com  Wed Dec 27 14:41:23 2000
From: johngrayson at home.com (John Grayson)
Date: Wed, 27 Dec 2000 19:41:23 GMT
Subject: tkinter and scroll
References: <slrn94k7u8.14h.jp@localhost.localdomain>
Message-ID: <92dgl2$ov3$1@nnrp1.deja.com>

In article <slrn94k7u8.14h.jp at localhost.localdomain>,
  jp at ulgo.koti.com.pl wrote:
> I want to create window with:
> - text (diffrent fonts)
> - images
> - buttons
> and scroll it (becouse it can be larger than screen)
> Is it possible to do it in Tkinter?
>
> from tutorial: (scrollbar)
>
> "This widget is used to implement scrolled listboxes, canvases, and
>  text fields."
>
> what about buttons and other widgets? what about all-in-one ?
>
> If it's impossible with Tkinter - is it possible with pygtk?
>


How about using a Text widget -- you can embed anything in that...

from Tkinter import *

root = Tk()
root.title('Text')
text = Text(root, height=26, width=50)
scroll = Scrollbar(root, command=text.yview)
text.configure(yscrollcommand=scroll.set)
text.tag_configure('bold_italics', font=('Verdana', 12, 'bold',
'italic'))
text.tag_configure('big', font=('Verdana', 24, 'bold'))
text.tag_configure('color', foreground='blue', font=('Tempus Sans ITC',
14))
text.tag_configure('groove', relief=GROOVE, borderwidth=2)
text.tag_bind('bite', '<1>',
              lambda e, t=text: t.insert(END, "I'll bite your legs
off!"))

text.insert(END, 'Something up with my banter, chaps?\n')
text.insert(END, 'Four hours to bury a cat?\n', 'bold_italics')
text.insert(END, 'Can I call you "Frank"?\n', 'big')
text.insert(END, "What's happening Thursday then?\n", 'color')
text.insert(END, 'Did you write this symphony in the shed?\n', 'groove')
button = Button(text, text='I do live at 46 Horton terrace')
text.window_create(END, window=button)
photo=PhotoImage(file='lumber.gif')
text.image_create(END, image=photo)
text.insert(END, 'I dare you to click on this\n', 'bite')
text.pack(side=LEFT, fill=BOTH, expand=YES)
scroll.pack(side=RIGHT, fill=Y)

root.mainloop()

You'll have to get your own lumber.gif image...

    John Grayson


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


From gregj at pobox.com  Fri Dec  8 20:36:34 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Sat, 09 Dec 2000 01:36:34 GMT
Subject: Python wrapper for C++ STL?
References: <3A31760E.D157EB54@erols.com>
Message-ID: <C0gY5.126062$U46.4232859@news1.sttls1.wa.home.com>

"Edward C. Jones" <edcjones at erols.com> wrote:

> Has anyone written a Python wrapper for the C++ Standard Template Library?

C++ templates are compile-time constructs, not run-time features. Templates
are analogous to macros in C. They are not library functions that can be
wrapped up and called from Python.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From jp at ulgo.koti.com.pl  Wed Dec 27 19:20:14 2000
From: jp at ulgo.koti.com.pl (Jacek =?iso-8859-2?Q?Pop=B3awski?=)
Date: 28 Dec 2000 00:20:14 GMT
Subject: split hangs
Message-ID: <slrn94l1pu.4so.jp@localhost.localdomain>

My program hangs when running that instruction:

ourlist=ourregex.split(s)

I can only kill it, or ctrl+z

ourregex is of course result of re.compile()

s is a long string

it works good with other strings, why it hangs with that one?
how can I fix it? how to find out what is bad?
how it's possible that split() hangs?

PS. python-1.5.2, Linux

-- 
Now the best steel goes through the fire
And you seem to burn
The dead man's always a liar                              "Hide In The Rainbow"
Look out! there's another one learning!                     - Ronnie James Dio


From tim.one at home.com  Sun Dec 31 07:00:31 2000
From: tim.one at home.com (Tim Peters)
Date: Sun, 31 Dec 2000 07:00:31 -0500
Subject: 2001 Enchancement Wishlist
In-Reply-To: <Pine.LNX.4.10.10012301008410.5079-100000@kev.ioc.ee>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEPMIFAA.tim.one@home.com>

[Pearu Peterson]
> ...
> For example, the following behavior cannot be handled in Python 2.0:
>
> class NonZeroInteger:
>     def __init__(self,num):
>         if num == 0:
>             return 0
>         self.num = num
>     <other number class methods>
>
> so that
> >>> type(NonZeroInteger(2)),type(NonZeroInteger(0))
> (<type 'instance'>, <type 'int'>)

I bet it can't be handled that way in Python 42.0 either <0.41 wink>:  on
the very face of it, the only *sensible* thing for a NonZeroInteger
constructor to do when handed a 0 is to raise ValueError -- else its name
lies about the nature of the thing it's constructing.

Guido has Pronounced on this in the past:  a class constructor creates an
instance of the class, end of story.  It's a little bit of semantics people
can and do rely on.  For example, you didn't show any other methods of your
NonZeroInteger class, but let's say it had a method xxx.  Then any Python
programmer can rely on that

    y = NonZeroInteger(anything_at_all)
    y.xxx()

won't blow up with an AttributeError on "xxx", and without doing dynamic
type checks etc first.

There's nothing wrong with writing a factory function that can return
objects of different classes.  It often makes code using them harder to work
with, though, and precisely because you *don't* know what pops out of them
in the absence of further type checks (however spelled).

> Sure, one can write wrapper functions (as some of you already
> suggested):
>
> class _Complex:
>     ...
> def Complex(real,imag):
>     if imag == 0: return Real(real)
>     return _Complex(real,imag)
>
> but I think that this way is actually a very unpythonic one,

It's the way Guido intended from the start.  Works fine.  But you already
know that(!).

> especially if there is much simpler/cleaner solution available
> (that is, possible), that is, allowing __init__ return arbitrary
> object.

Well, I sure don't like it.

> I have used these number classes only for demonstration, but there
> are real situations where 'rich' __init__ would be very useful.

Then a factory function will be exactly as useful -- with the killer
advantage that you can actually code one today <wink>.

> For example, in PySymbolic there are classes Plus, Symbol, etc and
> gmpy numbers that are objects. E.g. class Plus represents a sum of
> its elements (arguments):
>
> c = Plus(a,b,c)       -> c = a + b + c  # c is Plus instance
>
> where a,b,c are Symbol instances.
> Now, if one constructs a Plus instance only with one argument,
>
> c = Plus(a)           -> c = a          # c is Symbol instance
>
> ***lots of code*** could be avoided with rich Plus.__init__ that would
> return in this case Symbol instance a.

Wouldn't your "rich Plus.__init__" contain about as much code as a factory
function would today?  Both have to consider all the same cases.  A factory
function would get to omit the often-useless "self" argument, though <0.9
wink>.

I suggest you're off track in your modeling, confounding roles that should
be distinct:  that of Plus as a function and Plus as an object (representing
an only partially evaluated sum).  In this case I'd probably have a base
class Expression for all kinds of expresssions (incl. atoms like Symbols),
make Plus a function mapping zero or more Expressions (or subclasses) to an
Expression (or subclass), and make Sum a subclass of Expression that
represents a partially evaluated sum.  That's the way I *think* of the
domain, so that's the way I'd want to code it (and, no, I'm not bothered by
having to come up with distinct names for the verb sense (Plus) and the noun
sense (Sum) -- the distinction between nouns and verbs is usually helpful!).

symbolic-math-would-be-useless-if-we-always-wanted-the-verb-
    form<wink>-ly y'rs  - tim




From gregj at pobox.com  Wed Dec  6 02:31:59 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Wed, 06 Dec 2000 07:31:59 GMT
Subject: [DAO Database in win32] What is more efficient
References: <0g8X5.152$7U6.3922@m2newsread.uni2.es>
Message-ID: <PXlX5.117656$U46.3885672@news1.sttls1.wa.home.com>

"Arturo P?rez Mulas" <arturo_perez at wanadoo.es> wrote:

> I am accessing an Access database under Windows NT from Python. I am
> using Mark's python windows extensions to access through the DAO
> objects, and I have a question and a performance comment:

> a) When I open a connection and create a recordset executing an arbitrary
> SQL instruction, as follows:

> db = dbengine.OpenDatabase("filename")
> SQL_query = 'SELECT Table1.Column1, Table2.Colum2  FROM Table1 INNER JOIN
Table2 ON (Table1.Num =
> Table2.Num)  WHERE (((Table1.ID)="name"));'
> myrecordset = self.db.OpenRecordset(SQL_BCSet_query)
>
> Is there any way to check if the resulting recordset is empty?

If you aren't using ADO you should be. DAO was replaced by ADO some time
ago.

The recordset's EOF property will be true right after the OpenRecordset() if
the recordset is empty.

> b) I need to add a register to a database but being sure that one
particular field is not repeated;
> I have no access to the predefined structure of the database, so I need to
check manually if
> such a register exists before adding it; which of the following is more
efficient?:

I don't know what you mean by "register" in this context. Your examples are
simply checking a table to see if any row has the desired value in a column.
If you don't know the database structure you will have trouble writing
queries; SQL statements can't refer to columns that aren't defined. You'll
generate an error, not an empty recordset.

The simplest and fastest way to see if any row has a desired value in a
column is:

  rs = db.OpenRecordset("select count(*) as 'nrows' from table where field =
value")

The resulting recordset will have exactly one row, and one column named
'nrows.' The value of that column is how many rows were found where field =
value. If the number is 0 there were no matches.

Using the database engine to scan the table will be a lot faster than
looping over a cursor as your second example does.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From rzantow at my-deja.com  Wed Dec 13 13:20:28 2000
From: rzantow at my-deja.com (rzantow at my-deja.com)
Date: Wed, 13 Dec 2000 18:20:28 GMT
Subject: Problem with += (augmented assignment)
References: <9188og$q57$1@nnrp1.deja.com> <9188ue$q9m$1@nnrp1.deja.com>
Message-ID: <918el1$vs9$1@nnrp1.deja.com>

In article <9188ue$q9m$1 at nnrp1.deja.com>,
  mjjacob at my-deja.com wrote:
> > I tried this under Python 2.0 running on Win 2k:
> >
> > x = 0
> > while x <= 42:
> > 	x += 1
> >
> > Python always complains a Syntax Error: Invalid
> > Syntax on this:
> >
> >     x += 1
> >        ^
> > SyntaxError: invalid syntax

Are you sure you have 2.0 running? I clipped your code directly and it
ran fine (Python 2.0, Win2K).


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


From neilh at scintilla.org  Sat Dec 30 17:08:20 2000
From: neilh at scintilla.org (Neil Hodgson)
Date: Sat, 30 Dec 2000 22:08:20 GMT
Subject: Running Boa Contructor with the latest versions of Python & wxPython
References: <92ld65$4un$1@news.nuri.net>
Message-ID: <o1t36.38714$xW4.308615@news-server.bigpond.net.au>

June Kim:
> I tried to run BC on my windows machine installed with the lastest
versions
> of Python(2.0) and wxPython(2.2.2) and got the following error. ...

   This is a versioning problem that Riaan or Robin has commented on
somewhere - just can't find where at the moment. Have a search in the Boa
and wxPython mailing list archives. SetKeywords looks like a good search
term for this.

   Neil





From beno at TheWebsons.com  Sat Dec 23 06:14:42 2000
From: beno at TheWebsons.com (Ben Ocean)
Date: Sat, 23 Dec 2000 03:14:42 -0800
Subject: Setting Environment Variables
Message-ID: <LOBBJMHOIFOLFIPFANHJMEFCCFAA.beno@TheWebsons.com>

Hi;
I'm trying to set environment variables so I can define my own path for my
modules. My problem is I don't know where to set the PYTHONPATH. My book
says do it in the .cshrc file: where the heck's that? My Linux book says it
should be in the /home directory, but it isn't. I own this server and it's a
standard RedHat 6.2 install. Please advise.
TIA,
BenO




From fredrik at effbot.org  Tue Dec 19 18:42:37 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 19 Dec 2000 23:42:37 GMT
Subject: while true: !!!
References: <mailman.977268243.15476.python-list@python.org>
Message-ID: <NnS%5.482$Qb7.42792@newsb.telia.net>

"Delaney, Timothy" wrote:
> maybe_buffered_stream = MaybeBufferedFileInputStream(file_object)
>
> while maybe_buffered_stream.has_unread_contents:
>        line = maybe_buffered_stream.readline()

so you prefer slow and fragile code over fast and robust
code.  fine.  now please shut up, and let this thread die.

</F>




From Gareth.McCaughan at pobox.com  Tue Dec 12 18:46:49 2000
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: Tue, 12 Dec 2000 23:46:49 +0000
Subject: Python-like compiled language
References: <913jr3$dmv$1@sylvester.vcn.bc.ca>
Message-ID: <slrn93de79.j6.Gareth.McCaughan@g.local>

Alexander Lawhead wrote:
> This may seem like an odd question, but what is the most Python-like
> compiled language? I'm a new programmer and have been learning on Python.
> I'm trying to build some command-line driven programs to do some image
> processing and am trying to find a Python-like compiled language that
> would be faster than Python+NumPy. Ideally I guess I should learn C and
> combine Python + C extensions. For the moment, I'd like to start on a new
> compiled language, preferably one that's oriented towards math/science
> (hopefully with built-in or standard matrix libraries). I've fallen in
> love with Python+NumPy's matrix capabilities and am far more interested in
> a language that comes close in this regard. Object oriented capability is
> not necessary.

The compiled language that feels most Python-like to me is
Common Lisp: very dynamic, offers OO but doesn't force it
on you, good selection of basic data types, large library
of useful things. There are some very good CL compilers
out there, too. (Dylan would be another choice, if the
syntax of CL is too weird for you.)

I'm not sure how well this fits your needs, though, if what
you want is good performance for matrix operations. That's
the sort of thing for which C and C++ usually win.

Before you start looking for other languages, though: have
you checked that the presence of Python in the mix is actually
what's holding you back here? If you're making heavy use of
Numeric Python, then most of what you do is probably happening
inside a library written in C, and you're unlikely to improve
hugely on it.

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
sig under construc


From grey at despair.rpglink.com  Wed Dec 20 20:58:48 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Thu, 21 Dec 2000 01:58:48 -0000
Subject: Greetings from the Interpreter!!!
References: <mailman.977314029.15733.python-list@python.org>
Message-ID: <slrn942oun.eoe.grey@teleute.rpglink.com>

On Wed, 20 Dec 2000 12:05:55 -0000, gunter  <gb at a3design.de> wrote:
>import string,time,os

>print (string.join((lambda x:x[:10]+x[8:])(map(lambda x:
>string.center("*" * (lambda x: ((x < 24) * (x - 3)) + 3)(x) ,24),
>range(1,28,2))),os.linesep), "Please Wait...")[time.time()<977612400]

    Alright, maybe I am missing something but that does nothing at all.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From MarkH at ActiveState.com  Sat Dec 30 00:03:31 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sat, 30 Dec 2000 05:03:31 GMT
Subject: identifying user from Win9x?
References: <92ibvf$if2$1@nntp6.u.washington.edu> <1I536.42046$3B5.1723486@newsread2.prod.itd.earthlink.net>
Message-ID: <3A4D6B80.70203@ActiveState.com>

Christopher Tavares wrote:

> The Win32 API call that you want is WNetGetUser. I would expect that it's
> exposed from the win32net module in the python win32 extensions. 

It wasn't.  It is now tho ;-)

win32api.GetUserName() may also be useful.

Mark.



From moshez at zadka.site.co.il  Wed Dec 27 12:39:05 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Wed, 27 Dec 2000 19:39:05 +0200 (IST)
Subject: Python and Napster
In-Reply-To: <000c01c06fe3$efc29520$102b2bc1@cadlab.it>
References: <000c01c06fe3$efc29520$102b2bc1@cadlab.it>, <92ap0g12b6r@news1.newsguy.com>, <sx226.18970$RC1.511898@newsread1.prod.itd.earthlink.net> <20001227164900.DF7E6A84F@darjeeling.zadka.site.co.il>
Message-ID: <20001227173905.AC436A84F@darjeeling.zadka.site.co.il>

On Wed, 27 Dec 2000, "Alex Martelli" <aleaxit at yahoo.com> wrote:

> HTTP (and the Python tools around it) is great for file transfer (and
> suchlike) but it's not oriented to peer-to-peer at all.

Well, yeah, you'd have to implement the low-level part of peer-to-peer.
But the HTTP infrastructure is usually enough for that: broadcast by
performing a GET /?file=name to say you have a file called "name"
and the nameing service turns out to be composed of 302 redirects.

> I think that actually transferring a file's data and metadata, once
> you have a proxy for a remote object, is no big issue.  It seems to
> me that naming, exceptions, resuming transfer (not supported by HTTP,
> is it?)

You have insulted the prophet: your punishment shall be cruel indeed!
Of course HTTP supports resuming transfer. SimpleHTTPServer doesn't
though. I don't think it should be hard to add, though.

An advantage of HTTP is that it will be possible to use the protocol
with clients not written in Python. If what he wants is simple file
exchange, there is no reason to be limited to Python at the protocol
level.
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From emile at fenx.com  Fri Dec 22 11:46:42 2000
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 22 Dec 2000 08:46:42 -0800
Subject: Newbie Question
References: <05m64tkqev0aavea80ttjse1sj2gd447hg@4ax.com> <hqs64tc5jjndav3p6nvmu67t7hqaf5go00@4ax.com> <91vuj5$5n7is$1@ID-11957.news.dfncis.de> <42074t8a8e8604sgl77egadao721tk62ik@4ax.com> <92005r$5e67v$1@ID-11957.news.dfncis.de>
Message-ID: <9200gh$5o0sr$1@ID-11957.news.dfncis.de>

BTW, it also works with

    def run(self):
        os.path.walk(self.root, Class.func, self)


--

Emile van Sebille
emile at fenx.com
-------------------


"Emile van Sebille" <emile at fenx.com> wrote in message
news:92005r$5e67v$1 at ID-11957.news.dfncis.de...
> I'm still not sure, as this works as well:
>
> import os
>
> class Class:
>     def __init__(self, root):
>         self.root = root
>     def func(self, root, files):
>         print self
>         print root
>         print files
>     def run(self):
>         os.path.walk(self.root, Class.func, arg)
>
> root = r'C:\etvwin'
> arg = Class(root)
>
> arg.run()
>
>




From tim.one at home.com  Mon Dec  4 12:13:33 2000
From: tim.one at home.com (Tim Peters)
Date: Mon, 4 Dec 2000 12:13:33 -0500
Subject: (no subject)
In-Reply-To: <5.0.2.1.1.20001204105504.00aa6ec0@mail.abivest.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEFAICAA.tim.one@home.com>

[posted & mailed]

[Eric Stechmann]
> Sorry to waste bandwidth but...
>
>    I'm receiving little or nothing on this newsgroup.  Can anyone
>    fill me in on the situation, please?

Eric, do you mean the newsgroup, or do you mean the c.l.py mailing list?
The news<->mail gateways have been hosed "almost all the time" for a month,
so if you mean the mailing list "that's why".  Barry is in the process of
moving some of the python.org mailing lists to a new machine, which may (or
may not ...) help.

in-the-meantime-view-it-as-a-vacation-ly y'rs  - tim




From lars at stea.nu  Wed Dec 20 20:41:53 2000
From: lars at stea.nu (dv)
Date: Thu, 21 Dec 2000 01:41:53 GMT
Subject: Editing files
Message-ID: <Bdd06.5591$K6.674078@juliett.dax.net>

I'm trying to automate some configuration tasks with python. Is it possible
to do a "search and replace" or specify where (which line) in a file python
will write?

Thanks in advance

Lars




From garryknight at bigfoot.com  Sat Dec 23 20:29:59 2000
From: garryknight at bigfoot.com (Garry Knight)
Date: Sun, 24 Dec 2000 01:29:59 +0000
Subject: coding convention conversion
References: <mailman.977516469.1157.python-list@python.org>
Message-ID: <923jin$q2j$2@taliesin.netcom.net.uk>

In article <mailman.977516469.1157.python-list at python.org> "Issac
Trotts" <trotts at llnl.gov> wrote:

> Does anyone know of a utility to change between the coding conventions
>  of
> 
> def fooBarBaz(): pass
> 
> and 
> 
> def foo_bar_baz(): pass

Well, this turned out to be fairly trivial for functions with names
like fooBarBaz, but became a little less trivial for cases where the
author has defined functions with names like FooBarBaz. However, this
should do the job:

--------8X-cut here-----------------------------------------------
#!/usr/bin/perl

while (<>) {
  push(@lines, $_);
  if (/^\s*def\s+(\w+)\s*\(/) {
  $defs{$1} = "\l$1";
  }
}
@list = keys(%defs);
foreach $line (@lines) {
  foreach $def (@list) {
    if ($line =~ /$def/) {
      $line =~ s/$def/$defs{$def}/;
      $line =~ s/([A-Z])/_\l\1/g;
      last;
    }
  }
  print $line;
}
--------8X-cut here-----------------------------------------------

As before, copy and paste the text between the cut lines and save it
as, say, /usr/local/bin/redef (or the equivalent location on your
system) and use it as follows:
  redef <infile >outfile
Of course, I'm assuming you're running a *nix system with Perl
installed. If not, hopefully someone else will come up with a solution.

Thanks to David Porter and Ben Wolfson for helpful suggestions.

-- 
Garry Knight
garryknight at bigfoot.com


From slhath at home.com  Mon Dec 11 22:24:43 2000
From: slhath at home.com (Scott Hathaway)
Date: Tue, 12 Dec 2000 03:24:43 GMT
Subject: corba, soap, com, etc...
References: <pKsY5.58762$yR4.1496678@news1.rdc1.tx.home.com> <3QzY5.9825$dg6.637046@pouncer.easynews.com> <912ls3$bn8$1@violet.singnet.com.sg>
Message-ID: <%TgZ5.64106$yR4.1673707@news1.rdc1.tx.home.com>

Thanks to everyone for input.  I will probably go with xmlrpc because my
needs are simple.

Scott



"Ng Pheng Siong" <ngps at madcap.dyndns.org> wrote in message
news:912ls3$bn8$1 at violet.singnet.com.sg...
> According to Jiva DeVoe  <jiva at no-spam.com>:
> > Check out fnorb.  http://www.fnorb.org  It's an excellent open
> > source CORBA
> > implementation for python.
>
> Just downloaded Fnorb. According to the docu, it supports
> BOA. BOA has been superceded by POA.
>
> See the CORBA FAQ here:
>
>     http://www.javacommerce.com/CorbaFaq/toc-exhaustive.html
>
>
> OmniOrb3 supports POA:
>
>     www.uk.research.att.com/omniORB/omniORBpy/
>
>
>
> --
> Ng Pheng Siong <ngps at post1.com> * http://www.post1.com/home/ngps
>




From z3penguin at penguinpowered.com  Thu Dec  7 20:42:45 2000
From: z3penguin at penguinpowered.com (Z 3 Penguin)
Date: Thu, 7 Dec 2000 20:42:45 -0500
Subject: problems with urllib in Python20?
References: <90pdir$k7$1@bob.news.rcn.net> <90pe4i$7fk$1@panix6.panix.com>
Message-ID: <90peb3$5ef$1@bob.news.rcn.net>

Microsoft Windows 98 4.10.1998



From nigeljennison at cs.com  Tue Dec  5 20:12:30 2000
From: nigeljennison at cs.com (Nigel)
Date: Wed, 06 Dec 2000 01:12:30 GMT
Subject: Python Origins
Message-ID: <90k3pq$50p$1@nnrp1.deja.com>

 I am doing some research and I would be grateful if anybody could give
me some history on the origins of Python.

--
Worker
Disk Drive Computer Books
http://disk.vstorecomputers.com


Sent via Deja.com http://www.deja.com/
Before you buy.


From ansofive at bellsouth.net  Wed Dec  6 11:08:45 2000
From: ansofive at bellsouth.net (Dennis Dunn)
Date: Wed, 06 Dec 2000 21:08:45 +0500
Subject: Cookies seem to disappear
Message-ID: <wnCX5.218$4b3.42696@news4.mia>

Hi,

I am trying to use SmartCookie class to store state.  Everything works
perfectly until I exit my browser.  The next time I try to view the page
my state has been reset to the default values.  I set the 'max-age'
value like  so:

c = Cookie.SmartCookie()
c['state'] = stateDictionary
c['state']['max-age'] = '3600'
print c.output()
print

I've tried Netscape on WinNT and Linux and IE on WinNT, they all exhibit the
same behaviour.  The thing that is throwing me is that everything works until I
exit my browser.  It is as if I'm not setting the max-age correctly but
how difficult can that be?

Thanks for your help.

--dennism


From wolfson at midway.uchicago.edu  Fri Dec  8 22:01:45 2000
From: wolfson at midway.uchicago.edu (Ben Wolfson)
Date: Sat, 09 Dec 2000 03:01:45 GMT
Subject: Python Closures
Message-ID: <tghY5.161$E3.3547@uchinews>

More-or-less closures can be used in Python with, of course, the
lambdabuse module.  Observe:

>>> import lambdabuse as l
>>> l.Def('make-adder', lambda n: l.Def(l.ANONYMOUS,
				    lambda x: x + l.Val('n')))
Call_Wrapper instance: make-adder
>>> add4 = l.Call('make-adder', 4)
>>> add4(9)
13
>>> 

Source for the module is at 

http://home.uchicago.edu/~wolfson/Python/lambdabuse.py
                                        /LambdaDict.py
                                        /LambdaEnv.py
or + .html for colorization.

-- 
BTR
It infuriates me that Dracula may have needed my soul more than I do.
 -- Stanley Donwood, "Dracula"



From fredrik at effbot.org  Sat Dec  9 17:21:51 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sat, 09 Dec 2000 22:21:51 GMT
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu> <976153091.257544@clam-ext> <3A30A181.4302F102@caltech.edu> <p6qr93ik7ks.fsf@informatik.hu-berlin.de> <90u5mo$kr@gap.cco.caltech.edu>
Message-ID: <3gyY5.2864$O%.449093@newsc.telia.net>

Nathaniel Gray wrote:
> So why not provide a pointer instead of just flaming?

some people do, you know.

(or was my first post to this thread lost somewhere
in the vast usenet space?)

</F>




From goose at enteract.com  Tue Dec 12 19:46:27 2000
From: goose at enteract.com (Jerome Mrozak)
Date: Tue, 12 Dec 2000 18:46:27 -0600
Subject: [Q] Are Exceptions used that much in practice?
References: <3A35D546.A4731B5B@enteract.com> <3A36596A.89EEA210@alcyone.com>
Message-ID: <3A36C6E3.26181E3F@enteract.com>

I believe I've been misunderstood.  The problem is not "can I write code
without handling exceptions" but rather "are exceptions used
voluntarily, rather than by compulsion".

An example of compulsion is java.lang.Integer.parseInt(), which throws a
NumberFormatException you *must* deal with or the program won't
compile.  I'm aware of examples of Python such as (faking it here):

while 1:
  try:
    input = inputFileObject.readline()
  catch EOF:
    break
  else:
    process(input)

The point of the question is this -- is code like this sample a way of
exercising the try-catch, or is this an example of something I end up
doing all of the time because it is so much easier than the alternative?

Other posters to this question provided a fair sampling of examples.

Some, I'm disappointed to say, were, shall we say, not so supportive.  

The long-range goal of my question is to determine for my satisfaction
whether it is worth while for me to spend my limited time pursuing
Python as a compliment to (whatever other skills I might have) or to go
with the apparently much greater flow and use those resources on Perl. 
The 'exception' thing could be a determiner, esp. with nothing similar
in Perl, but only if I can use it in a *positive* way instead of just to
test a return value.  Heck, Perl has 'die' for that kind of testing.

Thanks for your concern anyways.

Jerome.





Erik Max Francis wrote:
> 
> Jerome Mrozak wrote:
> 
> > I realize that Python allows me to catch exceptions.  But being a
> > novice
> > I'm not sure if they are used much in the 'real world'.
>         ...
> > Another way of putting it is, if the try-catch-else didn't exist would
> > anyone be terribly put out?  (If not then the construct isn't much
> > used...)
> 
> Yes.  Yes.  Yes.
> 
> Anyone who has done substantial programming in Python, Java, or C++ has
> used exceptions extensively.  If you haven't found a significant use for
> exceptions, then you're either writing extremely trivial programs (ones
> so trivial they cannot fail, which has to be _really_ trivial) or you're
> not handling error conditions properly.  Truth be told, handling
> arbitrary error conditions properly is one of the more difficult things
> to do in computer science.  Exceptions, although they require a
> significant change in the way you write code, make this a lot easier.
> Draft Standard C++ programming without exceptions was obnoxious.
> 
> --
>  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
>  __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
> /  \ Love is not love which alters when it alteration finds
> \__/ William Shakespeare, _Sonnets_, 116
>     Alcyone Systems' CatCam / http://www.catcam.com/
>  What do your pets do all day while you're at work?  Find out.

-- 
Jerome Mrozak          "Never buy a dog and bark for yourself"
goose at enteract.com     --"Slippery" Jim DiGriz
                         (the Stainless Steel Rat)


From aleaxit at yahoo.com  Sat Dec 23 12:39:29 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 23 Dec 2000 18:39:29 +0100
Subject: A Suggestion for Python Colon Syntax
References: <mailman.977544967.21073.python-list@python.org> <CF316.26760$1t.879107@typhoon.ne.mediaone.net>
Message-ID: <922o3102a8b@news1.newsguy.com>

"David Abrahams" <abrahams at mediaone.net> wrote in message
news:CF316.26760$1t.879107 at typhoon.ne.mediaone.net...
    [snip]
> >     x = x + y
> >
> > could end up generating multimegabyte temp arrays while
> >
> >     x += y
> >
> > need not.  That is, for them it wasn't "just another way" to write the
> > former; it offers crucially different semantics.
>
> Yes! Any chance of getting this capability for extension classes?

Not sure what you mean by this question.  Both Python-written
classes (by implementing __iadd__) and extension types that are
written in C (by appropriate flags in their method tables) can do
in-place additions (&c).  What are 'extension classes'...?  Anyway,
this stuff is underdocumented, but, study the sources... it works.


Alex





From jon at bezek.dstc.monash.edu.au  Thu Dec 14 02:29:37 2000
From: jon at bezek.dstc.monash.edu.au (Jonathan Giddy)
Date: 14 Dec 2000 18:29:37 +1100
Subject: [Q] Are Exceptions used that much in practice?
References: <3A35D546.A4731B5B@enteract.com> <914mfd02ibt@news2.newsguy.com> <3A36F20B.7B766D0@cosc.canterbury.ac.nz> <uaea0ghlx.fsf@ctwd0143.fitlinxx.com>
Message-ID: <jon.976778599@bezek.dstc.monash.edu.au>

David Bolen <db3l at fitlinxx.com> writes:

] In the case of IOError that does imply parsing a string and assuming
] it starts with [Errno #], which is sort of crufty.  But other
] exceptions (such as those from win32all) are fairly standardized as a
] tuple with the code as the first element - something I also tend to do
] with my own user exceptions in my code.

Note that now that the standard exceptions are classes, IOError instances 
have an errno attribute, which is even nicer than using tuple values.

try:
    f = open(filename, 'r'):
except IOError, exc:
    if exc.errno == errno.ENOENT:
        pass
    else:
        raise



From grey at despair.rpglink.com  Thu Dec 21 11:40:01 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Thu, 21 Dec 2000 16:40:01 -0000
Subject: while true: !!!
References: <mailman.977268243.15476.python-list@python.org> <NnS%5.482$Qb7.42792@newsb.telia.net> <4bY%5.172261$U46.5444054@news1.sttls1.wa.home.com> <91q3gr$nbr$1@nnrp1.deja.com>
Message-ID: <slrn944cj1.km0.grey@teleute.rpglink.com>

On Wed, 20 Dec 2000 11:00:45 GMT, eff at my-deja.com <eff at my-deja.com> wrote:
>Greg Jorgensen wrote:
>> This thread is a great example of what newsreader filters
>> are good for.

>so?  this newsgroup is cross-linked with a mailing list.
>how about showing the people on that list some respect?

    Alrighty.  This thread is a great example of what mail filters are for.
There, respect shown.

>> Obviously quite a few of us are getting something out of it.

>lots of new names in the killfile, you mean?

    S T E V E <space> L A M B.  Don't want you to misspell it.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From Bjoern.Giesler at gmx.de  Wed Dec 13 04:40:08 2000
From: Bjoern.Giesler at gmx.de (Bjoern Giesler)
Date: 13 Dec 2000 09:40:08 GMT
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu>
Message-ID: <917g5o$qh$1@news.rz.uni-karlsruhe.de>

Hi,

since I'm considering to start working on a Python IDE that handles things
similarly to the Smalltalk IDEs I've seen (email me for more info), I'm very
interested in this debate. I'm not much of a Smalltalker -- the thing about
having to distribute entire system images that contain their very own GUI
etc. has always repelled me a bit, although I love the language for its
cleanness and the IDE for its efficiency.

Do you think that it would be possible to use Python as a Smalltalk
replacement if an IDE existed that hid some of Python's limitations? In the
following, I'm taking a look at each of the limitations that Russell
mentioned and try to examine if they could be remedied by a good IDE. Please
correct me if I'm wrong, a discussion of these points _before_ I start is
very important to my project.

In comp.lang.python Russell E. Owen <owen at astroNOJNK.washington.edu.invalid> wrote:
[snip]
: On the other hand, it is not as powerful, clean and regular as 
: smalltalk. Limitations include:
: - The language is somewhat cluttered and irregular because it has 
: procedures, functions and methods. Hence arbitrary decisions had to be 
: made in some areas as to how to implement things, which can lead to some 
: hunting through the manuals.

This refers to the standard Python library, I believe? I agree that it is
not as beautiful as it could be, but "hunting through manuals" fixes this.

: - It has lambda functions, but getting late binding behavior is too 
: difficult (though doable).

I'm not sure what you mean by this. Could you give me a short example?
(Python and/or ST.)

: - Only two scopes for variables: global (to a text file) or local (to a 
: function/method).

I can only think of two other scopes: local to an instance and local to a
class. The former can be remedied by assignment to self.something; the
latter can be simulated by text-file global variables. This could be hidden
by an IDE.

: - There are no class methods, though they can be faked reasonably well 
: with modules.

Same as the last -- this, also, could be hidden by an IDE.

: - Basic built in types (e.g. integers, floats and lists) are not 
: classes, in the sense that they do not have class methods.

Is this really a limitation, given Python's different control structures? Of
course, 0 to: 499 do: [:i|transcript put:i ]. doesn't work, but we have
for i in range(500): print i instead.

: - No boolean class; uses C's icky everything-is-a-logical idea.

This, too, is (perhaps) ugly, but not really a limitation IMHO.

: - I personally find not having to declare variables a negative (too 
: dangerous), though it's reasonable for simple scripting languages

Matter of taste, no? :-)

: - garbage collection is reference counted, though now enhanced in Python 
: 2.0 so it can handle most cases of circular references

This works really well in Python, and 2.0 has true GC, IIRC.

: - Collection classes are good, but no match for Smalltalk. (But then 
: what language comes close to ST in this area?). It does have 
: dictionaries and lists, both of which are great. But that's about it. It 
: also has "tuple", a minor and annoying (to me) variant of a list that is 
: immutable (more or less) and so can be used as a key in dictionaries.

Agreed, but could be remedied with a small addition to the Python standard
library.

: - Mediocre GUI support (a mediocre interface to Tk and a different 
: interface to wxWindows that isn't available on my platform--Mac)
: - Mediocre IDE. I hope this will be improved over time, along with the 
: GUI support.

Agreed. My work would address those two.

: - Rather ugly implementation of "package" (multi-file modules). This is 
: one of the few areas where I find the implementation to be ugly and 
: hastily done; generally it's a very clean language.

Agreed.

: - Like most languages, it uses text files to contain sets of code. I 
: really miss Smalltalk's more integrated approach.

Agreed, but this is another thing that I'd like to address.

Thanks for any and all input, and if anyone wants to participate or know
more, please email me.

Regards,
		--Bj?rn


From aleaxit at yahoo.com  Fri Dec 15 11:41:14 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 15 Dec 2000 17:41:14 +0100
Subject: Python advocacy
References: <91bln3$lmr$1@nnrp1.deja.com> <91d63k0msl@news1.newsguy.com> <8iak3t4a3lrqcqp9iqftmekdoahuico2b0@4ax.com>
Message-ID: <91dhke0179q@news1.newsguy.com>

"Steve Horne" <sh at ttsoftware.co.uk> wrote in message
news:8iak3t4a3lrqcqp9iqftmekdoahuico2b0 at 4ax.com...
    [snip]
> the books by Steven Pinker (one on the origins of language, the other
> more generally on human nature - I forget the titles).

"The Language Instinct" is (he admits grumblingly, as somebody
who has his own not-perfectly-consonant professional opinions
on linguistics -- but I've been out of the field for 10 years,
so I can't really argue against Pinker effectively!-) is a
rather good book, yes.  "How the Mind Works", I haven't read
(darn the 'net -- I can't get more than 20-30 books read per
year, any more -- so I'm _badly_ falling behind, as my huge
stacks of just-waiting-to-be-read books keep reminding me!).

Pinker's polemic against S. J. Gould, to be found at:
    http://www.mit.edu/~pinker/GOULD.html
also makes for pretty interesting reading -- though I wish
I had an URL for Gould's side of the coin, too, since that
is the one which I find more convincing.  However, something
both sides fail to note is that evolutionary psychology has
_intrinsic_ interest, when it fails to match the facts just
as much as when it's spot-on -- just as, and for the same
reasons as, economics.  Economics can be seen as a formalized
attempt to explain _every_ human behaviour "as if" humans
were perfect utility-maximizing machines (Steven Landsburg's
splendid "The Armchair Economist" is a good framing of that).

One does necessarily believe humans really *are* that way,
but it's very interesting to see what can be explained, and
what gets left out, by rigorously following that assumption
("If the fool would but persist in his folly, he would
become wise", as Blake put it:-).  I look on evolutionary
psychology as a similarly extremized attempt, except that
the assumption (which one had better not _truly_ believe in...)
is that every single aspect is reproductively adaptive (but
Dawkins himself originated the idea of 'meme', so how can
any disciple "believe" that _genes_' reproduction is what
is being maximized in each case...?).

The difference seems to me to be that some of the best
economists appear to be fully aware of the nature of their
game (Arrow and Landsburg come to mind at once), but I'm
not so sure about similar self-awareness in evolutionary
psychologists -- the mentions of possible non-adaptiveness
which Pinker quotes do happen (though much more rarely
than he leaves the impression of, it seems to me), and
that spoils the game a little bit.  Oh well...


> >Makes me wonder why Larry Wall didn't think to include at least
> >ONE strangeness, wart, or inconsistency *somewhere* in Perl...
> >thus making it so boringly perfect and un-idiosyncratic as to
> >leave its advocates without a suitable 'scapegoat feature'!
>
> Erm - either you haven't used Perl, or your an advocate yourself!

We're sorry, but our supply of emoticons is temporarily exhausted.

This emergency situation will be remedied shortly; in the meantime,
you are advised to try and ascertain the actual intent of utterances,
particularly which ones are intended to be taken "in jest", on the
basis of context.

(One great feature of 'Jeeves and Wooster', to go with the
nonpareil skill of Hugh Laurie and Stephen Fry of course,
was the _lack_ of a laugh-track.  It did run for 4 series,
so maybe I'm not the only deadpan-humour-lover left...?)


Alex





From tim.one at home.com  Wed Dec 13 22:32:36 2000
From: tim.one at home.com (Tim Peters)
Date: Wed, 13 Dec 2000 22:32:36 -0500
Subject: null bytes in re pattern - difference between 1.5.2 and 2.0?
In-Reply-To: <14904.14997.494451.840804@beluga.mojam.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCMENJIDAA.tim.one@home.com>

[posted and mailed]

[Skip Montanaro]
> I want to delete control characters from some strings.  Accordingly, I
> tried:
>
>     name = re.sub("[\000-\037\177]", "", name)
>
> This works in Python 2.0 but not in 1.5.2.  In 1.5.2 I find I need
> to use raw strings:

That's a good idea in 2.0 too, y'know.

>     name = re.sub(r"[\000-\037\177]", "", name)
>
> Accordingly, using raw strings in 2.0 fails.

Eh?  Prove it.  That is, submit a bug report with a specific failing example
if that's true.  Works for me:

Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
IDLE 0.6 -- press F1 for help
>>> import re
>>> allchars = [chr(i) for i in range(256)]
>>> fat = "".join(allchars)
>>> print len(fat)
256
>>> skinny = re.sub(r"[\000-\037\177]", "", fat)
>>> print len(skinny)
223
>>> 256 - 223
33
>>>

> Is there some form that will work both in 1.5.2 and 2.0?

The r-string form.  Or use the optional deletechars argument to
string.translate, which should run much faster in either version.

> Is this a change I should have expected?

No.

> I assume it has something to do with Unicode support in 2.0.

Much more mundane than that:  1.5.2 used a 3rd-party regexp engine (PCRE),
and its interface required passing in the pattern as a regular old C string.
So you couldn't pass a pattern with a literal null byte in 1.5.2.

ghosts-chasing-ghosts-ly y'rs  - tim




From aahz at panix.com  Sat Dec 16 10:19:52 2000
From: aahz at panix.com (Aahz Maruch)
Date: 16 Dec 2000 07:19:52 -0800
Subject: Alarms in threads - how to get/mimic?
References: <Pine.LNX.4.21.0012160248520.7483-100000@traal.eecs.berkeley.edu>
Message-ID: <91g16o$id8$1@panix6.panix.com>

In article <Pine.LNX.4.21.0012160248520.7483-100000 at traal.eecs.berkeley.edu>,
Luca de Alfaro  <dealfaro at eecs.berkeley.edu> wrote:
>
>I am trying to write a multi-threaded program to read web pages.  The
>program spawns (say) 10 threads, assigning to each thread a file to
>read with the ftp or http protocol; each thread reads it, and returns
>the info.
>
>The problem is that the threads sometimes hang for a very long time
>waiting for a connection (httplib seems to have very long timeouts??),
>whereas instead I don't care for the answer if it comes later than
>(say) a few seconds.

httplib does not timeout in its default state; the current best solution
is to download timeoutsocket.py.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

"It's 106 miles to Chicago.  We have a full tank of gas, a half-pack of
cigarettes, it's dark, and we're wearing sunglasses."  "Hit it."


From kmbotha at netinfo.ubc.ca  Sat Dec 30 20:37:10 2000
From: kmbotha at netinfo.ubc.ca (Kris Botha)
Date: Sun, 31 Dec 2000 01:37:10 GMT
Subject: createfilehandler on Win32
References: <3a4b186a.40583796@news.interchange.ubc.ca>
Message-ID: <3a4e8d78.7362859@news.interchange.ubc.ca>

>My question  -  is there a simple way to accomplish this task (i.e.
>program a client socket, 'waiting' for messages) without using the
>createfilehandler method?
>

NEVER MIND!

I was (am?) young and stupid when I wtote that msg. I didn't realize
how powerful a solution *threads* can be (provided you implement them
with caution). 

Cheers

~~====----..----====~~
        Kris Botha



From aahz at panix.com  Fri Dec 29 11:10:16 2000
From: aahz at panix.com (Aahz Maruch)
Date: 29 Dec 2000 08:10:16 -0800
Subject: Python vs. PHP (& Java?)
References: <69BC4C6AE83ED311BC9400805FA7B14AF84D2C@il93exp01.css.mot.com> <mailman.978059571.14515.python-list@python.org> <397C7FA100EE65A5.0FF0B81DCC646CA3.C3E3EFFAF818CC30@lp.airnews.net>
Message-ID: <92id18$js4$1@panix2.panix.com>

In article <397C7FA100EE65A5.0FF0B81DCC646CA3.C3E3EFFAF818CC30 at lp.airnews.net>,
Cameron Laird <claird at starbase.neosoft.com> wrote:
>
>  [Heisenberg]
>
>Metaphoric application to macroscopic phenomena is, as Moshe notes,
>often overdone.

And in some cases, it's underdone as a metaphor.  I remember reading an
article recently about how poor "focus groups" are for getting useful
results.  One of the examples given was about how the experimenters were
watching one focus group that was giving them bizarre results; the
experiementers were behind one-way glass.  As the test group filed out
of the room, one of the troublesome subjects turned to the one-way
glass, smiled, and waved.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

Ninth Virtual Anniversary: 2 days and counting


From erikd at zip.com.au  Wed Dec 20 02:01:32 2000
From: erikd at zip.com.au (Erik de Castro Lopo)
Date: Wed, 20 Dec 2000 07:01:32 GMT
Subject: length of wav files
References: <3A402D8E.81978F3B@uniserve.com>
Message-ID: <3A4057E4.9FAB4D1@zip.com.au>

Bob van der Poel wrote:
> 
> 
> I am looking at the docs for the wave library and can't seem to find the
> one bit of information I want to extract: the running time of a wav
> file. So, before I have to relearn my math, has someone some code which
> will deterime the time in seconds of a wav sound file?

import wave

wfile = wave.open ("somefile.wav", "r")
time = (1.0 * wfile.getnframes ()) / wfile.getframerate () 

The `1.0' above to to force the calculation to use floating point
arithmetic.

Erik
-- 
+-------------------------------------------------+
     Erik de Castro Lopo     erikd at zip.com.au
+-------------------------------------------------+
"I love deadlines. I especially like the whooshing sound 
they make as they go flying by."  -- Douglas Adams


From jh at cddk.dk  Mon Dec  4 08:10:16 2000
From: jh at cddk.dk (Jesper Hertel)
Date: Mon, 4 Dec 2000 14:10:16 +0100
Subject: Python And MS Word Com Server
References: <90bc4c$rhr$02$1@news.t-online.com>
Message-ID: <90g54c$chj$1@news.inet.tele.dk>

I have exactly the same problem, but no solution, unfortunately.

I think there must be a bug in win32com. The problem seems to arise when you
use keyword arguments. If you do not use keyword arguments, and instead
supply the arguments in the right sequence, it works. (I cannot test the
code below though, as I haven't access to Word 8 or above).

The problem with this "solution" is that you cannot omit arguments before
the arguments you want to supply to the function, so you have to supply the
default values explicitly for the arguments that should just take default
values, and which are placed before the arguments you want to supply. This
could be OK, but sometimes you cannot supply a default value explicitly to
Word, because sometimes the only way to tell Word to use the default value
of an argument is to omit the argument.

I tried looking in the source code for win32com for bugs, but it was too
complicated to understand in the limited time I had.

Does anyone have a solution to this?


Best regards,

Jesper Hertel
Forlaget Magnus A/S
Denmark

"Ralf Claus" <ralf.claus at t-online.de> wrote in message
news:90bc4c$rhr$02$1 at news.t-online.com...
> Hello,
> i have one special problem with the Word COM server !
>
> import win32com.client
> msw = win32com.client.Dispatch("Word.Document")
> msw.Application.PrintOut(OutputFileName = "D:\\test.prn",
> PrintToFile="True")
>
> The COM Server ignore the given filename, and the PrintToFile method .
> What ever i try, it was printed by the system standard printer!
>
> The same code in a VB macro in word works fine
> What's wrong?





From claird at starbase.neosoft.com  Fri Dec 29 09:15:15 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 29 Dec 2000 08:15:15 -0600
Subject: Python vs PHP (& Java?)
References: <92f3kp01cqe@news1.newsguy.com> <slrn94l86m.a3h.tim@degree.ath.cx> <mailman.978021012.21874.python-list@python.org> <3A4C8871.DB734376@stroeder.com>
Message-ID: <C56A958929D650C9.FC4D26C518326FAC.CD8FA9082F345503@lp.airnews.net>

In article <3A4C8871.DB734376 at stroeder.com>,
Michael =?iso-8859-1?Q?Str=F6der?=  <michael at stroeder.com> wrote:
>Moshe Zadka wrote:
>> 
>> Apache modules. Yes, now that we have mod_snake, you
>> can do these things in Python, in theory: but mod_snake is still
>> young, while mod_perl is quite mature.
>
>mod_python?
>
>Ciao, Michael.

Did you intend to supply readers with directions
to <URL: http://www.modpython.org/ >?
-- 

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


From sholden at holdenweb.com  Mon Dec 11 08:45:49 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 11 Dec 2000 08:45:49 -0500
Subject: Tkinter frame resizeing
References: <t30nhn9kq4a5a3@corp.supernews.com>
Message-ID: <TW4Z5.3380$Q54.129373@e420r-atl1.usenetserver.com>

Tony <colaveaj at sirinet.net> wrote in message
news:t30nhn9kq4a5a3 at corp.supernews.com...
> 1.  How do you turn on the resize feature off for frames?
>
> 2.  In Java you can modal windows?  How is this achieved with python using
> Tkinter?
>
I believe you've already had an answer for Q1.  As far as modal dialog(ue)s
go, the simplest way seems to be to call the mainloop() method of the dialog
window.  This suspends the root mainloop until the dialog is dismissed.

regards
 Steve





From cjensen at bioeng.ucsd.edu  Thu Dec 14 19:04:32 2000
From: cjensen at bioeng.ucsd.edu (Curtis Jensen)
Date: Thu, 14 Dec 2000 16:04:32 -0800
Subject: Capturing stdout from a console process?
References: <91b8q3$aqu$1@nnrp1.deja.com>
Message-ID: <3A396010.D14E1408@bioeng.ucsd.edu>

Sam Schulenburg wrote:
> 
> I am using <os.system("xcopy c:*.* d: /v")> within a python program to
> copy files from one directory to another under windows NT. When the
> program executes, a new console window is created with the results of
> the xcopy command being displayed in the console window. Is their a way
> to capture this output by python so I can log error messages?
> 
> Sam Schulenburg
> 
> Sent via Deja.com
> http://www.deja.com/

I think os.popen is what you want.

-- 
Curtis Jensen
cjensen at bioeng.ucsd.edu
http://www-bioeng.ucsd.edu/~cjensen/
FAX (425) 740-1451


From erik_who at my-deja.com  Wed Dec 13 15:30:33 2000
From: erik_who at my-deja.com (Erik Who)
Date: Wed, 13 Dec 2000 20:30:33 GMT
Subject: Tutorial for emacs python-mode?
References: <Xns9009D2CA0Latzikatz@139.174.2.56>
Message-ID: <918m9a$7bc$1@nnrp1.deja.com>

Start out with the mode documentation.  From a buffer that is in
python-mode, type: M-x describe-mode.  This will give the documentation
(keymappings, methods, variables, etc.) for the current major mode and
all minor modes.  To change a buffer's major mode to python-mode, type:
M-x python-mode.  X/Emacs is intimidating at first, but well worth the
effort.

-Erik


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


From aleaxit at yahoo.com  Mon Dec 18 12:28:14 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 18 Dec 2000 18:28:14 +0100
Subject: Solution: Direct access to Printer I/O lines
References: <917k9b$9i9$1@nnrp1.deja.com> <slrn93hrjn.2vp.ssthapa@localhost.localdomain> <91cnp3$g59$1@nnrp1.deja.com> <91d23f0jk7@news1.newsguy.com> <91l4co$jua$1@nnrp1.deja.com>
Message-ID: <91lhgi0qt@news1.newsguy.com>

<jkndeja at my-deja.com> wrote in message news:91l4co$jua$1 at nnrp1.deja.com...
> Hi Alex,
>
> OK, without (yet...) going too deep into the more-generalised-and-
> interesting patterns discussion developing here:
>
> If I/you/we don't want singleton objects, but _do_ want to avoid having
> multiple processes accessing a single port, how about this:

Seems nice, though I'd call the method which acquires the resource
'acquire' rather than 'claim' -- but that's quibbling.  __del__
would also be defined, and duplicate the functionality of .release
if need be (but it IS best to have an explicit releasing-method
as well, since, for example, client-code might need to do some writing,
then keep the port-handle around for some more reading, being able
to release it for writing in the meantime -- but also, for more
general reasons).

The only issue (which depends on the device attached to the
port, alas...) is IF reading is actually 'non-invasive' so it
can be allowed without requiring 'acquisition'.  I guess in
most cases it will be, but...


Alex





From frankmcgeough at my-deja.com  Thu Dec 21 21:43:36 2000
From: frankmcgeough at my-deja.com (frankmcgeough at my-deja.com)
Date: Fri, 22 Dec 2000 02:43:36 GMT
Subject: COM / Variants / VT_BOOL
Message-ID: <91uf4o$adi$1@nnrp1.deja.com>

Simple questions. How do I create a variant in Python? How do I set
that variant to hold a VT_BOOL?
More complicated question. Why is this info so hard to find? I would
think it would come up a lot if folks are trying to use the
win32com.client.
Thanks for any clues, Frank


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


From amk at mira.erols.com  Fri Dec 22 13:42:21 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 22 Dec 2000 18:42:21 GMT
Subject: mail processing module
References: <mailman.977504177.25049.python-list@python.org>
Message-ID: <slrn9478tg.1q7.amk@207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com>

On Fri, 22 Dec 2000 11:55:13 -0500, Barry A. Warsaw <barry at digicool.com> wrote:
>I've been working on a new package for dealing with message
>objects, including mime messages.  Mostly this is for use in
>Mailman, but they will be generally useful.  I want to

There was also the idea of separating Mailman's low-level mail
processing from the actual list management application; I think I'd
like to pursue in the not-too-distant future.  Features like detecting
all possible bounces can only really be developed through running the
software in production and is hard to get right, so it makes sense to
recycle it.  This is of interest to me because I have an application
that could use an e-mail interface.

Barry, should I start from the current CVS version or from the
2.0final release?  Any major upheavals going on in CVS now?

--amk



From s2mdalle at titan.vcu.edu  Tue Dec 26 22:58:42 2000
From: s2mdalle at titan.vcu.edu (David Allen)
Date: Wed, 27 Dec 2000 08:58:42 +0500
Subject: Python core dumping
Message-ID: <92deoj$b2q$1@bob.news.rcn.net>

I have a medium sized python application which is causing
python to segfault under Linux at very odd times.  I'd
like to get a core dump, so I can try to figure out
what on earth is going on with it, and possibly post
a bug report or something similar.  But as odd as this
sounds, the application will NOT dump core.  The
program I'm running does not use any C extensions,
it's written in pure python.

I've tried using:
Python 1.5.2 (#0, Apr  3 2000, 14:46:48)  [GCC 2.95.2 20000313 (Debian GNU/Linux)] on linux2
and

Python 2.0 (#1, Dec 17 2000, 18:36:14) 
[GCC 2.95.2 20000220 (Debian GNU/Linux)] on linux2

Both after issuing "ulimit -c unlimited" which gives
this:

[x at test pscr]$ ulimit -a
core file size (blocks)     unlimited
data seg size (kbytes)      unlimited
file size (blocks)          unlimited
max locked memory (kbytes)  unlimited
max memory size (kbytes)    unlimited
open files                  1024
pipe size (512 bytes)       8
stack size (kbytes)         8192
cpu time (seconds)          unlimited
max user processes          256
virtual memory (kbytes)     unlimited

So the lack of core dumping is not due to my 
environment as far as I can tell.

I did not compile 1.5.2, that's the default version
that's in Debian GNU/Linux (Woody) right now.  As
for 2.0, I only "./configure"'d it with a --prefix
option.

How can I get python to dump core so I can figure out
what's wrong with my program, or alternatively, what's
wrong with python?  Any help would be appreciated.

-- 
David Allen
http://opop.nols.com/


From jschmitt at vmlabs.com  Thu Dec 14 18:39:15 2000
From: jschmitt at vmlabs.com (jschmitt at vmlabs.com)
Date: Thu, 14 Dec 2000 23:39:15 GMT
Subject: Python advocacy
Message-ID: <91bln3$lmr$1@nnrp1.deja.com>

I just read this good article about Perl advocacy.  I think it applies
to Python equally.  I found the link mentioned from /. this morning.

http://www.perl.com/pub/2000/12/advocacy.html?wwwrrr_20001213.txt#why i
hate advocacy

John


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


From urner at alumni.princeton.edu  Tue Dec 19 14:12:17 2000
From: urner at alumni.princeton.edu (Kirby Urner)
Date: Tue, 19 Dec 2000 11:12:17 -0800
Subject: pow(-1,0) = 1, -1**0 = -1, why?
References: <7lav3t4ptquheuah74s5ff10ghg54u2dhf@4ax.com>
Message-ID: <plcv3tcb1ol0250dlodpeids8ctjtf5nf0@4ax.com>

Kirby Urner <urner at alumni.princeton.edu> wrote:

>
> Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)] on win32
> Type "copyright", "credits" or "license" for more information.
> IDLE 0.6 -- press F1 for help
> 
> >>> -1**0
> -1

But (-1)**0 = 1.

Slaps forehead, duh!

Thanks Joe!, I'll sleep better now. 

-- Kirby




From glandrum at my-deja.com  Wed Dec  6 09:30:42 2000
From: glandrum at my-deja.com (Greg Landrum)
Date: Wed, 06 Dec 2000 14:30:42 GMT
Subject: Sort
References: <NlnX5.1444$yv1.2700363@nnrp5.proxad.net> <oa8s2tsls5foll62r8fpa6nlcpisb6v1ek@4ax.com>
Message-ID: <90liij$805$1@nnrp1.deja.com>

In article <oa8s2tsls5foll62r8fpa6nlcpisb6v1ek at 4ax.com>,
  Steve Horne <sh at ttsoftware.co.uk> wrote:
>
> You can specify an alternative comparison function to the sort method.
> The easiest way to do this is with a lambda...
>
> l.sort (lambda a, b : a[2] < b[2])
>
> For comparison, the following should do the same as the basic sort
> (but slower)...
>
> l.sort (lambda a, b : a < b)
>

Is this something that changed in python 2.0?  The above example
does not work under 1.5.2 where, according to the docs:

>The sort() method takes an optional argument specifying a comparison
>function of two arguments (list items) which should return -1, 0 or 1
>depending on whether the first argument is considered smaller than,
>equal to, or larger than the second argument.

So the answer to the original question would be:

l.sort (lambda a, b : a[2] - b[2])

-greg


Sent via Deja.com http://www.deja.com/
Before you buy.


From mail at codex.pe.kr  Mon Dec 18 04:02:08 2000
From: mail at codex.pe.kr (Ahn, Jae-wook)
Date: Mon, 18 Dec 2000 18:02:08 +0900
Subject: Database solution for Mac?
Message-ID: <91kjtq$ns3$1@news1-2.kornet.net>

Hello,

I have an iBook. I usually do works on Linux or Windows, but sometimes I
need some database support for my small Mac machine.

I visited some pages, and found out a lot of database support for Python
including ODBC, but could not see anything that supports Macintosh. Please
recommend any solution which support database for Mac. Networked or local.

Ahn





From pbjorn at uswest.net  Wed Dec 13 12:40:31 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Wed, 13 Dec 2000 10:40:31 -0700
Subject: Dijkstra's Shortest Path algorithm
References: <Pine.GSO.4.21.0012090042370.15110-100000@y.glue.umd.edu> <dvrd3t40u9l7atgdro59m5ttg7dsfv7c4i@4ax.com> <Pine.GSO.4.21.0012122249480.19682-100000@y.glue.umd.edu>
Message-ID: <3A37B48F.FE8F112@uswest.net>

Roy Katz wrote:

> [snip]

> btw -- we're holding a competition for the fastest program in my C++
> class.  I'm probably in the top four in my section (1.77 seconds for 500
> nodes with 10k edges, 1 query). The average for my class is around 20
> seconds (this was the final project).  Weird! the biggest lags occur in
> reading in the data! but I guess, that's what we get with
> object orientation -- more layers to tunnel through.

Yes, unfortunately, if you want fast IO you shouldn't use the iostreams
library (the new standards compliant one is even slower than the old
iostreams.h version since it has to go through locale *sigh*).

fopen/fread etc. are still available when you need speed (and usually much
easier on the eye also <wink>).

-- bjorn



From moshez at zadka.site.co.il  Thu Dec 28 07:17:19 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: 28 Dec 2000 12:17:19 -0000
Subject: Is this a BUG?
In-Reply-To: <3A4B0C5E.1E7976D2@wielun.dhs.org>
References: <3A4B0C5E.1E7976D2@wielun.dhs.org>
Message-ID: <20001228121719.2375.qmail@stimpy.scso.com>

On Thu, 28 Dec 2000 10:48:14 +0100, QdlatY <qdlaty at wielun.dhs.org> wrote:

> print "!",form["index"].value,"!"
...prints...
> ! 1 !
> ----------------
>  ^ ^
> Why the hell there are spaces? index property has value '1'!
> not ' 1 '?!

Because print prints a space between its arguments.
The print statement is equivalent to 
``sys.stdout.write(" ".join("!", '1', "!")+'\n')'' 

Use sys.stdout.write for finer grain control, or

print "!%s!" % form['index']




From urner at alumni.princeton.edu  Tue Dec 19 11:33:50 2000
From: urner at alumni.princeton.edu (Kirby Urner)
Date: Tue, 19 Dec 2000 08:33:50 -0800
Subject: Python in "math through programming" curriculum
References: <2g2t3t06r2agrrjmfnr41eu4lc2u39a8qi@4ax.com> <3A3EFF47.67DE69CF@gte.net>
Message-ID: <513v3toslkjkggh9oumlf20qc9bi1gdtif@4ax.com>

"Clifford J. Nelson" <cnelson9 at gte.net> wrote:

>I don't think Python should be a standard at all, because of
>my experience with it on the Mac. Keep the options open for
>computer programming languages.
>

Keeping options open is fine (using Python to open doors
to other languages is a good way to go as well).  I don't
use Python for much of my programming work, but as one 
who used to write 8th grade level "math through programming" 
stuff for McGraw-Hill in the 1980s (using BASIC and Logo), 
I can say without a doubt that Python would have been 
pure gold for us.

Python on the Mac was disappointing for me for the brief
30 minutes I tried it.  IDLE is a great learning environment
and on Windows and Linux is in color, but on the Mac seemed 
to be all one color.  Was I missing a setting?  Maybe the Mac
people are working on alternatives/enchancements to IDLE?
Does anyone have some info on this?

>Both of those languages are OOP too, and easier to use than Python.
>

How much personal experience to you have with all three,
or are you reporting the assessments of others?  Do you have
a website where you show your work, review your experiences?  
I'm not sure I need to attach much weight to any of your 
opinions.  Others around here (such as the previous respondant 
to your post) seem to know a great deal more about these 
languages than you do, and I've never seen you post anything 
that suggests you've done much with Python, either.

Kirby



From mitch at garnaat.com  Wed Dec  6 08:29:06 2000
From: mitch at garnaat.com (Mitchell Garnaat)
Date: Wed, 06 Dec 2000 13:29:06 GMT
Subject: CGI File Upload
References: <p8ce0tcgtt2bksb3adrf9cfr2i8egr4ff5@4ax.com> <8u7fn7$jpf$1@la-mail4.digilink.net> <90j10o$5p5$1@nnrp1.deja.com>
Message-ID: <CarX5.5051$Ia1.421476@typhoon.nyroc.rr.com>

It sounds like you are on a PC platform and the input file is not being
opened in binary mode.  The default on Windows platforms is to open the file
in text mode.  When you hit the first zero byte in the binary file it will
be interpreted as an end-of-file.  Look in the CGI code to see where it is
opening files and make sure they are being opened in binary mode.

Mitch


"Bill Seitz" <fluxent at yahoo.com> wrote in message
news:90j10o$5p5$1 at nnrp1.deja.com...
> In article <8u7fn7$jpf$1 at la-mail4.digilink.net>,
>   "Pete Shinners" <pete at visionart.com> wrote:
> > "Simon Faulkner" <news at titanic.co.uk> wrote
> > > I would like users to be able to upload .jpeg files via a web page
> > > <input type=file>
> > >
> > > What python code do I need to be able to save the file to a
> directory
> > > on the hard disk?
> >
> > i have some code that does this.. you'll probably want to clean
> > it to suit your own purposes, but here it is. at some point i
> > wanted to add PIL support to resize the images to a certain
> > resolution and convert them to be JPG, but that day hasn't come
> > yet... heh. note it also allows the user to save a comment as a
> > .TXT file. anyways, you can use it as a start for whatever you want
>
> Hmm, this works for me with text data (though the file length is
> slightly different: maybe line-end changes from Win to *nix?). With
> binary it's totally off. Uploading a JPEG I consistently only get 109
> bytes written to the file system. With GIFs I consistently get ~500
> bytes (but not always exactly the same number). With an MsWord file I
> get a bigger file written, but still not the whole thing (60kb out of
> 75kb).
>
> What do I do next?
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.




From ljz at asfast.com  Sat Dec 16 22:56:03 2000
From: ljz at asfast.com (Lloyd Zusman)
Date: Sun, 17 Dec 2000 03:56:03 GMT
Subject: Alarms in threads - how to get/mimic?
References: <Pine.LNX.4.21.0012161749380.7913-100000@traal.eecs.berkeley.edu>
Message-ID: <ltd7erk8xd.fsf@asfast.com>

Luca de Alfaro <dealfaro at eecs.berkeley.edu> writes:

> [ ... ]
> 
> What is not clear to me is the following.  When there is a timeout,
> the exception Timeout is not caught by httplib.  Hence, if the thread
> that called httplib just continues its job giving up on the
> connection, httplib will not have a chance to do any clean-up, such as
> for example releasing/closing the socket. 
> 
> Hence, it seems to me that in order to abort a web page load cleanly
> upon timeout, I also have to modify the httplib module to handle the
> exception Timeout, and do clean-up work if that happens. 
> 
> Is my understanding correct? 

Not entirely.  When using timeoutsocket, all you have to do is start
catching timeout exceptions when you make the various method calls
within httplib.  As an example, I'll illustrate the use of
timeoutsocket with some sample code that I've based on a program of
mine that uses imaplib.  You could do exactly the same kind of thing
when using httplib:

   import sys, traceback, imaplib
   import timeoutsocket    # this needs to be imported prior to
                           # making any socket calls

   # ... snip ...

   # the following line sets a 15 second timeout for the
   # subsequent socket calls made within imaplib

   timeoutsocket.setDefaultSocketTimeout(15.0)

   try:
      connection = imaplib.IMAP4(host)
   except timeoutsocket.Timeout, msg:
      # this "except" block will be invoked if the
      # imaplib initialization fails
      print "imap initialization timed out"
   except:
      # this "except" block will be invoked if there
      # is any other exception that occurs during
      # imap initialization
      typ, val, tb = sys.exc_info()
      print "imap initialization exception: %s %s\ntraceback:" % \
         ( typ, val )
      traceback.print_tb(tb)
   else:
      # the imaplib connection is successful      
      print "successful imaplib connection"

It's recommended in Python that exception handling be done whenever
it's appropriate; therefore, by wrapping your httplib calls in this
same manner, you'll remain true to the Pythonic Way, and you won't
have to re-write any httplib code when using timeoutsocket.

Note that the timeoutsocket.setDefaultSocketTimeout() method call
should be made before making any httplib calls.  If you want to use
different timeout values when making different httplib calls (or other
calls socket-based methods), you'll have to re-invoke the
setDefaultSocketTimeout() method before each call that uses a changed
timeout value.

HTH

- Lloyd


-- 
 Lloyd Zusman
 ljz at asfast.com


From cbray at orchestream.com  Fri Dec 22 08:07:50 2000
From: cbray at orchestream.com (Carl Bray)
Date: Fri, 22 Dec 2000 13:07:50 -0000
Subject: Passing \r to PyRun_SimpleString syntax error?
Message-ID: <977490691.294533@lave>

I'm wondering if this is a bug??

We've embedded python into our application, we're then passing in a string
converted from a file. This file can be created on either Windows or UNIX.

We have found that passing in a \r 0x0D char causes the parser to return a
syntax error. This seems inconsistent with the Lexical analysis in the
reference manaul ...

2.1 Line structure
A Python program is divided into a number of logical lines.
2.1.1 Logical lines
The end of a logical line is represented by the token NEWLINE. Statements
cannot cross logical line boundaries
except where NEWLINE is allowed by the syntax (e.g., between statements in
compound statements). A logical line
is constructed from one or more physical lines by following the explicit or
implicit line joining rules.
2.1.2 Physical lines
A physical line ends in whatever the current platform's convention is for
terminating lines. On UNIX, this is the
ASCII LF (linefeed) character. On DOS/Windows, it is the ASCII sequence CR
LF (return followed by linefeed). On
Macintosh, it is the ASCII CR (return) character.




From mikael at isy.liu.se  Wed Dec 13 02:54:24 2000
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Wed, 13 Dec 2000 08:54:24 +0100 (MET)
Subject: while true: !!!
In-Reply-To: <VrtZ5.3567$Kd1.415273@newsb.telia.net>
Message-ID: <XFMail.001213085424.mikael@isy.liu.se>

On 12-Dec-00 Fredrik Lundh wrote:
 >  (well, in the current implementation, it isn't: if your file
 >  has more than 2147483647 lines, your code won't read
 >  all of them... ;-)

Let's hope that your computer is up that long <wink>.

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    13-Dec-00
Time:    08:52:22

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

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


From teyc at bigfoot.com  Mon Dec 11 10:02:31 2000
From: teyc at bigfoot.com (Chui Tey)
Date: Tue, 12 Dec 2000 01:02:31 +1000
Subject: Modifying running code? [HACK ALERT]
References: <912m65$pu1$1@news.rz.uni-karlsruhe.de>
Message-ID: <912qdf$b6s$1@bunyip.cc.uq.edu.au>

Doesn't quite answer the question, but ...

I read somewhere that VB compiles to an intermediate step before p-code is
generated. This technique allowed minor changes to the source when a program
is being run. Didn't seem to need continuations here.





From tbryan at python.net  Mon Dec 11 10:15:20 2000
From: tbryan at python.net (Thomas A. Bryan)
Date: Mon, 11 Dec 2000 10:15:20 -0500
Subject: Bug: Readline getting stuck on Linux and Solaris
References: <lxk8zqwupy2.fsf@ioasun3.epfl.ch> <90m53k$ovh$1@nnrp1.deja.com> <m3u28h48ja.fsf@atrus.jesus.cam.ac.uk> <slrn92u8al.76b.quinn@lira.ugcs.caltech.edu> <m3r93k4kqk.fsf@atrus.jesus.cam.ac.uk> <slrn932eoj.d1j.quinn@ngwee.ugcs.caltech.edu> <m34s0c4ik7.fsf@atrus.jesus.cam.ac.uk> <m31yvg4i0i.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <3A34EF88.20E84C43@python.net>

Michael Hudson wrote:

> But I've just d/l-ed readline and it's signal handler is (predictably)
> straight out of Preprocessor Hell, so I can't tell what's going on at
> all :-(.

If you know the flags on the target machine (Irix?), then you could run 
gcc with the -E option so that you can see what is left of the source 
after preprocessing.  If Quinn sent you the files that "make configure" 
creates on his box, then you should be able to get the preprocessed 
readline source that corresponds to his setup.

Perhaps makefiles should have a dummy target using the -E and -o options 
so that people could preprocess the entire tree and get the source code 
after the preprocessor is done with it.

---Tom


From tim.one at home.com  Sun Dec 24 14:16:43 2000
From: tim.one at home.com (Tim Peters)
Date: Sun, 24 Dec 2000 14:16:43 -0500
Subject: A Suggestion for Python Colon Syntax
In-Reply-To: <20001224010638.J5515@xs4all.nl>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEGFIFAA.tim.one@home.com>

[Tim]
> [...] (ditto the whole "while 1" thread
> that's been going on virtually non-stop for a decade).

[David C.Ullrich]
> and we'll never be certain whether it was an intentional
> joke or not. Keen.

[Thomas Wouters]
> No, I'm certain Tim did not make that horrible, horrible pun by
> accident. No way could he inflict such amounts of nausea and pain
> by accident. Just not possible :)

Winkery seems to irk some people so much that I have been experimenting with
leaving it out half the time.  It's a Bad Idea, I'm afraid.  Witness poor
Greg Ewing, who recently got beat up even in the presence of an explicit
";-)" thingie.  So I resolve to redouble my explicit winking efforts in
2001, and  those who can detect humor without it will just have to live with
the clues <wink>.

"while-1"-is-a-wonderfully-self-referential-thread-though-ly y'rs  - tim






From slhath at home.com  Tue Dec 19 19:41:07 2000
From: slhath at home.com (Scott Hathaway)
Date: Wed, 20 Dec 2000 00:41:07 GMT
Subject: xml-rpc module question
References: <dmC%5.75166$yR4.2151642@news1.rdc1.tx.home.com> <91ooea$9qb$1@bunyip.cc.uq.edu.au>
Message-ID: <DeT%5.76385$yR4.2212474@news1.rdc1.tx.home.com>

Chui,

Thanks for the response.  When I try your suggested code, here is the error
I get:

C:\Projects\core>python core_client.py
Traceback (most recent call last):
  File "core_client.py", line 5, in ?
    print testing.CubeIt(3)
  File "xmlrpclib.py", line 547, in __call__
    return self.__send(self.__name, args)
  File "xmlrpclib.py", line 627, in __request
    response = self.__transport.request(
AttributeError: 'int' object has no attribute 'request'


Any ideas?

Scott

"Chui Tey" <teyc at bigfoot.com> wrote in message
news:91ooea$9qb$1 at bunyip.cc.uq.edu.au...
> Not tested code:
>
> from xmlrpcserver import RequestHandler
>
> class my_handler(RequestHandler):
>
>     def CubeIt(self, x):
>         return x ** 3
>
>     def call(self, method, params):
>         if method == 'CubeIt':
>             return self.CubeIt( params[0] )
>
> if __name__ == '__main__':
>     server = SocketServer.TCPServer(('', 8000), my_handler)
>     server.serve_forever()
>
>
> Chui
>
>
>
>
>




From fgeiger at datec.at  Tue Dec  5 08:32:20 2000
From: fgeiger at datec.at (Franz GEIGER)
Date: Tue, 5 Dec 2000 14:32:20 +0100
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Dec  4)
References: <8C419A8040EAFDCB.1774103991763D9E.2F034C54E78FF82C@lp.airnews.net>
Message-ID: <90iqs8$n6k$1@newsreaderm1.core.theplanet.net>

>     Alex Martelli shows a clever way to overload a method based on
>     number of arguments.
>         http://deja.com/=dnc/getdoc.xp?AN=975947338

This link leads to "Invalid document request". How is the right one?

Thanks in advance and
best regards
Franz GEIGER


Grant Edwards <grante at visi.com> schrieb in im Newsbeitrag:
8C419A8040EAFDCB.1774103991763D9E.2F034C54E78FF82C at lp.airnews.net...
>     Alex Martelli discusses the issues involved in using Python
>     within Windows Script Control.
>         http://deja.com/=dnc/getdoc.xp?AN=698664345
>     Alex Martelli shows a clever way to overload a method based on
>     number of arguments.
>         http://deja.com/=dnc/getdoc.xp?AN=975947338
>
>     Dale Strickland-Clark and Alex Martelli on using Python
>     programs as drop targets under Win32.
>         http://deja.com/=dnc/getdoc.xp?AN=700537932
>         http://deja.com/=dnc/getdoc.xp?AN=700557153
>
>     Andrew H. Chatham announces Python bindings for Ogg/Vorbis (MP3
>     patent/licensing free replacement for MP3) -- and requests help
>     with dlopen().
>         http://deja.com/=dnc/getdoc.xp?AN=975946985
>
>     Nick Mathewson shows how to implement a general overloading
>     module for python.
>         http://deja.com/=dnc/getdoc.xp?AN=700076223
>
>     Gordon McMillan releases fixed Installer for Python 2.0
>         http://deja.com/=dnc/getdoc.xp?AN=700179180
>
>     Raymond Hettinger shows a module to provide dynamic inheritance
>     and convenient composition.
>         http://deja.com/=dnc/getdoc.xp?AN=700743048
>
>     Johannes Nix releases Python wrappers for ALSA (Advanced Linux
>     Sound Architecture) that enable real-time audio processing.
>         http://deja.com/=dnc/getdoc.xp?AN=975947772
>
>
> ========================================================================
>
> Everything you want is probably one or two clicks away in these pages:
>
>       Python.org's Python Language Website is the traditional
>       center of Pythonia
>           http://www.python.org
>       although PythonLabs.com bills itself as "The Python Source",
>       which is becoming increasingly true in 2000
>           http://www.pythonlabs.com/
>
>       PythonWare complements the digest you're reading with the
>       daily python url
>            http://www.pythonware.com/daily
>
>       comp.lang.python.announce announces new Python software.  Be
>       sure to scan this newly-revitalized newsgroup at least weekly.
>           http://deja.com/group/comp.lang.python.announce
>
>       Andrew Kuchling writes marvelous summaries twice a month of
>       the action on the python-dev mailing list, where the future
>       of Python is truly determined
>   http://www.amk.ca/python/dev
>
>       The Vaults of Parnassus ambitiously collects Python resources
>           http://www.vex.net/~x/parnassus/
>
>       Much of Python's real work takes place on Special-Interest Group
>       mailing lists
>           http://www.python.org/sigs/
>
>       Python Consortium emerges as an independent nexus of activity
>           http://www.python.org/consortium
>
>       Cetus does much of the same
>           http://www.cetus-links.de/oo_python.html
>
>       Python FAQTS
>           http://python.faqts.com/
>
>       Python To-Do List anticipates some of Python's future direction
>           http://www.python.org/cgi-bin/todo.py
>
>       Python Journal is at work on its second issue
>           http://www.pythonjournal.com
>
>       Links2Go is a new semi-automated link collection; it's impressive
>       what AI can generate
>           http://www.links2go.com/search?search=python
>
>       Archive probing trick of the trade:
>
http://www.dejanews.com/dnquery.xp?QRY=&DBS=2&ST=PS&defaultOp=AND&LNG=ALL&fo
rmat=threaded&showsort=date&maxhits=100&groups=comp.lang.python
>
>
> Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
>   http://purl.org/thecliff/python/url.html
> or
>
http://www.dejanews.com/dnquery.xp?QRY=~g%20comp.lang.python%20Python-URL%21
>
>
> Suggestions/corrections for next week's posting are always welcome.
> [http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
> e-mail us ideas directly.]
>
> To receive a new issue of this posting in e-mail each Monday morning,
> ask <claird at neosoft.com> to subscribe.  Mention "Python-URL!".
>
>
> -- The Python-URL! Team--
>
> Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
> sponsor the "Python-URL!" project.
> --
>
> Cameron Laird <claird at NeoSoft.com>
> Business:  http://www.Phaseit.net
> Personal:  http://starbase.neosoft.com/~claird/home.html




From gregj at pobox.com  Sun Dec 24 04:28:12 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Sun, 24 Dec 2000 09:28:12 GMT
Subject: string split
References: <slrn94b2go.srm.jp@localhost.localdomain>
Message-ID: <Mkj16.188569$U46.5832132@news1.sttls1.wa.home.com>

"Jacek Pop?awski" <jp at ulgo.koti.com.pl> wrote:

> I have string:
>
> s="one two <br> three"
> ...
> is it possible to split not by space (" ") but by "<" and ">" ?
>
> I need:
>
> ['one two','<br>','three']
>
> How to do it in simple way?

The regular expression (re) module split function may be what you want:

    import re

    # pattern matches <...>, but doesn't handle nested < ... < ...> ... >
    # the \s* serve to remove any whitespace before and after the < ... >
    rx = re.compile(r'\s*(<[^<>]*>)\s*')

    s = "one two <br> three"
    rx.split(s)        # ['one two', '<br>', 'three']

This won't handle nested <...>, e.g. <a <b>>, and it will insert an empty
element between adjacent <...><...> substrings, or before/after <...>
substrings that appear at the beginning or end of the string. Empty elements
can be removed with a filter() or a list comprehension:

    s = "one two <br><br> three<br>"
    t = rx.split(s)            # t=['one two', '<br>', '', '<br>', 'three',
'<br>', '']
    t = [i for i in t if i]    # t = ['one two', '<br>', '<br>', 'three']

or

    t = filter(lambda i: i, t)    # t = ['one two', '<br>', '<br>', 'three']

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From glenfant at equod.com.nospam  Wed Dec 13 15:10:28 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Wed, 13 Dec 2000 21:10:28 +0100
Subject: Source Code Management
References: <fb4f3tsgffnmvme64aol1cn397te7ln349@4ax.com> <918711$meh$1@reader1.imaginet.fr> <prff3tc59qtv0qr9afj16dali3jv789c7u@4ax.com>
Message-ID: <918krp$pls$1@reader1.imaginet.fr>

Dale,

RTFM ;-)
It should work (but I did not test with Windows CVS server)
Or do as we do, carry on using Windows CVS clients but use Linux servers
(more efficient for non M$ server softwares).

"Dale Strickland-Clark" <dale at out-think.NOSPAMco.uk> a ?crit dans le message
news: prff3tc59qtv0qr9afj16dali3jv789c7u at 4ax.com...
> "Gilles Lenfant" <glenfant at equod.com.nospam> wrote:
>
> >Have a look to cvs at http://www.gnu.org/software/cvs/cvs.html
> >http://www.wincvs.org
> >http://www.cvshome.org/
> >
> >The best of the open source version control tool.
> >There are python wrappers to CVS somewhere, I don't remember where.
> >
>
> I've just dug CVS out and downloaded it but I discover that the NT version
appears to require a
> mounted volume.
>
> That's a disappointment.
>
>
> --
> Dale Strickland-Clark
> Out-Think Ltd
> Business Technology Consultants



From clarence at netlojix.com  Wed Dec 27 13:23:51 2000
From: clarence at netlojix.com (Clarence Gardner)
Date: Wed, 27 Dec 2000 10:23:51 -0800
Subject: One Python 2.1 idea
References: <mailman.977905875.10357.python-list@python.org>
Message-ID: <977941508.74720040@news.silcom.com>

On Wed, 27 Dec 2000, Tim Peters wrote:
>
>This is pretty deep magic!  Don't go doing things like "a.f=a.f", and you
>can safely ignore it for decades.
>
>and-writing-"x=x"-can-erase-your-hard-drive<wink>-ly y'rs  - tim

I see now that Ayn Rand might have been a closet Python Wizard,
fond as she was of saying "A is A" :)



From neilh at scintilla.org  Wed Dec 13 17:49:26 2000
From: neilh at scintilla.org (Neil Hodgson)
Date: Wed, 13 Dec 2000 22:49:26 GMT
Subject: Problems With socket Under NT & PythonWin 2.0
References: <910jtt$nnh$1@nnrp1.deja.com> <fMQY5.12943$hp.1457332@typhoon.nyroc.rr.com> <915k4g$kcc$1@nnrp1.deja.com> <8QzZ5.1711$7g.340689@typhoon.nyroc.rr.com> <3A37F5FA.3020209@ActiveState.com>
Message-ID: <W1TZ5.16500$xW4.133590@news-server.bigpond.net.au>

Mark Hammond:
> Darrell wrote:
> ...
> > PythonWin could have started your app in another thread to avoid this.
>
> Exactly.  In general, this problem will exist for all IDEs that run code
> in the same process.  For this reason, most IDEs are moving to
> out-of-proc debugging.  Komodo has only out-of-proc debugging (and works
> very nicely too)

   While out-of-proc is good for debugging, I like also having the ability
to run things that change the main PythonWin environment. Makes it really
flexible and allows exploration with a combination of running stuff in files
and from the Interactive Window. Then you don't have to build throw-away
test scripts that starts from zero.

   Neil





From dsh8290 at rit.edu  Wed Dec 27 18:14:37 2000
From: dsh8290 at rit.edu (D-Man)
Date: Wed, 27 Dec 2000 18:14:37 -0500
Subject: classes (was Re: Same again please for OOP)
In-Reply-To: <92708d0g4g@news1.newsguy.com>; from aleaxit@yahoo.com on Mon, Dec 25, 2000 at 09:20:51AM +0100
References: <mailman.977623149.8184.python-list@python.org> <9257mf0sl1@news1.newsguy.com> <roy-E5B925.13324224122000@news.panix.com> <925khi019hs@news1.newsguy.com> <kubsu1t81x.fsf@lasipalatsi.fi> <92708d0g4g@news1.newsguy.com>
Message-ID: <20001227181437.A11403@westsidecnc.rh.rit.edu>

(Sorry if this posts twice, but I don't see my earlier post)

Originally I was in agreement with Erno, but Alex has convinced me
that Python has a better way.  I think it would be really cool if the
interpreter would automatically look for the __get_* function when a
client access a member, but for now Alex's solution will have to
suffice.

However, I can't get this sample code to work.  I keep getting
AttributeError's or infinitely recursive __getattr__ function.  If I
remove one of the leading underscores it works fine.  This is because
of the current name mangling used to "enforce" private members.

Am I missing something here, or should I only use 1 leading
underscore?

Thanks,
-D

On Mon, Dec 25, 2000 at 09:20:51AM +0100, Alex Martelli wrote:
<snip>
> class AccessorsMixin:
>     def __getattr__(self, name):
>         if name.startswith('__'): raise AttributeError,name
>         return getattr(self, '__get_'+name)()
> 
> and implement method __get_location(self) the same way as you'd
> implement location(self) in your style.  What could be easier?  If
> you already have a __getattr__, it's equally OK to add this couple
> of lines to it, or delegate to this one (delegating has pluses).
<snip>



From DavidL.nono.nospam at raqia.com  Thu Dec  7 15:08:10 2000
From: DavidL.nono.nospam at raqia.com (David Lees)
Date: Thu, 07 Dec 2000 15:08:10 -0500
Subject: Simple HTTP client bug?
Message-ID: <3A2FEE2A.D1A0A645@raqia.com>

I am trying to run the following simple HTTP client code, copied
directly from "Programming in Python with Medusa and the Async Sockets
Library", but I get the following error every time I run it (regardless
of what URL I use or whether I run under NT or Linux Redhat 6.2).  Can
someone tell me what I am doing wrong?

Thanks in advance.

david lees


>>> host,path= www.nightmare.com
starting asyncore loop
sending GET
GET  HTTP/1.0


<!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML 2.0//EN">
<HTML><HEAD>
<TITLE>400 Bad Request</TITLE>
</HEAD><BODY>
<H1>Bad Request</H1>
Your browser sent a request that
this server could not understand.<P>
Invalid URI in request GET  HTTP/1.0<P>
</BODY></HTML>

warning: unhandled close event



----
The code is:

# -*- Mode: Python; tab-width: 4 -*-

import asyncore
import socket
import string

class http_client (asyncore.dispatcher):

    def __init__ (self, host, path):
        asyncore.dispatcher.__init__ (self)
        self.path = path
        self.create_socket (socket.AF_INET, socket.SOCK_STREAM)
        self.connect ((host, 80))

    def handle_connect (self):
        print 'sending GET'
        print 'GET %s HTTP/1.0\r\n\r\n' % self.path
        self.send ('GET %s HTTP/1.0\r\n\r\n' % self.path)

    def handle_read (self):
        data = self.recv (8192)
        print data

    def handle_write (self):
        pass

if __name__ == '__main__':
    import sys
    import urlparse
    for url in sys.argv[1:]:
        parts = urlparse.urlparse (url)
        if parts[0] != 'http':
            raise ValueError, "HTTP URL's only, please"
        else:
            host = parts[1]
            path = parts[2]
            print 'host,path=',host,path
            http_client (host, path)
            print 'starting asyncore loop'
    asyncore.loop()


From andreas at andreas-jung.com  Mon Dec 18 07:05:46 2000
From: andreas at andreas-jung.com (Andreas Jung)
Date: Mon, 18 Dec 2000 13:05:46 +0100
Subject: cron with sched module
In-Reply-To: <91dgfa$q54$1@reader1.imaginet.fr>; from glenfant@equod.com.nospam on Fri, Dec 15, 2000 at 05:26:33PM +0100
References: <91dgfa$q54$1@reader1.imaginet.fr>
Message-ID: <20001218130545.A9245@yetix.sz-sb.de>

On Fri, Dec 15, 2000 at 05:26:33PM +0100, Gilles Lenfant wrote:
> I'm trying hardly to make a python "cron" like utility that may run
> functions/methods periodically (say every x seconds) and predetermined
> hours.
> 
> I can't use Window "AT" or Unix "cron" cause some functions/methods must be
> executed evry 15 seconds. Launching Python executable periodically such a
> manner will eat the disk I/O and
> 
> So, I'm trying with the "sched" module to do this but I can't add  "enter"
> to a scheduler object already running.
> So, what's the workaround ?
> Use directly lower level "threading" module in lace of sched ?

I suggest to use the thread or threading module. The threading
provides a higher level interface (it's build on the top
of the thread module). The way how to implement a solution for
your needs depends on your needs. E.g. you can create a worker
threads that runs in a loop. To periodic execution you could
use time.sleep() to delay execution for a period of time.

Andreas 



From frankn=nuws at cs.vu.nl  Mon Dec 25 09:02:11 2000
From: frankn=nuws at cs.vu.nl (Frank Niessink)
Date: Mon, 25 Dec 2000 14:02:11 GMT
Subject: another newbie question - regex
References: <slrn94eju1.1fk.jp@localhost.localdomain>
Message-ID: <927k13$kff@cs.vu.nl>

Jacek Pop?awski <jp at ulgo.koti.com.pl> wrote:

> then why:

>>>> print re.search(s,".*")
> None
>>>> print re.match(s,".*")
> None


Try reverse the arguments:

>>> import re
>>> s = 'Hi there!'
>>> m = re.search(".*", s)
>>> m.group()
'Hi there!'


Cheers, Frank
-- 
Reverse engineering. To me the flow of time is irrelevant. You decide what you
want. I then merely make sure that it has already happened.
	-- Douglas Adams, 'Mostly Harmless'


From dmlopez at ucauca.edu.co  Sun Dec 17 08:56:59 2000
From: dmlopez at ucauca.edu.co (=?iso-8859-1?Q?diego_L=F3pez?=)
Date: Sun, 17 Dec 2000 08:56:59 -0500
Subject: problems Mailman instalation
Message-ID: <00c901c06831$39ad0360$02ff10ac@ucauca.edu.co>

I followed all the step for the instalation, but when i create the test list,  when i am the list administrator, it retive to my email address failed deliveries once per minute.
The e-amil is like:


Traceback (innermost last):
  File "/usr/mailman/cron/qrunner", line 271, in ?
    lock.lock(timeout=0.5)
  File "/usr/mailman/Mailman/LockFile.py", line 219, in lock
    self.__write()
  File "/usr/mailman/Mailman/LockFile.py", line 350, in __write
    fp = open(self.__tmpfname, 'w')
IOError: [Errno 13] Permission denied: '/usr/mailman/locks/qrunner.lock.orion.2929'


I think the problem should be with directory where i have istalled the aplication (usr/mailman/)


i did-nt checked it it is with nosuid option, because i dont know how to konw it.
Thanks
===========================================================
Diego Mauricio L?pez G .
Docente Departamento de Conmutaci?n
(Ingenier?a Telem?tica)
Facultad de Ing. Electr?nica - Universidad del Cauca. Popay?n- Colombia
Tel: (+57-28) 233031 ext 192 Fax: (+57-28) 233031
website: http://www.ucauca.edu.co/~dmlopez
e-mail: dmlopez at ucauca.edu.co 
           dmlopez at halley.ucauca.edu.co
==========================================================

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20001217/de4a54ce/attachment.html>

From nospam at no.spam  Tue Dec 19 11:25:50 2000
From: nospam at no.spam (Carsten Geckeler)
Date: Tue, 19 Dec 2000 17:25:50 +0100 (CET)
Subject: while true: !!!
In-Reply-To: <91nc7h0s0p@news2.newsguy.com>
Message-ID: <Pine.LNX.4.21.0012191723380.30860-100000@justitia.jura.uni-tuebingen.de>

On Tue, 19 Dec 2000, Alex Martelli wrote:

> "Rainer Deyke" <root at rainerdeyke.com> wrote in message
> news:KTz%5.48644$x6.22165257 at news2.rdc2.tx.home.com...
> > "Steve Lamb" <grey at despair.rpglink.com> wrote in message
> > news:slrn93t7bv.le.grey at teleute.rpglink.com...
> > > On Mon, 18 Dec 2000 23:32:23 +0100 (MET), Carsten Geckeler
> > <nospam at no.spam>
> > > wrote:
> > > >`while true' would be more readable and clear than `while 1'.  By the
> > way,
> > > >explain a newbie that -1 is true. ;)
> > >
> > >     0 is true, all else is false.  And?  :P
> >
> > 0 is false, None is false, [] is false, () is false, {} is false, and x is
> > false where x is a class instance and x.__nonzero__() returns false.
> > Everything else is true.  I think.
> 
> Another couple of 'false's: a class instance which doesn't define
> __nonzero__ but does define __length__ when the latter returns 0;
> and, C-implemented user objects, similarly to class instances.
> 
> I don't _think_ there are any more...

Nice; It takes three tries of -- I assume -- experienced Python
programmers to find all cases of all statements evaluating as false. ;)

Cheers, Carsten
-- 
Carsten Geckeler:  carsten dot geckeler at gmx dot de
To get proper email-address replace `dot' and `at' by the corresponding symbols.




From ingoogni at .home.nl  Sat Dec  9 10:59:29 2000
From: ingoogni at .home.nl (ingo)
Date: Sat, 09 Dec 2000 15:59:29 GMT
Subject: PIL binary for Python 2.0 for Windows ?
References: <3a2f472f$0$235@hades.is.co.za> <KZOX5.2015$Kd1.294314@newsb.telia.net> <IlSX5.24763$R77.1876538@nnrp4.clara.net> <9JnY5.2716$O%.428568@newsc.telia.net>
Message-ID: <Xns9005AB80Aseed7@news1.zwoll1.ov.nl.home.com>

Fredrik Lundh wrote:

>also see the updated pil page:
>    http://effbot.org/pil

When I use these and run the viewer.py demo from 1.1.1 scource I get an 
error:

Traceback (most recent call last):
  File "D:\Ingo\Download\viewer.py", line 44, in ?
    UI(root, im).pack()
  File "D:\Ingo\Download\viewer.py", line 23, in __init__
    self.image = ImageTk.PhotoImage(im)
  File "d:\programming\python20\pil\ImageTk.py", line 86, in __init__
    self.paste(image)
  File "d:\programming\python20\pil\ImageTk.py", line 123, in paste
    import _imagingtk
ImportError: No module named _imagingtk

Did I miss something?

Ingo

-- 
Photography: http://members.home.nl/ingoogni/
Pov-Ray    : http://members.home.nl/seed7/


From gherron at aw.sgi.com  Thu Dec 28 19:57:36 2000
From: gherron at aw.sgi.com (Gary Herron)
Date: Thu, 28 Dec 2000 16:57:36 -0800
Subject: how do i check if a file exists with python?
References: <RlM26.98766$yR4.2639444@news1.rdc1.tx.home.com>
Message-ID: <3A4BE180.FF6AFEC5@aw.sgi.com>

Scott Hathaway wrote:
> 
> I have searched Deja, looked in two pyton books, and searched the python.org
> site trying to find out this simple question.
> 
> Can someone please give me the call and the module it lives in?
> 
> Thanks,
> Scott
> 
> --
> http://www.python.org/mailman/listinfo/python-list

This should work for any recent version of Python:
  import os.path
  os.path.isfile('somefile')

See the documentation of module os.path for several other isXXX
procedures.

-- 
Dr. Gary Herron <gherron at aw.sgi.com>
206-287-5616
Alias | Wavefront
1218 3rd Ave, Suite 800, Seattle WA 98101



From aleaxit at yahoo.com  Wed Dec  6 06:57:33 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 6 Dec 2000 12:57:33 +0100
Subject: Sort
References: <NlnX5.1444$yv1.2700363@nnrp5.proxad.net> <oa8s2tsls5foll62r8fpa6nlcpisb6v1ek@4ax.com>
Message-ID: <90l9ir02avk@news2.newsguy.com>

"Steve Horne" <sh at ttsoftware.co.uk> wrote in message
news:oa8s2tsls5foll62r8fpa6nlcpisb6v1ek at 4ax.com...
> On Wed, 06 Dec 2000 09:07:57 GMT, "Alfred" <scj at mcom.mcom.fr> wrote:
>
> >I'm trying to sort a List of List, on the third col, but I don't find how
to
> >do this.
>
> You can specify an alternative comparison function to the sort method.
> The easiest way to do this is with a lambda...
>
> l.sort (lambda a, b : a[2] < b[2])
>
> For comparison, the following should do the same as the basic sort
> (but slower)...
>
> l.sort (lambda a, b : a < b)

Faster for sorting a large list is a "Schwartzian Transform" approach:

    auxlist = [ (x[2], x) for x in thelist ]
    auxlist.sort()
    thelist = [ x[1] for x in auxlist ]

This does roughly the same thing as:

    thelist.sort(lambda a, b: cmp(a[2], b[2]))

but more rapidly.  ("Roughly", because: the Schwartzian Transform
approach is not in-place, which may matter if you're very tight in
memory or there are other outstanding references to thelist; also,
if two items have identical x[2] sub-items, the [random-ish] ways
in which the two approaches sort them will probably differ).


Alex





From fredrik at effbot.org  Sat Dec 30 16:08:07 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sat, 30 Dec 2000 21:08:07 GMT
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <3A4D6A05.EEC97FA6@uswest.net> <slrn94qqvg.mgg.grey@teleute.rpglink.com> <ije36.96813$x6.43604873@news2.rdc2.tx.home.com> <slrn94qt4f.mgg.grey@teleute.rpglink.com> <CMo36.97857$x6.44474719@news2.rdc2.tx.home.com> <slrn94s9qg.134.grey@teleute.rpglink.com> <gdq36.2696$Qb7.302774@newsb.telia.net> <slrn94shbe.150.grey@teleute.rpglink.com>
Message-ID: <X8s36.2731$Qb7.305339@newsb.telia.net>

Steve Lamb wrote:
> On Sat, 30 Dec 2000 18:56:12 GMT, Fredrik Lundh <fredrik at effbot.org> wrote:
> >from what world do you come from where initializing variables to arbitrary
> >values just to get your program flow right is a good thing?
>
>     I have never heard of a single programming book or course that says, "Oh,
> don't init variables."  In this case I feel you are squarely in the wrong
> with, oh, the entire CS community.

Who said "don't init variables?"

Of course you should initialize a variable when it has a meaning-
ful initial value.

But in this case, the initial value is just something you made up
to get into the loop.  The temperature is never zero degrees (nor
minus 4711, or whatever) -- the initial value is just a flag value
that you use to get into the loop.

>     No, but I don't think that when you can place the test into the while that
> one should use a while 1:break construct.

What's the difference between using a special flag value, and a
separate flag variable?  I'd say it's the same thing...

(besides, no matter what you think, "do-while" is spelled "while
1-if-break" in Python.  no amount of arguing on comp.lang.python
will change that)

</F>




From sh at ttsoftware.co.uk  Wed Dec  6 06:25:57 2000
From: sh at ttsoftware.co.uk (Steve Horne)
Date: Wed, 06 Dec 2000 11:25:57 +0000
Subject: Sort
References: <NlnX5.1444$yv1.2700363@nnrp5.proxad.net>
Message-ID: <oa8s2tsls5foll62r8fpa6nlcpisb6v1ek@4ax.com>

On Wed, 06 Dec 2000 09:07:57 GMT, "Alfred" <scj at mcom.mcom.fr> wrote:

>I'm trying to sort a List of List, on the third col, but I don't find how to
>do this.

You can specify an alternative comparison function to the sort method.
The easiest way to do this is with a lambda...

l.sort (lambda a, b : a[2] < b[2])

For comparison, the following should do the same as the basic sort
(but slower)...

l.sort (lambda a, b : a < b)

-- 
Steve Horne
Home : steve at lurking.demon.co.uk
Work : sh at ttsoftware.co.uk


From jvanar1 at my-deja.com  Fri Dec  8 16:41:29 2000
From: jvanar1 at my-deja.com (jvanar1 at my-deja.com)
Date: Fri, 08 Dec 2000 21:41:29 GMT
Subject: Link problems embedding python in Linux. HELP!
References: <90r5kf$q9j$1@nnrp1.deja.com> <p6qr93iiqf4.fsf@informatik.hu-berlin.de>
Message-ID: <90rki8$796$1@nnrp1.deja.com>

Thanks Martin but that didn't help any.  What values do PYTHONHOME &&
PYTHONPATH be set to??
-J


In article <p6qr93iiqf4.fsf at informatik.hu-berlin.de>,
  Martin von Loewis <loewis at informatik.hu-berlin.de> wrote:
> jvanar1 at my-deja.com writes:
>
> > I've read every resource available, but nothing tells me what to
link
> > with. here is my compilation line:
> >
> > gcc -o go -lpython1.5 -L/usr/lib/python1.5 go.c
>
> The order of libraries on the linker line is relevant; place
> -lpython1.5 after go.c.
>
> Hope this helps,
> Martin
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From tino at wildenhain.de  Sun Dec 24 05:16:01 2000
From: tino at wildenhain.de (Tino Wildenhain)
Date: Sun, 24 Dec 2000 11:16:01 +0100
Subject: Some Python 2.1 ideas
References: <mailman.977590932.19335.python-list@python.org> <kuelyyi8k3.fsf@lasipalatsi.fi>
Message-ID: <3A45CCE1.B5E8F1DC@wildenhain.de>

Hi Erno,

Erno Kuusela wrote:
> 
> >>>>> "Bob" == Bob Alexander <bobalex at home.com> writes:
> 
> [snip]
> 
> | ---- String method to remove trailing newline
> 
> i think it would be prettier and more useful to have methods
> to remove a given prefix/suffix from the beginning or the end of a
> string. maybe add optional arguments for lstrip/rstrip, or
> introduce new methods rtrim/ltrim?

You mean removing it only if it matches?
If yes, you can at least use for ltrim:

string.replace(str,begin_to_match,'',1)

if not, you can simply use <string>[len(begin_to_match):]

Regards
Tino



From grey at despair.rpglink.com  Tue Dec 12 18:44:03 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Tue, 12 Dec 2000 23:44:03 -0000
Subject: Recommendations for a 2.0 book?
Message-ID: <slrn93de23.eum.grey@teleute.rpglink.com>

    OK, just a little while ago I had dabbled in Python 1.5.2 and was using
the Beazley book as my reference.  I've since lost the book and since I'm not
far enough along to remember most of the language (that is months of actual
use away) I am now pondering a replacement.  What is a good replacement now
that we're on Python 2.0?  Could I get another copy of the Beazley book and be
able to switch between 1.5.2 and 2.0 fairly easily?

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From dalke at acm.org  Sun Dec 10 14:40:31 2000
From: dalke at acm.org (Andrew Dalke)
Date: Sun, 10 Dec 2000 12:40:31 -0700
Subject: Programming Habits in Python
References: <cspU5.82459$U46.2617014@news1.sttls1.wa.home.com> <8vvb6p$ct2$1@news.wrc.xerox.com> <XiMU5.87123$U46.2803855@news1.sttls1.wa.home.com> <900cpl02ss1@news2.newsguy.com> <Ue3V5.91231$U46.2956112@news1.sttls1.wa.home.com> <slrn92tv0u.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com>
Message-ID: <910m5d$t6v$1@slb3.atl.mindspring.net>

A.M. Kuchling wrote:
>I've started reading Knuth's TAoCP several times but never
>gotten very far, mostly because vol. 1 quickly embroils you
>in MIX, making it very hard to distill anything of practical
>value from the book.

I'll second that.  I tried for figure out one of the MIX
examples and found it to be quite difficult.  The examples
do not use structured programming, and the code I tried
could not easily be turned into a set of nested loops.
Instead, the loops were interweaved using gotos.

The problem is that for the last 10 years, my thoughts have
been tuned as it were to hierarchical loops and not arbitrary
tangles, so it was really hard to figure out the example.

I compare that to Kernigan and Plauger's "The Elements of
Programming Style" where even their examples of well written
code uses gotos for the loop constructs.  It's easier to
understand in that case because I could mentally change them
to while loops, but there's still the mental overhead of
having to verify that the jump destinations are correct.

As yet another comparison, I've been doing a lot of work with
mxTextTools, which uses a state table to do text processing.
It's very fast so I rewrote some of my Python parsing code
using it.  While there were only about 12 states, it took
a few hours to write and debug while the Python implementation
took about 15 minutes.

I ended up coding the example in C++ instead of Python because
C++ has gotos and because C++'s STL includes a priority queue.
The whole code ended up being about 5 times longer than the
MIX example, which isn't bad since MIX puts a lot of the
code on one line, and he didn't declare a priority queue.

It would be very hard to rewrite the code for Python because
there are no gotos, so almost by definition that means it's
difficult to "distill anything of practical value from the
book." :)  Theoretical value, yes.  Practical value, no.

                    Andrew
                    dalke at acm.org





From tim.one at home.com  Fri Dec 29 15:16:44 2000
From: tim.one at home.com (Tim Peters)
Date: Fri, 29 Dec 2000 15:16:44 -0500
Subject: Good introduction to functional programming with Python?
In-Reply-To: <86y9x0qxsq.fsf@localhost.my.domain>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEOCIFAA.tim.one@home.com>

[Brad Knotwell]
> The information in that [Erlang] book is a bit dated (even so it's
> still a good book).  More recent information is at the following URL:
>
>  http://www.erlang.org/doc/current/doc/extensions/funs.html
>
> IIRC, this book also doesn't talk about Erlang's list comprehensions.

Right, the passage I quoted said that, but disguised under the highfalutin
name "ZF comprehension" (from the so-called but not always-called "axiom of
comprehension" in Zermelo-Fraenkel (ZF) set theory, which is where the idea
really came from -- although it would be more accurate to call them "list
separations", because the ZF axiom of comprehension should be (and sometimes
is) called the axiom of separation instead <wink>).

Thanks for the link!  Anyone else playing along will find the frame-based
version helpful (it's got a table of contents in a frame for quick
navigation):

http://www.erlang.org/doc/current/doc/extensions/part_frame.html




From junaftnoon at nospamplzyahoo.com  Fri Dec  8 06:28:38 2000
From: junaftnoon at nospamplzyahoo.com (June Kim)
Date: Fri, 8 Dec 2000 20:28:38 +0900
Subject: Parallel Computing
References: <3A300B1D.223630F3@bioeng.ucsd.edu> <874s0fpzr2.fsf@lisboa.ifm.uni-kiel.de>
Message-ID: <90qgmt$1ko$1@news.nuri.net>

"Janko Hauser" <jhauser at ifm.uni-kiel.de> wrote in message
news:874s0fpzr2.fsf at lisboa.ifm.uni-kiel.de...
> There are modules which wrap parts of MPI and PVM, both are available
> from the starship or look at the vaults for a link to them. If you
> want to have parallel communication between Python programs there are
> many different ways, like Dopy, Pyro, CORBA, SOAP, XML-RPC...
>
> HTH,
> __Janko
>
> --
>   Institut fuer Meereskunde             phone: 49-431-597 3989
>   Dept. Theoretical Oceanography        fax  : 49-431-565876
>   Duesternbrooker Weg 20                email: jhauser at ifm.uni-kiel.de
>   24105 Kiel, Germany

Is "Python Parallelization Project" still alive? The link to the MPI
interface seems broken. Is there anyother place where I can get it?

http://starship.python.net/crew/hinsen/parallel.html




From root at rainerdeyke.com  Sat Dec 23 14:15:58 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Sat, 23 Dec 2000 19:15:58 GMT
Subject: <> vs != was Re: A Suggestion for Python Colon Syntax
References: <2kM06.120$Lm2.7518@news.clarkson.edu> <LNBBLJKPBEHFEDALKOLCKEDEIFAA.tim.one@home.com> <mailman.977594956.22001.python-list@python.org>
Message-ID: <OR616.67063$x6.31192247@news2.rdc2.tx.home.com>

"Thomas Wouters" <thomas at xs4all.net> wrote in message
news:mailman.977594956.22001.python-list at python.org...
> On Sat, Dec 23, 2000 at 12:15:25AM -0500, Tim Peters wrote:
>
> > Best guess is that "<>" will go away in P3K.  I jokingly told Guido it
was
> > clearly wrong for an inequality operator to look so symmetric, and he
forgot
> > to laugh <wink>.
>
> Well, excuse me, but it's clearly wrong for an equality test operator to
> look so much like an augmented assignment statement ! I definately prefer
> '<>' myself, anyway, I just wish I could use it in C without turning
towards
> preprocessor macros :)

I dislike <> because to me A <> B implies A > B or A < B.  This does not
logically apply to some data types where no ordering is defined.  Of course
in Python, A <> B (and A != B) does imply A > B or A < B because of the way
comparison is implemented.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From aleaxit at yahoo.com  Tue Dec 26 13:37:39 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 26 Dec 2000 19:37:39 +0100
Subject: Can't install any packages
References: <92acdv$eid$1@nnrp1.deja.com>
Message-ID: <92ap0f02b6r@news1.newsguy.com>

"EdK" <e_d_k at my-deja.com> wrote in message
news:92acdv$eid$1 at nnrp1.deja.com...
> I'm runing Python 1.6 on Windows98.  When I start Python it identifies
    [snip]
> None of the zip files I have downloaded contained any .dll files, but
> they do contain .ptd files. I can not find any documentation for whay a
> .pyd file is. These .pyd file all exist on my PYTHONPATH.

A .PYD file is actually a DLL, renamed to avoid confusing it with
"normal" DLL's.  I suspect the packages you download support
Python 2.0, which came out so soon after 1.6; upgrading to 2.0
might solve it (unless some very old packages only support the
1.5 version).  1.6 ruled the roost for such a short time (its
final was out the same day as 2.0's beta!) that few or no pre-built
binaries are likely to support it.


Alex





From jicman at cinops.xerox.com  Thu Dec 21 11:09:46 2000
From: jicman at cinops.xerox.com (Jose Isaias Cabrera)
Date: Thu, 21 Dec 2000 11:09:46 -0500
Subject: How to get the os.system data into a variable
References: <91o5ih$2kv$1@news.wrc.xerox.com> <20001219172933.B29727@ActiveState.com>
Message-ID: <00a401c06b68$7257c760$5aa4af0d@dssc.mc.xerox.com>


> On Tue, Dec 19, 2000 at 05:23:29PM +0000, jicman wrote:
> > I know that
> > 
> > t = os.system('ls -l')
> 
> o = os.popen('ls -l')
> output = o.read()
> o.close()
> 
> "output" now contains the output. Also look at os.popen2, os.popen3, etc for
> when you want handles on stderr and stdin inaddtion to the usual stdout.

thanks.




From mikael at isy.liu.se  Tue Dec 12 11:17:46 2000
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Tue, 12 Dec 2000 17:17:46 +0100 (MET)
Subject: Inheriting the @ sign from Ruby
In-Reply-To: <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu>
Message-ID: <XFMail.001212171746.mikael@isy.liu.se>

On 12-Dec-00 Roy Katz wrote:
 >  shorthand??

If you want a shorthand, why not just do

    Class X:
        def __init__( I, c ): 
            I.c = c

It is substantially shorter than self. and myself., and only one 
character longer than your @-expression, with the advantage that 
you can use it at once without any changes to Python itself.

And-I-think-that- at var-is-ugly-ly y'rs

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    12-Dec-00
Time:    17:11:13

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

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


From loewis at informatik.hu-berlin.de  Wed Dec 27 20:17:52 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 28 Dec 2000 02:17:52 +0100
Subject: Python 2.0 parser module bug(?) and a Q
References: <Pine.LNX.4.10.10012272204060.330-100000@kev.ioc.ee>
Message-ID: <p6qofxxe43j.fsf@informatik.hu-berlin.de>

Pearu Peterson <pearu at ioc.ee> writes:

> I have found that parser module may have a bug in its functions tuple2ast
> or suite.

Please report bugs on sourceforge.net/projects/python; on this list,
they might easily be overlooked.

> And my question is (not really related to above):
> 
>   Is there inverse function for parser.suite? 

A colleague of mine has an "unparser" module for Python AST; that is
not (yet) released to the public, and works not (yet) for 2.0. Please
contact me if you need more details.

Regards,
Martin


From aleaxit at yahoo.com  Thu Dec 14 12:25:48 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 14 Dec 2000 18:25:48 +0100
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <tcmay-FCA22F.14073713122000@news.got.net> <918v7a02scn@news1.newsguy.com> <tcmay-04E80C.16181313122000@news.got.net> <91a1ed02a4j@news2.newsguy.com> <3A389B1A.4F33FFD4@cfmu.eurocontrol.be> <91ae8s02llf@news2.newsguy.com> <3A38CF3E.8BBD2216@cfmu.eurocontrol.be>
Message-ID: <91avrn08ql@news2.newsguy.com>

"Ian Wild" <ian at cfmu.eurocontrol.be> wrote in message
news:3A38CF3E.8BBD2216 at cfmu.eurocontrol.be...
> Alex Martelli wrote:
> >
> > Forget C++ -- we're talking OO _analysis_.  Assume you're
> > trying to perform OOA on the universe of discourse "immutable
> > planar 2D geometric polygons", for the sake of definiteness.
> >
> > If Rectangle and Square are both concrete (instantiable)
> > classes, with *overlapping* requirements for membership,
> > how do you classify a given rectangular object which
> > happens to have equal sides?  Toss a coin?
>
> Both.
>
> Unless you're using classes purely as an implementation
> shortcut, why need an object be a member of precisely *one*
> class?

Because that's how object-oriented analysis works -- at
least, all OOA methodologies I've ever encountered.  Please
supply references to OOA methods working differently --
they'll surely be interesting to study.  (If you don't
have any such reference, then go away for a while, write
up your new and brilliant 'one object -> many concrete
classes at a time' OOA methodology, have Addison-Wesley
publish it, make a mint as it sweeps away all the old and
dusty 'one object -> one class' alternatives and you
become the most sought-after consultant in the industry,
THEN provide the reference, thanks).


> What if your analysis is later implemented in
> a language without classes (eg Self?), or one where
> classes exist only by convention (Perl)?

'only by convention'!?  Have they abrogated the 'bless'
keyword, now?

Anyway, when I do OOA (and sometimes OOD), and then
have to implement the results in a non-OO language,
or a language whose object-model is miles apart from
that implied in the OOA/OOD methodologies I have
chosen to employ, then I may have more bridging to do
than if the object models were a good match for each
other (each at the appropriate level of abstraction).
So what else is new?

I fail to see any benefit in a methodology assigning
each object to several concrete-classes, as opposed
to one identifying _interfaces_ (abstract classes),
and working along classical lines -- each object is
in a (one!) class, each class implements N interfaces,
etc.  This is a good match for the object models in
COM, Java (though you may have to further identify a
pattern of implementation inheritance then -- but this
is clearly a later issue, for design-time), .NET/C#
(ditto), Sather -- and easily implementable in languages
with richer class-structures, such as Eiffel and C++
(and in languages without a formalized concept of
'interface', such as Smalltalk and Python, where
one just defines the interface as a bunch of related
messages an object 'implementing the interface' will
appropriately respond to -- typical dynamic vs static
typing dichotomy).  I don't have enough practical
experience in OO design/coding in Self or Perl to be
sure, but I'd bet it wouldn't be beyond what's easy
in those languages, either.

What compensating advantage does your favourite
'many-CONCRETE-classes-per-object' OOA approach
offer, to offset its serious mismatch with many
available implementation-languages object models?
I can't see any, but am I just being short-sighted...?


Alex





From olczyk at interaccess.com  Sat Dec 23 09:59:06 2000
From: olczyk at interaccess.com (Thaddeus L. Olczyk)
Date: Sat, 23 Dec 2000 14:59:06 GMT
Subject: Problem with python mode
Message-ID: <3a44bcf1.73260781@nntp.interaccess.com>

When i try loading pdb.el in NTemacs I keep getting an error--
Symbol's function defiition void: toolbar-make-button-list.
Any idea what the problem is?
TIA


From dsavitsk at my-deja.com  Tue Dec 12 01:44:15 2000
From: dsavitsk at my-deja.com (dsavitsk at my-deja.com)
Date: Tue, 12 Dec 2000 06:44:15 GMT
Subject: Plan9 port of python
Message-ID: <914hfv$pi1$1@nnrp1.deja.com>

does a port of python exist for plan9 or inferno? and or will it ever?

thanks
doug


Sent via Deja.com http://www.deja.com/
Before you buy.


From brianknox at home.com  Sun Dec 24 17:11:50 2000
From: brianknox at home.com (Brian Knox)
Date: Sun, 24 Dec 2000 22:11:50 GMT
Subject: Python Database Interfaces - Any Standard?
References: <RUo16.204045$RP.37382354@news1.rdc1.md.home.com> <kuk88pu0cd.fsf@lasipalatsi.fi>
Message-ID: <Gwu16.205577$RP.37521696@news1.rdc1.md.home.com>

The Perl DBI consists of two parts: the non database specific DBI module
(DBI.pm). Then there is a DBD (DataBase Driver) module for each database.
So, if I wanted to say work with Oracle, I'd use DBI.pm and the Oracle DBD.

use DBI;
$dbh = DBI->connect( $dbi:Oracle:archaeo", "username", "password" );

To switch to say PostgreSQL, just change "Oracle" to "Pg" and make sure you
have the DBD.

Brian

"Erno Kuusela" <erno-news at erno.iki.fi> wrote in message
news:kuk88pu0cd.fsf at lasipalatsi.fi...
> >>>>> "Brian" == Brian Knox <brianknox at home.com> writes:
>
> it's ofcourse not possible to have a single module cover all databases
> since the various databases have different protocols for talking to
> them (modulo odbc/jdbc). well, unless the single module contained code
> for every conceivable database, but that would hardly be sensible.
>
>   -- erno





From jah at spam-no-no.mail.ru  Thu Dec 14 17:25:16 2000
From: jah at spam-no-no.mail.ru (Igor Poteryaev)
Date: Fri, 15 Dec 2000 03:25:16 +0500
Subject: Python for REAL Projects ?
References: <3a380d1e.409238@bendigo.news.origin.net.au>
Message-ID: <3A3948CC.50939621@spam-no-no.mail.ru>

Hello DELPHI dragon !

You definitely can get the best from both Delphi and Python worlds.
Have you ever heard about PythonForDelphi located at
http://www.multimania.com/marat/  ?
It is a nice example of both extending and embedding capabilities of 
python. You get all comfort of using Delphi forms for UI and flexibility 
of python. 


dragon at hitech.net.au wrote:
> 
> Hello Python People !
> I have been looking at, and trying out Python (under Win NT) and
> although Python SEEMS interesting can it "cut the mustad" on a real
> client project ??
> 
> Curently, I develop using DELPHI + Interbase SQL ( or Borland BDE
> tables) because , in the main, DELPHI offers "mostly" easy to setup
> forms and access to database tables ( not perfect but usable).
> 
> FORMS are a major feature of DELPHI and very complex forms may be
> built very rapidly  with AUTOMATIC code stub attachments for component
> events, ie place the component on the form, 'enter' the object
> inspector for that component and write code for the desired  event.

Hmm. Usually it leads to unnecessary dependencies between UI and 
program logic. But it is the theme for discussion in other newsgroup.


Regards

    Igor Poteryaev


From aleaxit at yahoo.com  Fri Dec 22 03:53:41 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 22 Dec 2000 09:53:41 +0100
Subject: How use urllib.urlopen()?
References: <9012313F4gustaflalgonetse@194.213.69.148>
Message-ID: <91v4r601v10@news2.newsguy.com>

"Gustaf Liljegren" <gustafl at algonet.se> wrote in message
news:9012313F4gustaflalgonetse at 194.213.69.148...
> Can't get the urlopen() function to work. I have a script on a UNIX web
> server that should go to another webpage when the user clicks a button. To
> isolate the problem, I made a test script:
>
> #!/usr/local/bin/python
>
> import urllib
> u = urllib.urlopen("http://www.yahoo.com")
> print u.read()
>
> When running this I get HTTP 500 - Internal server error. :-(
>
> What's missing?

This script works fine for me if I run it on a machine outside
our firewall.  Are there perhaps proxy/firewall issues with
your "UNIX web server"'s configuration/settings...?  Seems one
likely explanation.


Alex





From josh at open.com  Wed Dec 20 15:46:15 2000
From: josh at open.com (Joshua Muskovitz)
Date: Wed, 20 Dec 2000 15:46:15 -0500
Subject: while (thread.alive()):
References: <mailman.977283904.26696.python-list@python.org>
Message-ID: <3a4118f8_2@corp.newsfeeds.com>

if while 1: bothers you so much, then why not use while "forever": ?  (Only
half a smiley here.)

-- j




-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----


From qdlaty at wielun.dhs.org  Mon Dec 25 13:12:46 2000
From: qdlaty at wielun.dhs.org (QdlatY)
Date: Mon, 25 Dec 2000 19:12:46 +0100
Subject: html filtering
References: <3A46900D.BD118A0A@wielun.dhs.org> <mailman.977727367.19724.python-list@python.org>
Message-ID: <3A478E1E.97BD8B67@wielun.dhs.org>

Hello!

Moshe Zadka wrote:
> 
> On Mon, 25 Dec 2000, QdlatY <qdlaty at wielun.dhs.org> wrote:
> 
> > Big problem in writeing message boards is filtering html tags. I use
> > very simple method: Each occurence of "<" sign or ">" sign I replace
> > with &lt; and &gt;. Is this safe method? Maybe You've got better ideas?
> 
> Well, it is safe as long as you remember to replace every "&" with
> "&amp;" *before* you replace < and >.

Yes! Thanks. I forget about ampersands;)).


Yours
QdlatY


From amk at mira.erols.com  Sun Dec 10 11:24:25 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 10 Dec 2000 16:24:25 GMT
Subject: Programming Habits in Python
References: <cspU5.82459$U46.2617014@news1.sttls1.wa.home.com> <8vvb6p$ct2$1@news.wrc.xerox.com> <XiMU5.87123$U46.2803855@news1.sttls1.wa.home.com> <900cpl02ss1@news2.newsguy.com> <Ue3V5.91231$U46.2956112@news1.sttls1.wa.home.com> <slrn92tv0u.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com> <udn1e8k0qi.fsf@box.home.de>
Message-ID: <slrn937caf.h4.amk@207-172-146-162.s162.tnt3.ann.va.dialup.rcn.com>

On 07 Dec 2000 11:02:45 +0100, 
	Jochen K?pper <jochen at pc1.uni-duesseldorf.de> wrote:
>Not to destroy the metaphor, but you cannot learn modern physics from
>Principia! Classical Mechanics at most. But you can learn the
>physicists way of thought... [1]

In the back of my mind I had a quotation, which I cannot track down on
the Web.  The Principia uses geometrical methods throughout, even
where analytical methods would have been much simpler, and someone
(Halley, perhaps) said that reading the Principia was like walking
through an arsenal of giant weapons of a past age, and that one is
amazed that anyone could wield them at all.  I feel the same way about
TAoCP; amazing piece of work, but I'm not strong enough to make any
use of it.  Hmm, there's an essay lurking here somewhere...

--amk






From dale at out-think.NOSPAMco.uk  Sun Dec 31 12:52:55 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Sun, 31 Dec 2000 17:52:55 +0000
Subject: Printing with wxPython
Message-ID: <ecsu4t89nigkvgna3kr8a53c84pgf4l3mu@4ax.com>

I'm trying to get to grips with wxPython and printing on Windows.

Has anyone got a very simple sample that sets up a device context and prints some text or simple
graphics?

I can't find any printing examples.

Thanks for any help.


--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants


From deha at poczta.onet.pl  Tue Dec 12 06:32:09 2000
From: deha at poczta.onet.pl (DeHa)
Date: Tue, 12 Dec 2000 12:32:09 +0100
Subject: A question.
Message-ID: <9152a0$acb$1@panorama.wcss.wroc.pl>

Hi All!

    I am new to Python, and just try to write my first program. But I have a
small problem.

    My program should find first numbers in the set from 1 to 100000. I made
such a code:

first = range(100001)
first = first[1:]

for i in first:
    if i != []:
#until end of the list
        for j in first[first.index(i):]:                           #find all
numbers divided by i
            if j % i == 0:
                first[first.index(j): first.index(j)] = []    #and delete it
from the list

    And it doesn't work, and I don't konw why? Could anyone help me? I try
to find answer in the tutorial, but the information in it doesn't change
anything.

Dawid Hulisz





From MarkH at ActiveState.com  Sat Dec 30 00:10:07 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sat, 30 Dec 2000 05:10:07 GMT
Subject: Komodo
References: <B672A66F.72CC%bobhicks@adelphia.net>
Message-ID: <3A4D6D0B.20000@ActiveState.com>

Robert L Hicks wrote:

> Has anyone tried Komodo from ActiveState yet? 

Yes. *

> What are you thoughts on it?

I think it is pretty good ;-) *

> Will Komodo for Linux have to have the ActiveState versions of Perl and
> Python to work?

I can't speak for Perl.  But for Python, you certainly should be able to 
craft your own working Python version.  There is nothing special in the 
ActivePython binaries, other than everything Komodo needs in one 
package.  I made a valiant effort to continue running Komodo on Linux 
with Python built from sources, but gave up once I discovered I didnt 
have the XML package configured correctly.  Obviously people with a clue 
on Linux would fare better than I ;-)

Or, IOW, ActivePython is the only configuration Komodo is known to run 
under, but there are no obvious reasons why other configurations are not 
suitable.

Mark.

* For those who can't infer the obvious, I'm part of the Komodo dev team :-]



From fredrik at effbot.org  Tue Dec 12 17:16:50 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 12 Dec 2000 22:16:50 GMT
Subject: quoting styles (was Re: Label-Value)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <3A368E11.20C17E92@san.rr.com> <r0pwvd51gga.fsf@black.cs.washington.edu> <kupuix71b6.fsf@lasipalatsi.fi> <9167qo$g4s$1@panix2.panix.com>
Message-ID: <mtxZ5.3611$Kd1.423896@newsb.telia.net>

When you see the answer before the question.

Aahz wrote:
> What's "jeopardy style" quoting?




From wware at world.std.com  Fri Dec  8 09:01:44 2000
From: wware at world.std.com (Will Ware)
Date: Fri, 8 Dec 2000 14:01:44 GMT
Subject: My first attempt at python
References: <n2j13tcc0il8k1i3dtjkdckglo871ln744@4ax.com>
Message-ID: <G595Mx.KJ8@world.std.com>

Jason Stewart (kahuna01 at hotmail.com) wrote:
> 	This is the first piece of code that I wrote in Python...
> 	This same functionality might be in a module somewhere, but I
> wrote it for practice. Its a string tokenizer.

In the "string" module, there's a function called "string.split"
that does this same thing. The Python libraries are vast deep
repositories of boundless ingenuity:
http://www.python.org/doc/current/lib/lib.html

> def tokenize(string_to_chop):
> 	"Break a string up into individual words"
> 	last_space = 0
> 	token = []
> 	for index in range(len(string_to_chop)):
> 		if string_to_chop[index:index+1] == " ":
> 			token.append(string_to_chop[last_space:index])
> 			last_space =  index + 1
> 		elif (index+1 == len(string_to_chop)):
> 		    token.append(string_to_chop[last_space:index+1])
> 	return token

Instead of writing "string_to_chip[index:index+1]", you can just write
"string_to_chop[index]". Two lines later you write "last_space = index + 1"
which suggests that you expect the tokens on the line to always be
separated by exactly one space; otherwise you'll end up with confusing
results. Try tokenize("abc     def     ghi") and see what it does.
-- 
# - - - - - - - - - - - - - - - - - - - - - - - -
# Resistance is futile. Capacitance is efficacious.
# Will Ware	email:    wware @ world.std.com


From avv at quasar.ipa.nw.ru  Mon Dec  4 20:04:57 2000
From: avv at quasar.ipa.nw.ru (Alexander V. Voinov)
Date: Mon, 04 Dec 2000 17:04:57 -0800
Subject: mp3 module
References: <3A2BB89B.D4C30300@ira.uka.de>
Message-ID: <3A2C3F39.BEB46122@quasar.ipa.nw.ru>

Hi

Markus von Ehr wrote:
> is there a module to play mp3 files with python?

write(os.popen("mpg123 -", "w"), my.brand.big.blob)?

Alexander


From roy at panix.com  Sat Dec  9 20:53:50 2000
From: roy at panix.com (Roy Smith)
Date: Sat, 09 Dec 2000 20:53:50 -0500
Subject: Handling signals and performance issues.
References: <meRU5.211$Mx.12164@sjc-read.news.verio.net> <m9dgmAAe+tM6EwnY@gol.com>
Message-ID: <roy-FD4A85.20535009122000@news.panix.com>

> My second problem is that while the program handles a CTRL-C, etc 
> well-- cleaning up and the like, if it is sent a kill -9 or kill -HUP 
> by root or the user spawning the process it dies a horrible 
> death--ie, it doesn't clean up and flush its buffers, or run the 
> 'cleanup' functions I've defined with a try: except: block.

You cannot catch a kill -9 signal.  It is the "signal of last resort", a 
way to kill a process regardless of what it has done to itself in the 
way of signal processing/catching/ignoring.  This is as true in python 
as in any other language.

There always has to be some sort of un-catchable signal, otherwise a 
process would be able to make itself completely immune to being killed.  
There are still ways to make processes unkillable, but they all involve 
tickling kernel bugs :-)

As for watching the end of a log file without burning CPU, as somebody 
else already mentioned, sleeping for a second between reads would be one 
good way.  Another thing you might try is using a select() call to tell 
you when new data is readable on the file.

You might want to look at the source to the unix "tail" command (you 
should be able to find it on any of the many GNU/Linux distribution 
sites) and see how it implements the "-f" option, which is essentially 
what you are trying to emulate.


From etsang at my-deja.com  Tue Dec  5 15:16:38 2000
From: etsang at my-deja.com (etsang at my-deja.com)
Date: Tue, 05 Dec 2000 20:16:38 GMT
Subject: resource module??
References: <8v1p0d$7cl$1@nnrp1.deja.com> <p6qy9yh780r.fsf@informatik.hu-berlin.de>
Message-ID: <90jiet$ltn$1@nnrp1.deja.com>

Hi, I uncomment out the line #resource resource.c in Modules/Setup, amd
then rebuild theo whole thing again, but it is still not importing.
Before rebuilding I have use make clean.

Any advice?

In article <p6qy9yh780r.fsf at informatik.hu-berlin.de>,
  Martin von Loewis <loewis at informatik.hu-berlin.de> wrote:
> etsang at my-deja.com writes:
>
> > Hi has anyone who has used the resource module??
> > I am using Python 1.5.2 but when I try to import resource module,
it
> > is not there..
>
> It's a compile-time option.
>
> > Does anyone know anything about this module? does it come with
Python
> > 1.5.2??
>
> Sure, it comes with 1.5.2. Modify Modules/Setup to activate it.
>
> Regards,
> Martin
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From auto103690 at hushmail.com  Mon Dec 18 01:43:37 2000
From: auto103690 at hushmail.com (Jeff)
Date: Mon, 18 Dec 2000 06:43:37 GMT
Subject: PyOpenGL and Nvidia GL?
Message-ID: <tmi%5.60958$df5.2143169@news1.crdva1.bc.home.com>

A non-text attachment was scrubbed...
Name: not available
Type: text/plain  charset=iso-8859-1
Size: 444 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20001218/4fd0bc1d/attachment.bin>

From fredrik at effbot.org  Sun Dec 31 04:48:48 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sun, 31 Dec 2000 09:48:48 GMT
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <3A4D6A05.EEC97FA6@uswest.net> <slrn94qqvg.mgg.grey@teleute.rpglink.com> <ije36.96813$x6.43604873@news2.rdc2.tx.home.com> <slrn94qt4f.mgg.grey@teleute.rpglink.com> <CMo36.97857$x6.44474719@news2.rdc2.tx.home.com> <slrn94s9qg.134.grey@teleute.rpglink.com> <gdq36.2696$Qb7.302774@newsb.telia.net> <slrn94shbe.150.grey@teleute.rpglink.com> <X8s36.2731$Qb7.305339@newsb.telia.net> <3A4EB9B5.B4B60F4C@gte.net>
Message-ID: <4iD36.2776$Qb7.316530@newsb.telia.net>

Steve Williams wrote:
> Do you have trouble porting your code to other languages?

Nope.

But given the other Steve's reaction, I might have trouble
porting my thoughts to other languages.

Or maybe he's just stupid.

</F>




From fredrik at effbot.org  Thu Dec  7 11:33:46 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Thu, 07 Dec 2000 16:33:46 GMT
Subject: PIL binary for Python 2.0 for Windows ?
References: <3a2f472f$0$235@hades.is.co.za>
Message-ID: <KZOX5.2015$Kd1.294314@newsb.telia.net>

Leon Booyens wrote:
> Where can I find a binary distribution for Windows of the Python Image
> Library 1.1.1. (It is not as yet available from Pythonware as a binary and I
> unfortunatly have no means of compiling the code on a windows platform)

http://effbot.org/pil

</F>




From slhath at home.com  Sun Dec 24 02:09:01 2000
From: slhath at home.com (Scott Hathaway)
Date: Sun, 24 Dec 2000 07:09:01 GMT
Subject: xml-rpc server in python returning python lists...
Message-ID: <hih16.85504$yR4.2492277@news1.rdc1.tx.home.com>

Hello All,

I am writing an xml-rpc server in python that has a function which returns a
python list.  When I use the python client to connect to the python server,
it works great.  When I try from VB, ASP, PHP, and Delphi, they all have
trouble with the return type of the variable.  By the way, all the clients
mentioned work fine if the server's method returns a string or a number.

Any ideas?

Scott




From dalke at acm.org  Sun Dec 17 02:53:41 2000
From: dalke at acm.org (Andrew Dalke)
Date: Sun, 17 Dec 2000 00:53:41 -0700
Subject: Singleton vs Proxies DP (was Re: Solution: Direct access to Printer I/O lines)
References: <917k9b$9i9$1@nnrp1.deja.com> <slrn93hrjn.2vp.ssthapa@localhost.localdomain> <91cnp3$g59$1@nnrp1.deja.com> <91d23f0jk7@news1.newsguy.com> <91gfnp$7r4$1@animus.fel.iae.nl> <91gkh801e49@news2.newsguy.com> <InW_5.37707$x6.18119119@news2.rdc2.tx.home.com>
Message-ID: <91hrbn$v7a$1@slb3.atl.mindspring.net>

Rainer Deyke wrote:
>My preference for standardized singletons in Python is this:
>
>singleton C:
>  pass # class definition here
>
>The singleton statement would be equivalent to the following:
>
>class __tmp:
>  pass # class definition here
>C = __tmp()
>del __tmp

Wouldn't the cleaner equivalent be

class C:
  pass
C = C()

 ?

In that way the class name used by the default repr would
still show a 'C' and you would have the nagging worry of
accidentally affecting someone else's use of __tmp in the
current namespace.  It also looks cleaner.

If that's the case, do we really need a new keyword for
what can be done with one extra line of code and no new
functionality?

                    Andrew Dalke
                    dalke at acm.org





From joconnor at cybermesa.com  Sat Dec 16 10:04:16 2000
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Sat, 16 Dec 2000 15:04:16 +0000
Subject: while true: !!!
 
Message-ID: <3A3B8470.C5193211@cybermesa.com>

>Yeah, you're right.. Things like that are definitely bad when someone
>else looks at your code, but for something that you're sure nobody will
>ever see, i might try using this.. except for doing true = 0, of course 


Nah...get creative.

You need to define true=1; false = 0..

and then you get code like 

true = false      #exits the loop
true = ! false 

etc...

-- 


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

"God himself plays the bass strings first when He tunes the soul"



From aleaxit at yahoo.com  Wed Dec 13 08:18:19 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 13 Dec 2000 14:18:19 +0100
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de>
Message-ID: <917sva02flj@news2.newsguy.com>

"Bjoern Giesler" <Bjoern.Giesler at gmx.de> wrote in message
news:917r2b$4n7$1 at news.rz.uni-karlsruhe.de...
> Hi,
>
> In comp.lang.python israel raj thomas <israelt at optushome.com.au> wrote:
> :>Do you think that it would be possible to use Python as a Smalltalk
> :>replacement if an IDE existed that hid some of Python's limitations?
>
> : Would it perhaps not be better to use Ruby ?
>
> Not for me, for the following reasons:
> a) I don't know Ruby well enough
> b) Ruby is "even more" unknown in The Real World(tm) than Python
> c) I think Ruby code looks ugly (what with @, #, $, no 'return' etc.)
> d) I don't see any of the features Ruby has over Python as 'killer'
>    features.

I might, personally, agree, but a really-rabid Smalltalker might
differ with [d] -- being able to inherit from *anything* is, I
think, the main extra of Ruby wrt Python, and isn't all of
Smalltalk about inheriting from integers, &c?-)


Alex





From qdlaty at wielun.dhs.org  Wed Dec 27 10:55:42 2000
From: qdlaty at wielun.dhs.org (QdlatY)
Date: Wed, 27 Dec 2000 16:55:42 +0100
Subject: textarea parsing
Message-ID: <3A4A10FE.B8C1B18C@wielun.dhs.org>

Hello!

I'm creating some multi-channels message board in python.
I have question: how textarea form item is passed to python script?
Edit line items are passed as strings. I though that textarea items will
be passed as normal string with "\n"/"\r" signs.
f.e.
form = cgi.FieldStorage()
#Now extracting contets of textarea formitem
print form["text"].value

Why I see MiniStorage object?
How to easily extract lines of textarea from this object?

Yours
QdlatY


From aleaxit at yahoo.com  Mon Dec 11 09:15:56 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 11 Dec 2000 15:15:56 +0100
Subject: 3rd argument to os.path.walk(): Why?
References: <912gdb$1oqr$1@news.idiom.com>
Message-ID: <912nj60ccr@news2.newsguy.com>

"James T. Dennis" <jadestar at idiom.com> wrote in message
news:912gdb$1oqr$1 at news.idiom.com...
>   In the _Python_Essential_Reference_, by David Beazley
>  (New Riders Publishing, 2000) I notice that os.path.walk()
>  takes three arguments which he lists as:
>
> path, visitfunc, arg

So it does.

>  It notes that the function referenced in visitfunc is called
>  with (arg, dirname, namelist) for every directory that is traversed
>  by the os.path.walk.

And, indeed, that's what happens.


>  I don't understand what the "args" is for.

But you do, as you say later:

>  ignoring the 'args' argument in my functions.  I guess I
>  could write my functions to behave differently based on
>  the contents of the args.  (I can even imagine that as
>  useful, though I don't need it for anything right now).

Exactly.  Could come in useful at some time (make the
function reentrant by avoiding having it refer to a
global variable, etc).

E.g., say you want to _count_ the files:

import os

class Counter:
    def __init__(self):
        self.count = 0
    def inc(self, n=1):
        self.count += n

def count_names(count, dir, flist):
    count.inc(len(flist))

count = Counter()
os.path.walk("/", count_names, count)

print count.count,"files in all"


You could do it without the argument by having the
function access count as a global variable, but if
global-variables can be easily avoided, so much
the better, don't you think?


> Why is this third argument required in os.path.walk()?

Why required as opposed to optional?  Dunno, it seems
that it might well have been made optional, defaulting,
say, to None.  But this seems to be a very minor issue --
it's not hard to pass in the third-argument as None in an
explicit way rather than letting it default, for example.


Alex





From etsang at my-deja.com  Thu Dec  7 19:49:36 2000
From: etsang at my-deja.com (etsang at my-deja.com)
Date: Fri, 08 Dec 2000 00:49:36 GMT
Subject: How to keep a windows frame on top??
References: <90p9ss$chi$1@nnrp1.deja.com>
Message-ID: <90pb6v$dfb$1@nnrp1.deja.com>

by the way I am usin Tkinter for gui.
thanks
In article <90p9ss$chi$1 at nnrp1.deja.com>,
  etsang at my-deja.com wrote:
> This is the function for calling my window. Can anyone tell how can I
> make the windows to always stay on top when called?
>
> def resultsData(self):
>
>         okButton = None
>
>         self.inputWindow = Toplevel(self.master)
>         self.inputWindow.geometry("+%d+%d" % (self.master.winfo_rootx
()
> +180,
>                                   self.master.winfo_rooty()+180))
>         self.inputWindow.title("Enter data")
>         Label(self.inputWindow, text ="Please enter your initials and
> the test object").grid(row=0, columnspan=2, pady=6)
>         Label(self.inputWindow, text="Initials").grid(row=1, sticky=W)
>         Label(self.inputWindow, text="Test object").grid(row=2,
> sticky=W)
>         self.e1 = Entry(self.inputWindow, textvariable=self.name,
> width=35)
>         self.e2 = Entry(self.inputWindow, textvariable=self.iut,
> width=35)
>         self.e1.grid(row=1, column=1, sticky="EW", pady=1)
>         self.e2.grid(row=2, column=1, sticky="EW", pady=1)
>         okButton = Button(self.inputWindow, text="Ok",
> command=self.getData)
>         okButton.grid(row=3, column=1, sticky=W, pady=6)
>         if self.automatic == FALSE:
> 	   self.inputWindow.wait_window()
> 	else:
> 	   self.getData()
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From ldw at us.ibm.com  Tue Dec  5 14:39:43 2000
From: ldw at us.ibm.com (Larry Whitley)
Date: Tue, 5 Dec 2000 13:39:43 -0600
Subject: Graphs from python
References: <90gse2$14uo$1@news.rchland.ibm.com> <90ihje$hmb$1@reader1.imaginet.fr>
Message-ID: <90jg9u$159e$1@news.rchland.ibm.com>

Thanks for the pointer.  I'll check it out.

Larry

"Gilles Lenfant" <glenfant at equod.com.nospam> wrote in message
news:90ihje$hmb$1 at reader1.imaginet.fr...
> Hi,
>
> You can use reportlab package to draw your graphs in a PDF file. It's use
is
> identical whatever OS you use (I did not test BeOS or Macintosh but it's
> supposed to work too).
> reportlab provides a low level API to the PDF document but making classes
> for various charts should not be that difficult.
> Get reportlab (free) from http://reportlab.com
>
> "Larry Whitley" <ldw at us.ibm.com> a ?crit dans le message news:
> 90gse2$14uo$1 at news.rchland.ibm.com...
> > The simulation models I write and run generate a lot of data and to make
> > sense of it all I put the data into a set of graphs.  The graphs are the
> > typical spread sheet fare, bar graphs, X/Y graphs, scatter plots, etc.
> > Today, I read the data into spreadsheets from the spreadsheet and create
> my
> > graphs by hand and I'd like to automate the process.
> >
> > I'm looking at using win32com (by Mark Hammond) to drive excel or
> > piddle+graphite to generage graphs directly from python.  While I'm sure
> it
> > will get the job done, the excel object model is a bit overwhelming.
> Piddle
> > by itself isn't at the right level.  Graphite appears get to the right
> level
> > but the pre-alpha rating makes me think that it's a bit early to jump
in.
> >
> > Are there other approachs I should consider?  Whatever I end up with,
I'd
> > like it to work from either Unix or Windows.  If I can't have both, I
can
> > make do with just windows.  Of the two approaches above, is one
> > significantly better than the other (by whatever measure you want to
> > choose)?
> >
> > Thanks,
> > Larry
> >
> >
>




From smoriano at sicon.net  Mon Dec 11 07:48:38 2000
From: smoriano at sicon.net (Sergio Boix Moriano)
Date: Mon, 11 Dec 2000 13:48:38 +0100
Subject: Validating xml files
Message-ID: <3A34CD26.DDE8F341@sicon.net>

    Hi all,

        I?m using the PyXML-0.5.5.1 library to work with xml files. We
generate the xnl files from a database and we need to validate this
files cause later we have to do an insert in another dataabse with the
xml contents, and the file must only have fields and tables that exits
in the database.

       I?ve tried to use the DTDParser() class, but lokking at the code
i?ve seen that the functions in the class doesn?t do nothing. The code
of this classes in the saxlib module is:
class DTDHandler:
    def notationDecl(self, name, publicId, systemId):
        "Handle a notation declaration event."
        pass

    def unparsedEntityDecl(self, name, publicId, systemId, ndata):
        "Handle an unparsed entity declaration event."
        pass


    Does anyone know anyway to read an xml file validating it with a
DTD, in a way that if the xml is bad formed it gives me an error.

    Thanks in advance,






From aleaxit at yahoo.com  Mon Dec 11 04:57:19 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 11 Dec 2000 10:57:19 +0100
Subject: Fault trees structure
References: <D_IY5.49$17.2785@m2newsread.uni2.es>
Message-ID: <9128e902uo3@news2.newsguy.com>

"Arturo P?rez Mulas" <arturo_perez at wanadoo.es> wrote in message
news:D_IY5.49$17.2785 at m2newsread.uni2.es...
Hi,

I am creating a fault-trees (kind of binary trees, but with unlimited inputs
in every gates and several different types of gates/inputs) management
package; thouhg I have created a basic structure for the class, I am not so
happy with it and I am certain that it can be improved greatly.

Any one knows of an Object Oriented description of a similar structure, or
is interested in this so we can discuss further?


Fault-tree analysis does appear like an interesting subject, but I
can't easily find a precise model of "a fault-tree" around the web.

Can you point to a summary description ("fault-tree analysis" gets
a lot of hits, but I wonder if a "fault-tree object" can be described
well enough without getting into the depth of the analysis issues)?

Discussing how to model something in terms of Python classes &c seems
to be a rather basic purpose of this forum, so I think it would be
all right to pursue a discussion here, but you'd have to supply the
domain-expertise (at least if I'm the other discussant...!).


Alex






From tlownds at my-deja.com  Sun Dec 24 12:23:49 2000
From: tlownds at my-deja.com (tlownds at my-deja.com)
Date: Sun, 24 Dec 2000 17:23:49 GMT
Subject: PEP 222 draft
References: <slrn947bj1.1rs.amk@207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com>
Message-ID: <925bf1$5tk$1@nnrp1.deja.com>

In article
<slrn947bj1.1rs.amk at 207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com>,
  akuchlin at mems-exchange.org wrote:
>     Utility function: build a query string from a list of 2-tuples
>

Perhaps it would be better to extend urllib.urlencode to handle this
case. I think it should be extended to handle lists as values as well:

>>> urlencode({'x': ['a', 'b']})
'x=%5b%27a%27,+%27b%27%5d'

would be:

'x=a&x=b'

>     The cgi.py module would be deprecated.  (XXX A new module or
>     package name hasn't been chosen yet: 'web'?  'cgilib'?)
>

cgilib gets my vote.

>     * An HTML generation module is not part of this PEP.  Several such
>       modules exist, ranging from HTMLgen's purely programming
>       interface to ASP-inspired simple templating to DTML's complex
>       templating.  There's no indication of which templating module to
>       enshrine in the standard library, and that probably means that
>       no module should be so chosen.
>

There are more ASP-like modules out there than other types...

>     Parsing functions: carry over most of the parse* functions from
>     cgi.py
>

It would be really nice to change the main parse-form-input function to
return a dictionary of plain strings instead of the FieldStorage
instances. When a file is uploaded, a FieldStorage instance for that
value makes sense. You've suggested this before, I may have missed it in
your PEP.

>
>     # Module-level function
>     def wrapper(func, logfile=sys.stderr):
>         """
>         Calls the function 'func', passing it the arguments
>         (request, response, logfile).  Exceptions are trapped and
>         sent to the file 'logfile'.
>         """
>         # This wrapper will detect if it's being called from the
command-line,
>         # and if so, it will run in a debugging mode; name=value pairs
>         # can be entered on standard input to set field values.
>         # (XXX how to do file uploads in this syntax?)
>

Regarding debugging: A command-line debugging ability is very useful!
But why tie that feature to the wrapper feature? You can put it in the
parse-form-input function. Also, allow name=value pairs as command-line
arguments. name=value pairs would still be read even with command-line
arguments, unless a final command-line argument "go" is present.

One debugging feature that would be very useful in a wrapper, is a way
to start pdb either right away or at an exception.

The wrapper passes (request, response, logfile) - why not pass an object
with .request, .response, and .logfile attributes. Using a tuple now
makes it harder to extend later. A cgi.log-like method on that "context"
object would be useful too!

-Tony Lownds



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


From pbjorn at uswest.net  Tue Dec 26 13:36:12 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Tue, 26 Dec 2000 11:36:12 -0700
Subject: Formatting for hh:mm:ss
References: <Pine.GSO.4.21.0012261309370.13920-100000@z.glue.umd.edu>
Message-ID: <3A48E51C.1B346AE6@uswest.net>

This should work (although it might be easier doing it directly...):

 def hms(secs):
  import time
  time_tuple = (0,0,0,0,0,secs,0,0,0)
  time_val = time.mktime(time_tuple)
  time_local = time.localtime(time_val)
  return time.strftime('%X', time_local)

-- bjorn


Roy Katz wrote:

> Hello,
>
> I've searched the 'time' and 'locale' modules  for any function which
> would take a difference of two time.time()'s and format the result as
> hour:minutes:seconds, but I can't find any.  All I am doing is this:
>
>   t1 = time.time()
>   # some processing here
>   t2 = time.time()
>
>   total = t2 - t1
>
>   print_time( total )
>
> where print_time() would format the time as above. Does anyone know if
> this function already exists?
>
> Thanks,
> Roey



From fredrik at effbot.org  Tue Dec 12 02:33:24 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 12 Dec 2000 07:33:24 GMT
Subject: [Q] Are Exceptions used that much in practice?
References: <3A35D546.A4731B5B@enteract.com>
Message-ID: <8xkZ5.3256$O%.522600@newsc.telia.net>

Jerome Mrozak wrote:
> I realize that Python allows me to catch exceptions.  But being a novice
> I'm not sure if they are used much in the 'real world'.

have you written any Python code at all?

</F>




From keisari_ at hotmail.com  Thu Dec 21 10:20:08 2000
From: keisari_ at hotmail.com (joonas)
Date: Thu, 21 Dec 2000 17:20:08 +0200
Subject: Testing if a list exist
Message-ID: <3A421FA8.61A49D48@hotmail.com>

How can I test if a list exists or not.

I have used

##############
try:
 temp = mylist
except:
 #do stuff if mylist does not exist
else:
 #do sfuff if mylist exists
##############


Is there any better way to do that?

Joonas.


From fcnpereira at home.com  Sat Dec 30 00:21:17 2000
From: fcnpereira at home.com (Fernando Pereira)
Date: Sat, 30 Dec 2000 05:21:17 GMT
Subject: Erlang (was Re: Good introduction to functional programming with Python?)
References: <mailman.977984412.30211.python-list@python.org> <92f0il01a5u@news1.newsguy.com>
Message-ID: <301220000021179048%fcnpereira@home.com>

In article <92f0il01a5u at news1.newsguy.com>, Alex Martelli
<aleaxit at yahoo.com> wrote:

> "Tim Peters" <tim.one at home.com> wrote in message
> news:mailman.977984412.30211.python-list at python.org...
>     [snip]
> > I haven't used Erlang at all, so that makes it an excellent battlefield:
> no
> > unpleasant haggling over facts <wink>.  I'll hazard a guess based on the
> 
> Right!
> 
> > In the absence of higher-order functions, I can believe programmers don't
> > have much trouble reasoning about types without declarations.  They don't
> 
> The _scarcity_ (not _absence_:-) of HOF in typical Erlang use may be a
> cultural issue (or maybe they just don't buy one all that much in its
> typical use cases); and from that may follow the relative lack of
> interest in its (existing, but 'add-on') type-inferencing possibilities.
The historical reason is that Erlang was originally inspired on logic
and first-order functional programming ideas rather than on
higher-order functional programming. If I remember correctly, the first
Erlang implementation was built on top of Prolog. However, much of
Prolog's machinery (backtracking in particular) was in the way for
concurrent programming (lots of research and ink went into sorting this
out in the logic programming community), so the Erlang team simplified
the language and implementation to still have pattern-matching a la
Prolog but a simple non-backtracking rewriting semantics. The
higher-order additions came later, I believe.

-- F


From tim at degree.ath.cx  Wed Dec 13 01:24:29 2000
From: tim at degree.ath.cx (Tim Hammerquist)
Date: Wed, 13 Dec 2000 06:24:29 GMT
Subject: [Q] Are Exceptions used that much in practice?
References: <3A35D546.A4731B5B@enteract.com> <3A36596A.89EEA210@alcyone.com> <3A36C6E3.26181E3F@enteract.com>
Message-ID: <slrn93e5kj.6o8.tim@degree.ath.cx>

Jerome Mrozak <goose at enteract.com> wrote:
> I believe I've been misunderstood.  The problem is not "can I write code
> without handling exceptions" but rather "are exceptions used
> voluntarily, rather than by compulsion".

I thought the overwhelming response was "YES!", but maybe I'm just
delirious.

> The long-range goal of my question is to determine for my satisfaction
> whether it is worth while for me to spend my limited time pursuing
> Python as a compliment to (whatever other skills I might have) or to go
> with the apparently much greater flow and use those resources on Perl. 
> The 'exception' thing could be a determiner, esp. with nothing similar
> in Perl, but only if I can use it in a *positive* way instead of just to
> test a return value.  Heck, Perl has 'die' for that kind of testing.

Perl has signal-trapping on top of the 'eval(...);if($@){...}' struct, on top
of being able to disguise the later as the much more recognisable:

: try {...}
: catch {...}

along with the ability to pass custom objects, in addition to the normal
strings, via the $@ variable.  And although Python's implementation is
much more graceful, Perl can by all means accomplish the same task.
Both languages are extremely powerful and flexible.  You won't be
limiting yourself with either language in terms of what you can
accomplish.  As far as I can tell, the choice between Python and Perl is
more a matter of what you're comfortable with and which fits your
personality.  I must have multiple personalities, because I won't
choose.  =)

And besides, if you ask "What's better: Python or Perl?" in a python NG,
what answer do you expect?

Try both.  Perl is a significantly different style from most people's
experience, unless that experience consisted highly of Unix shell
programming.  Python may suit your current skills (whatever, if any,
they might be) better.  Don't let us decide for you.

And yes, exceptions are used voluntarily and frequently by thousands of
programmers daily because it makes their jobs easier and they can be
fun.

And if THAT doesn't answer your question...  =)

-- 
-Tim Hammerquist <timmy at cpan.org>

Not all who wander are lost.
	-- J.R.R. Tolkien


From pete at visionart.com  Tue Dec  5 13:34:23 2000
From: pete at visionart.com (Pete Shinners)
Date: Tue, 5 Dec 2000 10:34:23 -0800
Subject: debugging python extensions in c
References: <90iv44$mho$1@news.kth.se>
Message-ID: <90jchf$5s0$1@la-mail4.digilink.net>

"Oivvio Polite" <oivvio at cajal.mbb.ki.se> wrote
> We've just started writing a python extension in c.
>
> I know how to use the python debugger and I've  done some c debugging
> before.
>
> But how do I go about to debug pointerstuff and memory leaks in python
> extension.
>
> We're writing this for both windows and linux.

i've been getting along with my trusty debugger, "printf"

sorry it's not the most elegent thing, but it's been solving
my bugs here and there. you may also want to change the compile
flags and recompile python. there are a lot of internal #define
flags that can be used for finding leaked references, more error
checking, etc.






From DanielK at jBASE.com  Thu Dec  7 13:18:20 2000
From: DanielK at jBASE.com (Daniel Klein)
Date: Thu, 7 Dec 2000 10:18:20 -0800
Subject: Python naming conventions
Message-ID: <wAQX5.28$Mk1.9001@typhoon.aracnet.com>

Are there any formal online documents describing Python naming conventions
and taboos?

Thanks,
Daniel Klein (owner of 4 parrots but none of them are dead yet :^) )




From jkndeja at my-deja.com  Mon Dec 18 08:44:57 2000
From: jkndeja at my-deja.com (jkndeja at my-deja.com)
Date: Mon, 18 Dec 2000 13:44:57 GMT
Subject: Solution: Direct access to Printer I/O lines
References: <917k9b$9i9$1@nnrp1.deja.com> <slrn93hrjn.2vp.ssthapa@localhost.localdomain> <91cnp3$g59$1@nnrp1.deja.com> <91d23f0jk7@news1.newsguy.com>
Message-ID: <91l4co$jua$1@nnrp1.deja.com>

Hi Alex,

OK, without (yet...) going too deep into the more-generalised-and-
interesting patterns discussion developing here:

If I/you/we don't want singleton objects, but _do_ want to avoid having
multiple processes accessing a single port, how about this:

class directlpt:
    def __init__(self, LPTNum):
        # gives me access to an LPT port
        ...
    def inb(self, offset):
        # I'm always allowed to read the port
        ...
    def inw(self, offset):
        # likewise
        ...
    def claim(self, timeout = ??):
        # I have to claim the resource to be able to write to it
        ...
    def outb(self, offset)
        # otherwise I get an exception (IOERrror?) here...
    def outb(self, offset)
        # and here

    def release(self):
        # I release it for other processes as soon as I'm done



    Thanks
    Jon N


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


From aleaxit at yahoo.com  Tue Dec 19 14:49:13 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 19 Dec 2000 20:49:13 +0100
Subject: Events in win32com
References: <Y1g=Op6IfWSoJn1ISFtOyuvqwBb3@4ax.com> <FY0=OlZFLy71DtnK62u5TEPlhOts@4ax.com>
Message-ID: <91oe6301tfj@news2.newsguy.com>

"Paul Moore" <paul.moore at uk.origin-it.com> wrote in message
news:FY0=OlZFLy71DtnK62u5TEPlhOts at 4ax.com...
> On Tue, 19 Dec 2000 13:49:23 +0100, Paul Moore
> <paul.moore at uk.origin-it.com> wrote:
>
> >If I create an object using
> >
> >obj = win32com.client.Dispatch(progid)
> >
> >can I see events from the object?
>
> After scanning Deja, I found an article describing DispatchWithEvents.
> The example in the documentation (OnVisible) works fine. But the
> OnQuit event doesn't seem to work :-( Basically, the event routine is
> never fired. Is this because the event routine has a ByRef "Cancel"
> parameter? Or is it something else?

If the event expects to call a method with one parameter (as it
will, if it's [in,out]; but not if it's just [out] -- this is a key COM
level distinction, but VB tends to the former, not supporting the
latter...) then it will silently fail to call a method that takes no
parameters even if the names coincide.  (Make sure to return the
value you receive as input as your result, to get [in,out] effects).

Also, you need (in general) for makepy to have been run on that
type library, or something equivalent that has processed the tlib
for Python used and placed it in the cache directory of known
type libraryes (the, I believe, EnsureDispatch method of the
gencache module, for example).  Python has to guess if it does
not know the specifics of the type library (just as VB would have
to guess in the same case) and it won't always guess right!-)


Alex





From aahz at panix.com  Tue Dec 26 23:50:25 2000
From: aahz at panix.com (Aahz Maruch)
Date: 26 Dec 2000 20:50:25 -0800
Subject: One Python 2.1 idea
References: <mailman.977771949.17911.python-list@python.org> <slrn94ie9k.anf.neelk@alum.mit.edu> <92bamo$roe$1@panix3.panix.com> <slrn94ihni.anf.neelk@alum.mit.edu>
Message-ID: <92bseh$noh$1@panix3.panix.com>

In article <slrn94ihni.anf.neelk at alum.mit.edu>,
Neelakantan Krishnaswami <neelk at alum.mit.edu> wrote:
>
>A webpage on what threading is (it's unrelated to multithreading, just
>a case of overloaded terminology) and to what extent it is a good
>thing can be found at:
>
>  http://www.complang.tuwien.ac.at/forth/threaded-code.html

<grin>  Yeah, and it's the *older* meaning (I remember reading about
threaded interpreters in conjunction with Forth back around 1980).
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

Ninth Virtual Anniversary: 5 days and counting


From darrell at dorb.com  Tue Dec 26 17:14:50 2000
From: darrell at dorb.com (Darrell Gallion)
Date: Tue, 26 Dec 2000 17:14:50 -0500
Subject: One Python 2.1 idea
References: <LNBBLJKPBEHFEDALKOLCCEIFIFAA.tim.one@home.com>
Message-ID: <037601c06f89$43d9e510$6401a8c0@home>

From: "Tim Peters"
> [Thomas Wouters]
> > ...
> > Actually, it's even worse :) Methods are never stored in an
> > instance's dict, but are created 'on the fly' from the class's
> > unbound method. The binding of the method isn't too expensive,
> > but it isn't totally free either (if anything, it's a dict 'miss'
> > and an aditional dict lookup in the class's dict. More than one
> > if you use deep or multiple inheritance (or both.))
> >
> > You *can* speed it up, by doing 'x.foo = x.foo' somewhere early
> > in your call chain :) ...
>
> Note:
>
> >>> class Oops:
> def foo(self):
> pass
>
> >>> x = Oops()
> >>> x.foo = x.foo
> >>> x.foo.im_self is x
> 1
> >>>
>
> Oops!  That is, x is now involved in a reference cycle (x's dict contains
a
> bound method that points back to x).  Before 2.0, that's a memory leak.
In
> 2.0, it's a leak only if the class (unlike Oops above) has a __del__
method
> (or you've disabled cyclic gc).
>

You lost me here Tim.

>>> class A:
...     def f(self):
...             pass
...
>>> a=A()
>>> a.f.im_self is a
1
>>> a.f=a.f
>>> a.f.im_self is a
1
>>>

Don't see what's changed after a.f=a.f

--Darrell







From fgeiger at datec.at  Tue Dec 26 10:02:15 2000
From: fgeiger at datec.at (Franz GEIGER)
Date: Tue, 26 Dec 2000 16:02:15 +0100
Subject: Exception Handling and Inheritance
References: <921sdr$mlq$1@newsreaderg1.core.theplanet.net> <922les$bkr$1@panix2.panix.com>
Message-ID: <92c41f$n1p$3@newsreaderg1.core.theplanet.net>

> No, but you could easily implement a function to do that using the
> traceback module.
>

Admitted, that's almost always a solution if other things do not work. But I
wondered..., well, you know.

> WARNING: it's generally considered a Bad Idea to catch exceptions
> willy-nilly.  For example, do you really want to trap the keyboard
> interrupt?

Well, I do not agree here. I do not want the user seeing any uncaught
exception without any message of MY program. Visual Basic gives dreaded
examples here: "Runtime error 5 in MY.EXE" or something like that.

If tests yield, that there are exceptions needing special care, I will
handle them differently. If there where a feature like exception decoration,
the language could force me to consider different exceptions differently
from the beginning. But as I understand, scripting people do not like to be
forced to anything ;-)

Best regards
Franz GEIGER

"Aahz Maruch" <aahz at panix.com> wrote in message
news:922les$bkr$1 at panix2.panix.com...
> In article <921sdr$mlq$1 at newsreaderg1.core.theplanet.net>,
> Franz GEIGER <fgeiger at datec.at> wrote:
> >
> >But what about Exceptions? They are not caught by my TException handler!
And
> >Exception does not have my Formatted method.
>
> No, but you could easily implement a function to do that using the
> traceback module.
>
> >Do I really have to catch them in a 2nd handler? Or can I add some
> >functionality somehow to Exception instead of implementing it in a
derived
> >TException?
>
> Yes, you'll need to catch other exceptions explicitly.
>
> WARNING: it's generally considered a Bad Idea to catch exceptions
> willy-nilly.  For example, do you really want to trap the keyboard
> interrupt?
> --
>                       --- Aahz (Copyright 2000 by aahz at pobox.com)
>
> Androgynous poly kinky vanilla queer het    <*>
http://www.rahul.net/aahz/
> Hugs and backrubs -- I break Rule 6
>
> '"Crisp" is a good quality for crackers; less so for pot roast.'  --pnh










From e_d_k at my-deja.com  Tue Dec 26 10:10:58 2000
From: e_d_k at my-deja.com (EdK)
Date: Tue, 26 Dec 2000 15:10:58 GMT
Subject: Help: Can't install any packages
Message-ID: <92acdv$eid$1@nnrp1.deja.com>

I'm runing Python 1.6 on Windows98.  When I start Python it identifies
itselas:
Python 1.6 (#0, Sep  5 2000, 08:16:13) [MSC 32 bit (Intel)] on win32

I have no problems as long as I use only packages that came with the
instalation, but when I try to install new packages I always get :

ImportError: DLL load failed: One of the library files needed to run
this application cannot be found.

None of the zip files I have downloaded contained any .dll files, but
they do contain .ptd files. I can not find any documentation for whay a
.pyd file is. These .pyd file all exist on my PYTHONPATH.

Can anyone help me? I'm running out of ideas here. I like python, but if
I can't use any third party packages I'll have to go back to C++ to get
any real work done.

   -EdK




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


From tdelaney at avaya.com  Tue Dec 12 18:41:15 2000
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 13 Dec 2000 10:41:15 +1100
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B0FBCA3@natasha.auslabs.avaya.com>

> > (Hint: Everything is a reference to an object.  Variables are
> > named references, not actual objects.  Some objects can be
> > modified in place.  Some cannot be modified.  That's all)
> 
> Okay! tell me, which are the objects which cannot be modified 
> in place? I
> assume these are int's, long's, float's, etc.  Strings, 
> lists, objects and
> dictionaries are call-by-reference.  Pardon my choice of 
> diction, but I am
> against inventing new terminiologies (modified-in-place??). 

Objects are generally documented as mutable or immutable. As an example:

	tuple is immutable
	list is mutable

In both cases, you can pass a reference to the object to a function. The
function aquires a *separate* reference to the object.

Now, if you assign something else *to that reference* you change what that
reference is referring to. When you exit the function the original reference
still refers to the original object.

However, you can modify the contents of the list that the reference is
referring to. In this case, the contents of the object that the original
reference is referring to (which is the same object as the reference in the
function is pointing to) will change.

If you try to modify the contents of the tuple that a reference is referring
to you will have an exception thrown, because it is immutable.

Likewise, you pass around references to other immutable items such as
integers. The difference is that they are immutable *and* they don't have
"contents" as such. You can change object the reference to the integer is
pointing to, but every other reference to that integer remains unchanged.

Note: you list string as "call-by-reference". Yes, it is. However, you can't
modify the contents of a string (they are immutable). You can assign a
different string to the reference - but the original reference still points
to the original string.

> I want to know where this is documented.  Specifically, where is the
> dichotomy of ints/longs/floats vs. 
> objects/strings/lists/dictionaries. 

There is no dichotomy. Everything behaves the same.

You appear to be thinking of python in C terms. In this case, consider
python references as pointers. You *only* pass pointers. Some of these
pointers may point to a const memory location. Some don't. If you change
what a copy of the pointer is pointing to, you don't change what the
original is pointing to. If you change the contents of the memory location
is that is being pointed to, dereferencing any pointer which points there
will expose the change.

> While I'm brainstorming with Python grammar changes, what 
> about a real &
> operator.  No more of this hokey is-it-modifiable-in-place-or-not
> business.  Explicit control is important. 

IMO there should be an explicit immutable() function and an immutable
keyword.

The function would return the original object, but mark it as immutable
(thus, you could simulate tuples by calling immutable() on a list).
Attempting to modify such an object would throw an exception.

The keyword would be used to mark parameters to functions.

Note that the immutability would be based on *which reference you accessed
the contents through*. Thus you could have a multi-threaded app which has
one thread with a mutable reference, and all the others with immutable
references. Suddenly you don't need to worry about locking an object to
write to it!

In both cases, I envisage that it would go from one dictionary holding
references in each namespace, to two - one for references to immutable
objects, the other holding references to (possibly) mutable objects (i.e.
any which have not been marked as immutable).



From jp at ulgo.koti.com.pl  Sun Dec 24 08:08:51 2000
From: jp at ulgo.koti.com.pl (Jacek =?iso-8859-2?Q?Pop=B3awski?=)
Date: 24 Dec 2000 13:08:51 GMT
Subject: string split
References: <slrn94b2go.srm.jp@localhost.localdomain> <Mkj16.188569$U46.5832132@news1.sttls1.wa.home.com>
Message-ID: <slrn94btb3.14v.jp@localhost.localdomain>

Greg Jorgensen wrote:
>    rx = re.compile(r'\s*(<[^<>]*>)\s*')

this solution is better, that this with separator in string.split, becouse
it doesn't eat it (['one','br','three']), but - I must learn this regexp
stuff one day! 
thanks :)

-- 
Love is automatic pleasure
Virtual reality
Terminal hate - it's a calculation 
Send in the child for connection             "Computer God" - Ronnie James Dio


From mfletch at tpresence.com  Mon Dec 18 14:55:02 2000
From: mfletch at tpresence.com (Mike Fletcher)
Date: Mon, 18 Dec 2000 14:55:02 -0500
Subject: Timers in Python?
Message-ID: <CC59868A0A76D311B1F50090277B321C1F3758@VRTMAIL>

class TimeScheduled(threading.Thread):
    def __init__(self, function, delay):
        self.function = function
        self.delay = delay
        threading.Thread.__init__( self )

Python doesn't automatically call the superclass __init__, you need to do it
explicitly if you want it called.

HTH,
Mike

-----Original Message-----
From: Daniel.Kinnaer at AdValvas.be [mailto:Daniel.Kinnaer at AdValvas.be]
Sent: Monday, December 18, 2000 2:18 PM
To: python-list at python.org
Subject: Re: Timers in Python?


On Mon, 18 Dec 2000 18:46:13 +0100, "Gilles Lenfant"
<glenfant at equod.com.nospam> wrote:

>Have a look to "threading" package and time.sleep(...) function.
>Seems that's all you need!
>Important, use threading.Lock() objects if threads may use not sharable
>resources.
>Warning, binary distributions of Python/Unix are not always compiles with
>threads options.
>You may need to compile Python from sources if required.
>

Dear Gilles,
Thanks for the reply.  It seems I ran into trouble when trying the
above code.  Python replied me the following :

timers.py :

import threading, time

Mylock = threading.Lock()

class TimeScheduled(threading.Thread):
    def __init__(self, function, delay):
        self.function = function
        self.delay = delay

    def run(self):
        self.alive = 1
        while self.alive:
            Mylock.acquire()
            # optional (depends whether unsharable resources used or
not)
            self.function()
            Mylock.release() # optional
            time.sleep(self.delay)

    def kill(self):
        self.alive = 0

def sayhello():
    # Very complex stuff
    print "Hello world"

# Every 2 minutes
scheduledHello = TimeScheduled(sayhello, 120)
scheduledHello.start() #this is line28

scheduledHello.kill()



Reply by Python : 

Traceback (innermost last):
  File "C:\Python20\AProject\timers\timers.py", line 28, in ?
    scheduledHello.start()
  File "c:\python20\lib\threading.py", line 351, in start
    assert self.__initialized, "Thread.__init__() not called"
AssertionError: Thread.__init__() not called

Is there something that can be done about it?  Thanks

Daniel
-- 
http://www.python.org/mailman/listinfo/python-list



From fredrik at effbot.org  Wed Dec 13 03:57:04 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 13 Dec 2000 08:57:04 GMT
Subject: Thank you, Fredrik. (was Re: Label-Value)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <WntZ5.3566$Kd1.414873@newsb.telia.net> <Pine.GSO.4.21.0012121550530.19682-100000@y.glue.umd.edu> <9009628F0PaCmAnRDLM@161.48.76.20>
Message-ID: <ARGZ5.3707$Kd1.429323@newsb.telia.net>

Fred Pacquier wrote:> As Fredrik says, this info certainly exists elsewhere and could be pieced
> together from existing material. But having it summed up clearly and in few
> words in one place can be a revelation, so this definitely needs to go in
> python-snippets or FAQTS or somewhere...

note to snippet/faqts maintainers (and other interested parties):

I've fixed a couple of typos, and posted the article here:

    http://effbot.org/guides/python-objects.htm

if you repost the material, please include a pointer to that page.

thanks /F




From qrczak at knm.org.pl  Mon Dec 18 15:31:18 2000
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 18 Dec 2000 20:31:18 GMT
Subject: while true: !!!
References: <B43D149A9AB2D411971300B0D03D7E8B0FBCBB@natasha.auslabs.avaya.com> <v04220801b663769e1800@[192.168.11.33]> <mailman.977147642.18708.python-list@python.org> <91lclt02qr3@news1.newsguy.com>
Message-ID: <slrn93st0m.6c1.qrczak@qrnik.zagroda>

Mon, 18 Dec 2000 17:05:51 +0100, Alex Martelli <aleaxit at yahoo.com> pisze:

> I think it would be inconsistent, un-Pythonic, to add a
> fundamental type for 'conceptual' reasons, without any
> practical, measurable benefit.  For example, Python does
> away with the distinction between 'a character' and 'a
> string of length 1' -- a distinction that is religiously
> maintained in C, Pascal, Java, etc,

OTOH Python does not equate characters with integers. C, C++ and Java
do. The difference is conceptual, not in the representation. I don't
agree that having a separate boolean type would be inconsistent.

Scheme is dynamically typed and does have separate true and false
values. Lisp does not, but equates Booleans with the empty list and
symbol t rather than 0 and 1. Smalltalk is dynamically typed and has
separate Booleans.

I think it is only C and languages drawing conventions from C that
equate booleans with numbers, with "any non-zero" meaning "true"
and the canonical "true" value of 1.

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


From quinn at dinar.ugcs.caltech.edu  Sat Dec 16 21:17:02 2000
From: quinn at dinar.ugcs.caltech.edu (Quinn Dunkan)
Date: 17 Dec 2000 02:17:02 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <r0pzoi11lhz.fsf@black.cs.washington.edu> <5jvZ5.3588$Kd1.418770@newsb.telia.net> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com> <kuu28971kp.fsf@lasipalatsi.fi> <3A380E4E.8537A0FB@san.rr.com> <91a1vr02b02@news2.newsguy.com> <Co9_5.25465$x6.13756579@news2.rdc2.tx.home.com> <91bbq102mtk@news1.newsguy.com> <rWd_5.26820$x6.14119694@news2.rdc2.tx.home.com> <91d5el0me7@news1.newsguy.com> <rOr_5.31120$x6.15241856@news2.rdc2.tx.home.com>
Message-ID: <slrn93o8gj.tu0.quinn@dinar.ugcs.caltech.edu>

On Fri, 15 Dec 2000 16:38:47 GMT, Rainer Deyke <root at rainerdeyke.com> wrote:
>Yes, in-place modification should be supported.  I had something like this
>in mind:
>
>class UserInt:
>  def __init__(self, value):
>    self.data = value
>  def __add__(self, other):
>    if isinstance(other, UserInt):
>      return self.__class__(self.data + other.data)
>    else:
>      return self.__class__(self.data + other)
>  __radd__ = __add__
>  def __iadd__(self, other):
>    if isinstance(other, UserInt):
>      self.data += other.data
>    else:
>      self.data += other
>  # Repeat for all other operations
>
>UserInt as written is more properly UserNumber, since it supports longs ints
>and floats.

Just as a matter of style, would it be superior to add a __coerce__ method
that knows how to coerce ints, longs, and floats into UserInts, or make an
explicit check on each method?  __coerce__ saves a lot of typing, but I'm
generally wary of implicit things like coercion.  When I implemented a class
like this (a vector, actually), I used __coerce__, but after trying to wrap my
brain around the control flow, I was thinking maybe I wouldn't next time, even
though it seems to work fine.

Is this fear unfounded?  Is there a better way?


From garryknight at bigfoot.com  Fri Dec 22 20:29:00 2000
From: garryknight at bigfoot.com (Garry Knight)
Date: Sat, 23 Dec 2000 01:29:00 +0000
Subject: coding convention conversion
References: <mailman.977516469.1157.python-list@python.org>
Message-ID: <920v4s$oqj$2@taliesin.netcom.net.uk>

In article <mailman.977516469.1157.python-list at python.org> "Issac
Trotts" <trotts at llnl.gov> wrote:

> Does anyone know of a utility to change between the coding conventions
>  of
> 
> def fooBarBaz(): pass
> 
> and 
> 
> def foo_bar_baz(): pass
> 
> It is a pain to do this by hand.

Much though I hate to do this in a Python newsgroup, here's a Perl
script that should do the job. It reads lines from STDIN and copies
them to STDOUT. If a line starts with optional whitespace followed by
the word "def" followed by more whitespace followed by a name, then if
that name contains capital letters, each capital is lower-cased and has
a "_" prepended.

Copy and paste the text between the cut marks and save it as, say,
/usr/local/bin/redef (or the equivalent location on your system) and
use it as follows:
  redef <infile >outfile
It's about as simple as it can be and hasn't been tested exhaustively.
If there are errors, or if you need it enhanced, please let me know.

--------8X-cut here-----------------------------------------------
#!/usr/bin/perl

while (<>) {
  if (/^\s*def\s+\w+/) {
    s/([A-Z])/_\l\1/g;
  }
  print;
}
--------8X-cut here-----------------------------------------------

-- 
Garry Knight
garryknight at bigfoot.com


From ullrich at math.okstate.edu  Sat Dec 23 12:50:49 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Sat, 23 Dec 2000 17:50:49 GMT
Subject: A Suggestion for Python Colon Syntax
References: <mailman.977446446.16022.python-list@python.org>
Message-ID: <922olo$e43$1@nnrp1.deja.com>

In article <mailman.977446446.16022.python-list at python.org>,
  "Tim Peters" <tim.one at home.com> wrote:

> [...] (ditto the whole "while 1" thread
> that's been going on virtually non-stop for a decade).

and we'll never be certain whether it was an intentional
joke or not. Keen.

--
Oh, dejanews lets you add a sig - that's useful...


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


From dale at out-think.NOSPAMco.uk  Wed Dec  6 07:58:02 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Wed, 06 Dec 2000 12:58:02 +0000
Subject: update
References: <3A2D9A08.43F8FBE8@hypermath.org>
Message-ID: <8sds2t8hfgq54ad8ka43te25hvgcv6f4ot@4ax.com>

leonrom <leonrom at hypermath.org> wrote:

>I run v.1.5.2, do I have to uninstall it before installing v.2.0

I haven't seen recommendations one way or the other elsewhere so here's one I made earlier:

I recommend uninstalling the old version before intalling the new one.
--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants




From bradh at mediaone.net  Wed Dec  6 17:43:05 2000
From: bradh at mediaone.net (Brad Howes)
Date: Wed, 06 Dec 2000 22:43:05 GMT
Subject: DCOracle Memory Leak?
Message-ID: <m3g0k1qih5.fsf@h0050047c4794.ne.mediaone.net>

Has anyone seen a Python process using DCOracle grow and grow with each fetch
from an Oracle database? I can write a really simple loop that does a fetch but
does not save the value, and the Python process keeps growing. Are there known
bugs with Oracle 8i client libraries for Linux?

-- 
    "Were people this stupid before TV?" -- _White Noise_ by Don DeLillo


From syver at NOSPAMcyberwatcher.com  Wed Dec 20 15:55:26 2000
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Wed, 20 Dec 2000 21:55:26 +0100
Subject: Please help
References: <91qqjf$au4$1@nnrp1.deja.com>
Message-ID: <91r6gp$eo2$1@troll.powertech.no>

>Would you kindly help clear up
> some of my doubts about Python?
I'll try ;-)
> 1. I want to use IIS 4.0 for my webserver, can I use scripts compiled
> by Python in it?
Yes, but the scripts are not compiled.
You'll have to:
1. setup IIS to run .py files when requested in the url.
To do this select the default web site from inetmgr. Select the home
directory page. Click the configuration button. A new property page dialog
appears. Select the app mappings page and add a new mapping.

In the path edit box write the path to the python interpreter python.exe
followed by "%s" %s. ex: c:\folder\python.exe "%s" %s.
Set the extension to .py. That should be all. Test that it really works by
making a .py file with only this content:

2. Make a test python script.
if __name__ == '__main__':
   import cgi
   cgi.test()

Put the file in a folder that is shared by the IIS and enter the path to the
file in your favorite browser. It should print out a whole load of stuff.

3. Read the documentation on the cgi module to find out more about cgi
programming under python.



> 2. What about the ASP? If I want to use ASP, can I use Python to do
> such jobs like user authentication?
Yes.

>How well can Python be used in Web
> programming on Windows platform?
About as well as VbScript and Jscript I suppose if you're talking ASP. The
only limitation in my experience is that you can't do like this:
<%
if so and so:
%>
    html code here
<%
else:
%>
    html code here

Python has got libraries that you can use to the fullest extent where you'll
have to build COM objects in either VB (urhkhh!) or C++.

>It seems that I can not make a .exe file for it to execute.
Yes.
>It is dependent on the interpreter?
Yes. But there are ways making installations that packs together the
interpreter and your code, I don't have any experience with these, so that's
really all I can say.

> 4. If I really sit my self to Python, does it reward equally for the
> same effort in learning C? In the aspect of the quality of programs.
> Thank you in advance for answering my questions.

If you're thinking about productivity I guess you'll be handsomely rewarded
for choosing Python over C. What can only be expressed in C by writing many
lines of code can often be written in Python in just a few. On the other
hand there might be things you wish to do that is not possible in Python but
can be done in C, depends on what you want to do. Many modules in Python are
in fact written in C.




From darcy at vex.net  Sat Dec 30 12:53:06 2000
From: darcy at vex.net (D'Arcy J.M. Cain)
Date: 30 Dec 2000 17:53:06 GMT
Subject: Controlling processes (popen2)
References: <Wyo36.73$KD3.19486@typhoon.aracnet.com>
Message-ID: <92l7e2$1pbu$1@news.tht.net>

Daniel Klein <DanielK at aracnet.com> wrote:
> The problem is that in Python I have to terminate all output to the BASIC
> process with a \r (rather than \n). In Java, I can use either \n or \r.

I don't use Windows or BASIC so this is just a hunch.  What happens if
you popen with "wt" as the second argument?  That extra 't' is benign
under Unix.

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


From aleaxit at yahoo.com  Fri Dec 15 08:08:15 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 15 Dec 2000 14:08:15 +0100
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <915jcf0c6o@news2.newsguy.com> <9162p3$7t2$1@animus.fel.iae.nl> <MBWZ5.334$h67.37748@newsread1.prod.itd.earthlink.net> <3A397C74.6C8906E1@cosc.canterbury.ac.nz>
Message-ID: <91d5520m6j@news1.newsguy.com>

"Greg Ewing" <greg at cosc.canterbury.ac.nz> wrote in message
news:3A397C74.6C8906E1 at cosc.canterbury.ac.nz...
> Oldayz wrote:
> >
> > I thought something like forever: would be even more lucid
>
> If there is a break in the loop somewhere, it's
> not really forever, though, is it?
>
> So perhaps it should *really* be
>
>    for a while:
>       ...
>
> :-)

Why not have a 'built-in constant' it_lasts, of value 1,
so we can write:

    while it_lasts:
        ...

This requires no change to existing Python syntax.  For
the incurable romantics, besides allowing a 'for' statement
form with a single parameter so cynics can write

    for awhile:
        ...

and romantics can write:

    for ever:
        ...

we should also introduce an 'until' statement, and a
death_do_us_part built-in constant, of course.


Alex





From aleaxit at yahoo.com  Tue Dec 19 07:39:53 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 19 Dec 2000 13:39:53 +0100
Subject: problem with pickle.loads
References: <mailman.977223781.9501.python-list@python.org>
Message-ID: <91nkvi0143l@news2.newsguy.com>

"Oleg Broytmann" <phd at phd.pp.ru> wrote in message
news:mailman.977223781.9501.python-list at python.org...
> On Tue, 19 Dec 2000, Yusuf L?le wrote:
> > In a python script I create an object (an instance of a class) and call
the
> > pickle.dumps function
> > (serobj = pickle.dumps(obj)).
> > Then I put this serialized result (serobj) in an html page in an hidden
> > input box and sent it  to the next python script.
>
>    Wow! Please, give me the URL, I want to hack you! :)))
>
>    Do you understand that it is security hole? No, it is SCURITY HOLE!
> Don't do it. Find a different way to provide "session objects".

Encrypting would suffice, though keeping encrypted state data
in cookies seems like a better idea than using hidden input
fields for the purpose.


> > The second python script now reads the serialzed object from the
environment
> > of the webserver.
>
>    The problem is, perhaps, that either you screwed the object up, or
> browser did it for you. Pickles are binaries, so you need to
> urllib.urlquote_plus them.

Quoting is always a good idea, but you can also use the optional
second parameter to dumps to make its pickling 'less binary...':

>>> import pickle
>>> pickle.dumps(23.45,0)
'F23.449999999999999\012.
>>> pickle.dumps(23.45,1)
'G at 7s33333.'
>>> pickle.dumps(2345,1)
'M)\011.'
>>> pickle.dumps(2345,0)
'I2345\012.'
>>>

Hmmm, maybe it doesn't matter as much as all that, actually!-)


Alex





From aleaxit at yahoo.com  Mon Dec  4 03:36:31 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 4 Dec 2000 09:36:31 +0100
Subject: Python programs (or scripts) as drop targets
References: <hv5k2tg9cdf33j8v333qpkk7jo45ht596n@4ax.com> <90d9hi029hm@news1.newsguy.com> <3A2ACF21.4060406@ActiveState.com>
Message-ID: <90fl26022in@news2.newsguy.com>

"Mark Hammond" <MarkH at ActiveState.com> wrote in message
news:3A2ACF21.4060406 at ActiveState.com...
> Alex Martelli wrote:
>
> > I think IDropTarget must be implemented somewhere in Mark's code,
> > since *running* Python programs with win32all can be drop-targets.
> > (Dunno about IPersistXXX, but that shouldn't be a big problem).  But
> > that doesn't necessarily help with implementing what you desire, i.e.,
> > making a *shortcut* into a droptarget.
>
> Actually, the other way of getting it for a running program is Windows
> messages, and this is how win32all generally does it.  Indeed, win32gui

For a program that's *already* running, yes, there are several
ways it can handle having stuff being dropped on it (though I
think the COM solution is richer and more general).

> could knock this up OK.  It doesnt handle the "shell extensions"
> features Dale pointed at, where a new object is instantiated to handle
> the drop.

Right -- you do need COM for that.


> > But how this works for a _shortcut_ doesn't seem to be documented
> > at all.
>
> I understand the shortcut is simply resolved using the other shell APIs
> first.

So one loses the precious information that the shortcut has in
addition to the program-to-be-run, such as the flags, current
directory, etc?  Seems a suboptimal architecture to me (not that
this would be the first suboptimal architecture in the Win32 API
and surrounding subsystems, of course:-).


> > One thing we could do is invent a "new filetype" to be used *instead*
> ...
>
> To be quite honest, while I see that the "batch file" wrapper solution
> isnt as "clean" as it could be, I don't believe a whole bunch of C++
> code or registry additions are worth saving this one trivial file.

If saving a trivial file is indeed the only advantage of implementing
the shell extension, then you may be right (though quibbling about the
"registry additions" needed to _register_ a shell extension seems pretty
peculiar to me).


Alex





From root at rainerdeyke.com  Mon Dec 25 11:48:25 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Mon, 25 Dec 2000 16:48:25 GMT
Subject: One Python 2.1 idea
References: <mailman.977590932.19335.python-list@python.org> <9237md$orm$1@nnrp1.deja.com> <924bjt0e3n@news1.newsguy.com> <925baq$5sj$1@nnrp1.deja.com> <mailman.977737448.26875.python-list@python.org> <927q24$ond$1@nnrp1.deja.com>
Message-ID: <tTK16.73794$x6.34384327@news2.rdc2.tx.home.com>

<rturpin at my-deja.com> wrote in message news:927q24$ond$1 at nnrp1.deja.com...
> In article <mailman.977737448.26875.python-list at python.org>,
> Yes, this is a good performance optimization. BUT it
> requires NO CHANGE in how loops are written. The WRONG
> way to do this is to provide another construct in the
> language, making it more complex, but giving it no
> more expressive power. The RIGHT way to do this is
> for the compiler to recognize the common construct:
>
>     for i in range(<exp1>: <exp2>):
>
> The compiler then generates efficient byte code that
> does precisely what you describe. It can generate a
> minimal "ghost" of a list for the iteration, because
> there is no way that future code can do anything else
> with that list.

_range = range
ranges = []

def my_range(*args):
  r = _range(*args)
  global ranges
  ranges.append(r)

__builtins__.range = my_range


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From sdm7g at virginia.edu  Wed Dec  6 14:16:53 2000
From: sdm7g at virginia.edu (Steven D. Majewski)
Date: Wed, 6 Dec 2000 14:16:53 -0500 (EST)
Subject: Schwartzian Transform (was Re: Sort)
In-Reply-To: <90lrhl06t@news2.newsguy.com>
References: <NlnX5.1444$yv1.2700363@nnrp5.proxad.net> <oa8s2tsls5foll62r8fpa6nlcpisb6v1ek@4ax.com> <90l9ir02avk@news2.newsguy.com> <3A2E5F2A.9E650D50@erols.com> <90lrhl06t@news2.newsguy.com>
Message-ID: <Pine.NXT.4.21.0012061411450.230-100000@localhost.virginia.edu>


On Wed, 6 Dec 2000, Alex Martelli wrote:

> The technique is named after its inventor Randall
> Schwartz, who thought it up in the context of Perl
> (but it works even better in Python:-).  The key
> idea is that, in scripting languages such as Perl
> and Python, 'sort' slows down a lot if each of its
> O(N log N) comparison involves interpreting script
> code, rather than just the fast, built-in "lexical
> comparison" that the language provides.


This is however, a pretty ancient Python idiom, so I wonder
about attributing it to Randall's invention. 
( Unfortunately, it doesn't distill into an obvious regex 
  to grep thru the old mailings archives for a first use.) 

-- Steve Majewski <sdm7g at Virginia.EDU> 




From jwbnews at scandaroon.com  Tue Dec 26 00:03:29 2000
From: jwbnews at scandaroon.com (John W. Baxter)
Date: Mon, 25 Dec 2000 21:03:29 -0800
Subject: os.exec?
References: <LXw16.8561$KY1.23035@news1.rivrw1.nsw.optushome.com.au> <mailman.977727428.19738.python-list@python.org> <jwbnews-82EF1A.10532625122000@news.olympus.net> <3A47A179.64548C79@uswest.net>
Message-ID: <jwbnews-147D43.21032925122000@news.olympus.net>

In article <3A47A179.64548C79 at uswest.net>, Bjorn Pettersen 
<pbjorn at uswest.net> wrote:

> "John W. Baxter" wrote:
> 
> > [snip]
> > Or, moving to a higher level, there is the commands module, which wraps
> > those things mentioned above.  I tend to be lazy and use that.
> 
> The commands module is hopelessly unix centric when it doesn't need to be 
> so I
> consider it pretty much useless...

Well, I only do Unix (and Mac, where the issue doesn't arise).  So that 
has never bothered me.

   --John

-- 
John W. Baxter   Port Ludlow, WA USA  jwbnews at scandaroon.com


From rpm at wag.caltech.edu  Thu Dec 21 23:49:27 2000
From: rpm at wag.caltech.edu (Richard P. Muller)
Date: Thu, 21 Dec 2000 20:49:27 -0800
Subject: Good introduction to functional programming with Python?
Message-ID: <3A42DD57.9BCF7FA1@wag.caltech.edu>

Can someone recommend a decent introduction to functional programming in
Python? I haven't found any suitable intro, and I'd like to learn more
about map(), apply(), lambda, and the like. I understand the syntax, but
I'd like to learn something of the functional programming context these
commands come from.

Thanks in advance.

Rick


From ssthapa at harper.uchicago.edu  Thu Dec  7 03:26:35 2000
From: ssthapa at harper.uchicago.edu (Suchandra Thapa)
Date: Thu, 07 Dec 2000 08:26:35 GMT
Subject: parallel port access?
References: <3A268378.4DA1B831@sympatico.ca> <3A272881.6F444750@engcorp.com> <90lm04$b3r$1@nnrp1.deja.com>
Message-ID: <slrn92str2.12g.ssthapa@localhost.localdomain>

jkndeja at my-deja.com <jkndeja at my-deja.com> wrote:
>It is not advisable to be able to do this on a platform such as the PC,
>since you can crash your machine etc. very easily like this. Depending
>on the Operating System you are running, you might not even be allowed
>to do this even in Assembly language - it involves a change in
>operating system privileges which you may not be allowed to do.

    You can directly read and write ports on windows 95/98/me(?) but
you need to create a vxd to access ports on NT and 2000.  From my experience
it is relatively safely to directly modify the parallel port as long as
nothing else is trying to access it.  If you have control over the workstation,
you can insure the latter condition by not printing and/or only running 
one program that interfaces with the parallel port.  
    This approach has worked well in trying to control some 
electronics (a laser and some PMTs to be exact) over the parallel port.
The parallel port faq has some code samples and some pointers.  There is
also a DLL floating around but that is only a wrapper around inport and
outport calls.  You can probably use SWIG to get a python interface for
that.

-- 
------------------------------------------------------------------
			    |
Suchandra Thapa             | "There are only two kinds of math books. 
s-thapaNO at SPAMuchicago.edu  | Those you cannot read beyond the first 
			    | sentence, and those you cannot read 
			    | beyond the first page."
			    |                       -C.N. Yang
------------------------------------------------------------------


From emile at fenx.com  Fri Dec 22 13:21:18 2000
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 22 Dec 2000 10:21:18 -0800
Subject: Serial I/O (Windows) Documentation
References: <3A438D85.2C52324A@email.mot.com>
Message-ID: <92061t$5q220$1@ID-11957.news.dfncis.de>

It looks like the readme that comes with sio-151
(ftp://ftp.python.org/pub/python/contrib/sio-151.zip)is the
extent of the documentation.  You might take a look at using
COM, then http://members.aol.com/Bbirthisel/SerialPort.html
may also help some.

HTH,

--

Emile van Sebille
emile at fenx.com
-------------------


"Mark Wyatt" <tmf002 at email.mot.com> wrote in message
news:3A438D85.2C52324A at email.mot.com...
> Newbie Question:
>
> Have found several references to WIN 32 serial I/O. No one
seems to
> mention documentation however. What is available? What I
would really
> like is to find a tutorial on serial I/O programming on
the WIN 32/DOS
> platforms (UNIX is way down my list of things to support
for this little
> project). Failing that, any kind of documentation would be
a help.
>
> Thanks
>
> Mark




From wade at lightlink.com  Tue Dec  5 07:20:12 2000
From: wade at lightlink.com (wade at lightlink.com)
Date: Tue, 05 Dec 2000 12:20:12 GMT
Subject: codecs UTF-8 StreamReader ignores errors arg
Message-ID: <90imhq$t5m$1@nnrp1.deja.com>

I've run into a problem using the codecs module to decode data that is
mostly UTF-8, but with some bogus characters thrown in. The
StreamReader class seems to ignore an 'errors' argument passed to its
constructor, so it uses the default, which is 'strict'.

A short session illustrating the problem is shown below. Any advice
appreciated.

Wade Leftwich
Ithaca, NY

------------------------------------

>>> import codecs
>>> from StringIO import StringIO
>>> encode, decode, reader, writer = codecs.lookup('UTF-8')
>>> s = 'ab\346c'
>>> decode(s, 'replace')
(u'ab\uFFFDc', 4)
>>> fh = StringIO(s)
>>> sr = reader(fh, 'replace')
>>> sr.read()
Traceback (innermost last):
  File "<interactive input>", line 1, in ?
  File "c:\python20\lib\codecs.py", line 208, in read
    return self.decode(self.stream.read())[0]
UnicodeError: UTF-8 decoding error: unexpected end of data


Sent via Deja.com http://www.deja.com/
Before you buy.


From dsh8290 at rit.edu  Wed Dec 27 16:20:53 2000
From: dsh8290 at rit.edu (D-Man)
Date: Wed, 27 Dec 2000 16:20:53 -0500
Subject: classes (was Re: Same again please for OOP)
In-Reply-To: <92708d0g4g@news1.newsguy.com>; from aleaxit@yahoo.com on Mon, Dec 25, 2000 at 09:20:51AM +0100
References: <mailman.977623149.8184.python-list@python.org> <9257mf0sl1@news1.newsguy.com> <roy-E5B925.13324224122000@news.panix.com> <925khi019hs@news1.newsguy.com> <kubsu1t81x.fsf@lasipalatsi.fi> <92708d0g4g@news1.newsguy.com>
Message-ID: <20001227162053.A11064@westsidecnc.rh.rit.edu>

I had agreed with Erno with accessor methods, but Alex has convinced
me that accessing fields as fields is cool with Python's ability to
transparently convert it to a function call (giving the same results
as Erno's accessors and mutators, but allowing better looking client
code).  It would be really cool if the interpreter would do the magic
of looking for __get_attributename automatically, but Alex's solution
will have to do for now.

However, I can't get Alex's example code to work.  I either get
AttributeError or an infinitely recursive __getattr__.  The reason is
if I make a class Foo that has a function __get_value, the name is
mangled to _Foo__get_value.  If I reduce the leading underscores to
only 1 underscore, it works fine.

Have I missed something here Alex?

Thanks,
-D

On Mon, Dec 25, 2000 at 09:20:51AM +0100, Alex Martelli wrote:

<snip>
> 
> class AccessorsMixin:
>     def __getattr__(self, name):
>         if name.startswith('__'): raise AttributeError,name
>         return getattr(self, '__get_'+name)()
> 
> and implement method __get_location(self) the same way as you'd
> implement location(self) in your style.  What could be easier?  If
> you already have a __getattr__, it's equally OK to add this couple
> of lines to it, or delegate to this one (delegating has pluses).
> 
<snip>



From mail at codex.pe.kr  Sun Dec 24 00:28:59 2000
From: mail at codex.pe.kr (Ahn, Jae-wook)
Date: Sun, 24 Dec 2000 14:28:59 +0900
Subject: Can't load jpeg image with PIL (win32)
Message-ID: <9241ls$ks3$1@news1-2.kornet.net>

Hi.

I'm trying PIL(Python Image Library) on Windows 2000. I downloaded version
1.0 from Pythonware and tested. I extracted the zip file directly to the
Python directory (Active Python 2.0), and everything's ok except that it
can't load jpeg images.

It loads other image file types including gif, png, and bmp. The munual says
it supports jpeg images, and a file named 'JpegImagePlugin.py' is included
in the package.

Do I have to install other components like IJG jpeg library? Please let me
know how I can read and handle jpeg images with Python on Windows platform.






From matthewmNOSPAM at ihug.co.nz  Mon Dec 11 14:36:31 2000
From: matthewmNOSPAM at ihug.co.nz (Matthew)
Date: Tue, 12 Dec 2000 08:36:31 +1300
Subject: Newbie Thread question continues...
Message-ID: <9139uc$mrm$1@lust.ihug.co.nz>

Thanks for the replies to my first post.

The python docs state

There are two ways to specify the activity: by passing a
callable object to the constructor, or by overriding the run() method in a subclass. No other methods (except for the constructor) should be
overridden in a subclass. In other words, only override the __init__() and run() methods of this class. 

Okay, so you can have stuff like

class Dog(threading.Thread):
	def run(self):
		do stuff

dog = Dog()
dog.start()

I've got that far. However, it's generally been my understanding to NOT
subclass Thread unless it directly builds on Thread's functionality. So a
few more questions:-

a) Howto in Python do something like Java's:- 
new Thread(new Dog()).start() where Dog doesn't subclass Thread, but
rather wraps a Dog instance inside it's own thread.

b) In Thread, can someone please explain the use of 'target',
'args','kargs', and preferable give
a quck example to use the Thread constructor as in:

Thread (group=None, target=None, name=None, args=(), kwargs={}) 
       This constructor should always be called with keyword arguments. Arguments are: 

       group Should be None; reserved for future extension when a ThreadGroup class is implemented. 

       target Callable object to be invoked by the run() method. Defaults to None, meaning nothing is called. 

       args Argument tuple for the target invocation. Defaults to (). 

       kwargs Keyword argument dictionary for the target invocation. Defaults to {}.


THANK-YOU. matthew


From jlpeyret  Wed Dec  6 02:28:45 2000
From: jlpeyret (jluc)
Date: Wed, 06 Dec 2000 07:28:45 GMT
Subject: Mahagony EMail Client
References: <383a150e.12007736@news.telus.net>
Message-ID: <NUlX5.760104$8u4.13087197@news1.rdc1.bc.home.com>

no, you're not the only one.  multiplatform + python-scripted sounds great.
kinda cruel of them to dangle the premise of windoze support.  although i
think their website or sourceforge mentions that "the sources have been
known to compile on windoze", so maybe you can take that as an invite to
brew your own.

"David" <root at 127.0.0.1> wrote in message
news:383a150e.12007736 at news.telus.net...
>
> Am I the only Windows user who's dying for a full release of Mahagony?
>
> Please shout out your need -- perhaps it'll motivate the development team
> to live up to their month-and-a-half old promise that we'd see the Windows
> version "within a few days."
>
> :)
>
> [I've been patiently waiting move than two years, I'm certain.  Sigh.]




From scarblac at pino.selwerd.nl  Sun Dec 31 06:58:19 2000
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 31 Dec 2000 11:58:19 GMT
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <slrn94shbe.150.grey@teleute.rpglink.com> <X8s36.2731$Qb7.305339@newsb.telia.net> <slrn94smds.150.grey@teleute.rpglink.com> <FOt36.17$vd1.2224@e420r-atl2.usenetserver.com> <slrn94t1vj.150.grey@teleute.rpglink.com>
Message-ID: <slrn94u7ks.m8h.scarblac@pino.selwerd.nl>

Steve Lamb <grey at despair.rpglink.com> wrote in comp.lang.python:
> On Sat, 30 Dec 2000 23:00:53 GMT, Kragen Sitaker <kragen at dnaco.net> wrote:
> >Hahaha!  You just plonked Fredrik Lundh in comp.lang.python.  He was
> >being helpful and polite, and you plonked him because you disagreed
> >with him.  I can guess how much help you'll get next time you have a
> >Python question.  :)
> 
>     I can tell you I'll get far better without him than with him when he is
> giving advice that runs contrary to about 10+ years of programming that I've
> been taught from various different sources.

"I don't care if he's one of the few best posters ever on this newsgroup.
He's saying something I don't want to hear!! *plonk*".

> I don't care if he is God
> himself, advocating not initing a variable is stupid, plain and simple.

Nothing is ever as simple as that. You're giving that variable two different
meanings - first you use it as a flag that lets you enter the loop, then
you use it to store a temperature in. Yes, you initialize it, but you're
not initializing temperature! You are like those people who 
"#define THIRTEEN 13" because they should #define all the constants they
use in their program. 

One clue - if the guard of the loop ever changes,
you might need to change your initialization - that's an alarm bell. The
initial temperature didn't change, did it? If you need both a flag and
a temperature, use two variables.

> He
> wanted to argue the point for nothing more than argument's sake, so be it.  I
> don't need his advice any further.
 
You're applying some rule just for the rule's sake.

A mother was teacher her child to prepare a big meat dish, the way she had
done it for 20 years, and that her mother taught her. She explained which
spices to use, how to prepare the meat, and then two cut off two bits from
the sides and put it in the pan. The daughter was puzzled by this. "Why do
you cut off the sides?" "That's the way it's done. It's better. My mother
taught me, and doing it any other way is just plain stupid." The daughter
wasn't quite convinced, and asked her grandmother some other day. "Well, I
don't know why *she* does it, but I did it because my pan wasn't big
enough!"

The normal reason for always initializing a variable in many languages is
that their value might be undefined (random) otherwise. Not so in Python. In
Python, using an uninitialized variable will fail loudly, meaning you have a
control flow bug, most likely. Initializing variables when you don't need to
hides bugs. Don't simply apply rules.

-- 
Remco Gerlich


From grey at despair.rpglink.com  Mon Dec 18 13:51:49 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Mon, 18 Dec 2000 18:51:49 -0000
Subject: while true: !!!
References: <mailman.977127721.4398.python-list@python.org> <slrn93rsl2.qb2.grey@teleute.rpglink.com> <3A3E217C.5F2F4996@uswest.net>
Message-ID: <slrn93sn65.tvg.grey@teleute.rpglink.com>

On Mon, 18 Dec 2000 07:38:52 -0700, Bjorn Pettersen <pbjorn at uswest.net> wrote:
The most common reason for using "while 1:" (at least in newspostings ;-) is
>because saying what you _really_ mean <0.5 wink>:

>   for line in f.readlines():
>      ...

    Which...

>has the potential of crashing your machine with an out-of-memory error if you
>don't know the sizes of files you're operating on. 

    ...wastes memory.

>So, the status quo is that because the straight-forward way of doing it is
>too slow, we're all using an optimized idiom that while easy to understand if
>coming from a CS background, doesn't model what you're doing (at least not
>until you've programmed enough Python that reading a file seems like an
>endless loop <wink>).

    Or doing it in the 2-3 other languages I know of which use while in the
exact same manner.  I'm sure if we took a pole of the computer languages known
here you'd find that it is pretty much a standard across many languages, not
just Python.

    Hell, if you want to grouse about readability in Python, why not attack
some of the more obvious targets?  

while (x < 10):
  do.something(here)
  x = x + 1
else:
  print "What the hell does an else mean on a while loop?"
for x in y:
  do.something(here)
else:
  print "And here it is on for.  How do you else a loop?  Let's try this."
while 1:
  do.something(here)
  if (some.condition):
    break
else:
  print "So, is this executed when the loop completes, or when it breaks?"
  print "Which is more obvious to you?  If-else suggests the negative of the"
  print "if statement.  What's the negative of a loop, completion or the"
  print "break?  The answer lies in the same place that while 1: becomes"
  print "quite obvious, the documentation."

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From debl at NOtheworldSPAMMY.com  Fri Dec 29 20:57:07 2000
From: debl at NOtheworldSPAMMY.com (David Lees)
Date: Sat, 30 Dec 2000 01:57:07 GMT
Subject: Great New Book
References: <t4o1vr8hktk531@corp.supernews.com>
Message-ID: <3A4D40F3.134D839B@NOtheworldSPAMMY.com>

Actually, another new Python book has also come out recently, called
Python Developers's Handbook by Lessa.  It is comparable length
(actually about 150 pages more) than Core Python.  They both look like
they are aimed at newbies like me.

david lees


Jason Cunliffe wrote:
> 
> This morning I wes very happy to see in my local Barnes&Noble today a
> prominent stack of an excellent new book:
> 'Core Python Programming' by Wesley Chun.
> 
> http://www.amazon.com/exec/obidos/ASIN/0130260363/o/qid=978061249/sr=8-1/ref
> =aps_sr_b_1_3/002-1140792-5636848
> 
> IMHO This is #1 the best written book on Python I have yet set eyes on.
> Wesley Chun has a deceptively clear prose style, and strikes a clean balance
> between the obvious but essential, yet carefully elucidating famous pythonic
> topics.
> 
> I really thingk this is the most suitable entry text for Python which
> presently exists. It is writen with grace and skill by an author who
> obviously wnows his topic. It lays a great foundation.
> 
> I truly look forwards to any more volumes he may produce in this series.
> 
> The only shame is not the author's fault - it is the still crazy high price
> $45 and thick paper which bulks the 760pages to 2.5" thick.
> I wish american publishers would stop this telphone book web publishing
> madness & follow the example set by Korean technical books I have seen -
> thin paper, larger comfortable format, plenty of color and excellent step by
> step illustrations.
> 
> - Jason
> ___________________________________________________________
> Jason CUNLIFFE = NOMADICS['Interactive Art and Technology']


From andrer at stud.cs.uit.no  Mon Dec 11 09:27:20 2000
From: andrer at stud.cs.uit.no (=?iso-8859-1?Q?Andr=E9?= Larsen Risnes)
Date: Mon, 11 Dec 2000 15:27:20 +0100
Subject: Multiple Threads & IRC
References: <3A3415E0.1080607@bellsouth.net>
Message-ID: <3A34E448.8B42C1BC@stud.cs.uit.no>

Take a lok at:

http://www.cs.uit.no/~andrer/rfc1459.py

RFC 1459 is the document that describes the IRC communication
protocol (which, btw, not all irc servers comply to :-( ).
I've used rfc1459.py to implement a few irc robots.


-Andr? Risnes


(RFC available @ http://www.ietf.org/rfc/rfc1459.txt?number=1459)


From Doug.Hellmann at ZapMedia.com  Sat Dec 16 10:08:37 2000
From: Doug.Hellmann at ZapMedia.com (Doug Hellmann)
Date: Sat, 16 Dec 2000 15:08:37 GMT
Subject: DCOracle Memory Leak?
References: <m3g0k1qih5.fsf@h0050047c4794.ne.mediaone.net>
Message-ID: <3A3B8599.984D15F3@ZapMedia.com>

The only memory leak I've seen associated with DCOracle was using LOBs. 
The handles weren't properly freed because of a decref bug.  This has
been fixed in the latest version though, so if you're running the newest
code that wouldn't be your problem.

Doug

Brad Howes wrote:
> 
> Has anyone seen a Python process using DCOracle grow and grow with each fetch
> from an Oracle database? I can write a really simple loop that does a fetch but
> does not save the value, and the Python process keeps growing. Are there known
> bugs with Oracle 8i client libraries for Linux?
> 
> --
>     "Were people this stupid before TV?" -- _White Noise_ by Don DeLillo


From aleaxit at yahoo.com  Mon Dec 18 15:09:21 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 18 Dec 2000 21:09:21 +0100
Subject: float to string conversion
References: <91lmt3$4b8$1@nnrp1.deja.com>
Message-ID: <91lr82022p6@news2.newsguy.com>

<deragon at 8d.com> wrote in message news:91lmt3$4b8$1 at nnrp1.deja.com...
> Greetings.
>
>
>   I am embarrassed to ask this since it is trivial, but
>   I cannot figure out how to convert a float to a string (obviously,
>   I am new to python).  Somehow, I could not find in the 2.0
>   documentation the syntax to perform the operation.

To convert 'anything' to string, you can apply the str() built-in
function:


>   The following in my function:
>
>     return MSISDN + " check FAILED:  $" + balance + errormsg

return MSISDN + " check FAILED:  $" + str(balance) + errormsg


But you also have alternatives!  The % operator, with a string
as its left operand, does a 'format' -- using format-characters
modeled on C (sorry, I don't know where to find documentation
of what Python does that doesn't depend on docs for C's own
printf!).

return "%s check FAILED: $%s %s" % (
    MSISDN, balance, errormsg)

this is one equivalent alternative to the above, as, at each
place, it's using the '%s' format, which indicates transforming
to string.  But you can make it better, since you can use a
more specialized '%f' fixed-point format and specify two
decimals...:

return "%s check FAILED: $%.2f %s" % (
    MSISDN, balance, errormsg)

this may give, e.g., '$7.00', which is more likely to be the
format you desire than 'whatever string happens'.


Alex





From aleaxit at yahoo.com  Fri Dec 15 08:13:20 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 15 Dec 2000 14:13:20 +0100
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <r0pzoi11lhz.fsf@black.cs.washington.edu> <5jvZ5.3588$Kd1.418770@newsb.telia.net> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com> <kuu28971kp.fsf@lasipalatsi.fi> <3A380E4E.8537A0FB@san.rr.com> <91a1vr02b02@news2.newsguy.com> <Co9_5.25465$x6.13756579@news2.rdc2.tx.home.com> <91bbq102mtk@news1.newsguy.com> <rWd_5.26820$x6.14119694@news2.rdc2.tx.home.com>
Message-ID: <91d5el0me7@news1.newsguy.com>

"Rainer Deyke" <root at rainerdeyke.com> wrote in message
news:rWd_5.26820$x6.14119694 at news2.rdc2.tx.home.com...
    [snip]
> > > class MutableInt(UserInt.UserInt):
> > >   def set(self, value):
> > >     self.data = value
> >
> > UserInt is not in the Python 2 docs, though it's nice to hear it
> exists:-).
>
> Actually it doesn't, but it's trivial (if tedious) to implement.

But, if UserInt is not a funky metaclass, how will MutableInt live
up to its name?  Surely it should at least redefine all of the
in-place methods...:

    x = y = MutableInt(23)
    x += 1

I would expect y's value-as-integer to be 24 now.  So MutableInt
should implement __iadd__ (which the Python 2 docs only mention
in passing in one place, while talking of sequence and mapping
types, but that one _is_ a doc booboo, right?).


Alex





From donn at u.washington.edu  Thu Dec 14 13:02:09 2000
From: donn at u.washington.edu (Donn Cave)
Date: 14 Dec 2000 18:02:09 GMT
Subject: How to use mxTextTools
References: <2bc4OuYTMVp87e49nlcz01wv0ybq@4ax.com>
Message-ID: <91b1v1$bi0s$1@nntp6.u.washington.edu>

Quoth Paul Moore <paul.moore at uk.origin-it.com>:

| I'm looking at mxTextTools to see if it would be suitable for some
| types of text parsing work I am interested in (nothing concrete yet,
| so I can't give specifics...)
|
| The example in the documentation of tagging HTML looks fine - I
| understand what's going on there, and as I understand it, this will
| give me back a taglist, which is (effectively) the text stream with
| portions tagged as I ask.
|
| What I dont't see (yet), and I can't find any good examples for, is
| what to do with the resulting taglist. There seem to be no functions
| for working with taglists, and the lists themselves seem like
| relatively complex data structures, so is it right that I should be
| manipulating them "by hand"?
|
| More information, or better still, some complete examples, would be
| very helpful. (All the examples in the distribution just use
| print_tags() to display the tags, and don't do anything with them...)

Yes, I think that's right, you get to manipulate the tag lists by
hand.  That is, the TextTools.tag() function hands you the result
of analysis according to your table structure, and you make of it
what you will.  That will naturally depend on your application,
that's why it's up to you.

It's a convenient division of responsibility for me, but the Python
world also has some more classical parsers, where the parser calls
application functions as it analyzes the input.  I have tried to use
kwParsing, for example.  If that's more like what you're looking for.

	Donn Cave, donn at u.washington.edu


From sholden at holdenweb.com  Wed Dec 20 08:39:58 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 20 Dec 2000 08:39:58 -0500
Subject: NewBie question: cannot import TKinter and os
References: <90m7fc$r6m$1@nnrp1.deja.com> <p6qitouk6ow.fsf@informatik.hu-berlin.de> <3A368893.FDC9CAC7@boi.hp.com>
Message-ID: <GG206.19$cv5.2974@e420r-atl1.usenetserver.com>

Bill Anderson <anderson at boi.hp.com> wrote in message
news:3A368893.FDC9CAC7 at boi.hp.com...
> Martin von Loewis wrote:
> >
> > etsang at my-deja.com writes:
> >
> > > 1. at the very first beginning, it says import site error. I know this
> > > module site is no longer required in Python 2.0 after dereading some
> > > doc in pytho.org. how can I make thie error go away??
> >
> > That is incorrect. Where did you read that? site.py is still needed.
>
> Perhaps from the site.py included in 2.0?
> """
> Append module search paths for third-party packages to sys.path.
>
> ****************************************************************
> * This module is automatically imported during initialization. *
> ****************************************************************
>
> In earlier versions of Python (up to 1.5a3), scripts or modules that
> needed to use site-specific modules would place ``import site''
> somewhere near the top of their code.  Because of the automatic
> import, this is no longer necessary (but code that does it still
>         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> works).
>
> """
> (emphasis mine)
>
> I, too am having problems importing site.py and os, at least Zope is
> anyway. I can do it just fine from the interpeter... it seems that not
> allcode that does it still works. :(
>
> --
> Bill Anderson               Linux Specialist
> Modular Network Storage     R&D
> Random Quote:
>     Portable: survives system reboot.
>
"This is no longer necessary" in the code refers to the explicit

    import site

statement which used to be required in Python programs (I understand).
Since 1.5 final the import of this module has been automatic.  Site
customizations should be specified in ...

sitecustomize.py

imported, if importable, automatically by site.py

regards
 Steve





From neelk at alum.mit.edu  Sun Dec 24 20:06:48 2000
From: neelk at alum.mit.edu (Neelakantan Krishnaswami)
Date: 25 Dec 2000 01:06:48 GMT
Subject: Why is Python slow? (was Re: One Python 2.1 idea)
References: <mailman.977590932.19335.python-list@python.org> <9237md$orm$1@nnrp1.deja.com> <924bjt0e3n@news1.newsguy.com> <m38zp6nk8k.fsf@localhost.localdomain> <925mpe$jjg$1@panix6.panix.com>
Message-ID: <slrn94db6a.rc.neelk@alum.mit.edu>

On 24 Dec 2000 12:37:02 -0800, Aahz Maruch <aahz at panix.com> wrote:
>In article <m38zp6nk8k.fsf at localhost.localdomain>,
>Lieven Marchand  <mal at bewoner.dma.be> wrote:
>>
>> I think Common Lisp shows that getting very good speed in Python is
>> quite feasable. The original CMUCL implementers weren't such a large
>> group. Their highly optimising compiler (which incidentally is also
>> called Python) has on occasion beaten FORTRAN at numerics. I don't
>> know why some people in the Python community think compiling Python is
>> such a problem. Practically all the problems have been tackled and
>> solved 20 years ago in the Lisp community.
>
> One of the reasons that gets brought up less often than it should is
> that Guido is somewhat fanatical about also keeping the CPython
> implementation clean and simple.  I'm pretty sure that -- Ghu forbid it
> become necessary -- I could manage to maintain CPython if I had to.
>
> This leaves less room for optimization than many people think.

Perhaps.

I started to take a look at the Ocaml 3.00 bytecode interpreter the
other day, and it is quite clean and readable. It is also right around
10-20 times faster than the Python interpreter, measuring just the
basic busy-loop. The only filips in the Ocaml interpreter that jumped
out at me were a) some macrology to enable the gcc computed labels
extensions when it was available, and b) it uses tagging to avoid
allocating on the heap when handling integers.

The Python code was:

def loop(n):
    t = time.time()
    while n != 0:
        n = n - 1
    return time.time() - t

def loop2(n):
    t = time.time()
    for i in xrange(n):
        pass
    return time.time() - t

When passed n=10000000, the median runtime for loop1 was 7.05 seconds,
and for loop2 was 3.89 seconds.

Some equivalent code in Ocaml:

let rec loop n = if n=0 then () else loop(n-1);;

let time n = let t = Sys.time()
              in loop n;
                 Sys.time() -. t;;

With n=10000000, this ran in a median of 0.4 seconds. This is the
performance of the interpreter, mind -- Ocaml also has a compiler
which I didn't exercise. I think this suggests there is room for
*huge* performance improvements in interpreted Python.

I recall running into a webpage where Vladimir Marangozov had a patch
that used gcc's computed labels to do indirect threading in Python.
Anyone know what the performance improvement from that was?


Neel


From sbrunning at bigfoot.com  Tue Dec 12 12:04:54 2000
From: sbrunning at bigfoot.com (Simon B.)
Date: Tue, 12 Dec 2000 17:04:54 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu>
Message-ID: <915lrg$lp9$1@nnrp1.deja.com>

> From:	Roy Katz [SMTP:katz at Glue.umd.edu]
> Okay! tell me, which are the objects which cannot be modified in
place? I
> assume these are int's, long's, float's, etc.  Strings, lists,
objects and
> dictionaries are call-by-reference.  Pardon my choice of diction, but
I am
> against inventing new terminiologies (modified-in-place??).

The correct terminology is mutable (for objects which can be modified in
place), and immutable (for objects which cannot). See
<http://www.python.org/doc/current/ref/objects.html>.

> I want to know where this is documented.  Specifically, where is the
> dichotomy of ints/longs/floats vs.
objects/strings/lists/dictionaries.

The built-in types are documented here -
<http://www.python.org/doc/current/ref/types.html> - including whether
or
not they are mutable.

> While I'm brainstorming with Python grammar changes, what about a
real &
> operator.  No more of this hokey is-it-modifiable-in-place-or-not
> business.  Explicit control is important.

Ugh.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk


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


From nickatvideosystemdotcodotuk  Thu Dec 14 12:27:40 2000
From: nickatvideosystemdotcodotuk (Nick Trout)
Date: Thu, 14 Dec 2000 17:27:40 -0000
Subject: what is a closure?
Message-ID: <3a39033b@news.xtml.co.uk>

Sorry, this is probably not the place to ask this but you lot are pretty
friendly :-)

I have seen the term "closure" used... what is it exactly? Is it just a
class member function?

Ta,
N






From steinarrune at my-deja.com  Mon Dec 18 09:24:32 2000
From: steinarrune at my-deja.com (steinarrune at my-deja.com)
Date: Mon, 18 Dec 2000 14:24:32 GMT
Subject: Stored Procedures (ODBC) from Python?
Message-ID: <91l6mq$lge$1@nnrp1.deja.com>

I am developing in Pythonwin, and have been using
the odbc module successfully for a while. Still I
have not yet called Stored procedures over this
interface.

Is it possible ? Any examples out there ?

All the best

Steinar Eriksen


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


From ivnowa at hvision.nl  Mon Dec 11 10:00:11 2000
From: ivnowa at hvision.nl (Hans Nowak)
Date: Mon, 11 Dec 2000 15:00:11 GMT
Subject: Pascal -> Python 'record' structures
References: <2y4Z5.7634$KY1.20696@news1.rivrw1.nsw.optushome.com.au>
Message-ID: <3A34EA80.3CB4@hvision.nl>

deadmeat wrote:
> 
> Can anyone tell me how to read in file records that were created by Pascal
> and C programs? Will I have to write a wrapper and parse the raw data
> manually?

That's a way to do it, but the struct module might do what you want:

  http://www.python.org/doc/current/lib/module-struct.html

--Hans Nowak


From  at home.com  Thu Dec 21 19:30:54 2000
From:  at home.com (John Q. Public)
Date: Fri, 22 Dec 2000 00:30:54 GMT
Subject: Overriding sys.stdout
References: <3A417FA8.29EBC549@home.com> <3A418329.D1871B68@uswest.net>
Message-ID: <3A42A0BD.EB8BBE6D@home.com>

Thank you for the help.  I can finally remove my hack from IDLE!

Jonathan Polley

jwpolley(a)collins(d)rockwell(d)com

Bjorn Pettersen wrote:

> Something like this should work...
>
> -- bjorn
>
> import StringIO
> import sys
>
> class Logger:
>  def __init__(self):
>   self.buffer = StringIO.StringIO()
>   self.save_stdout = sys.stdout
>  def close(self):
>   sys.stdout = self.save_stdout
>  def getvalue(self):
>   return self.buffer.getvalue()
>  def write(self,txt):
>   self.buffer.write(txt)
>   self.save_stdout.write(txt)
>
> >>> import log
> >>> import sys
> >>> l = log.Logger()
> >>> sys.stdout = l
> >>> print 5,6,7
> 5 6 7
> >>> print "hello %s" % "world"
> hello world
> >>> l.getvalue()
> '5 6 7\012hello world\012'
> >>> l.close()
>
> "John Q. Public" wrote:
>
> > I would like to be able to take a snapshot of stdout while a script is
> > running, but still allow stdout to be written to the terminal.  My
> > current, working, solution is to hack IDLE but I would like a generic
> > solution so we would not be bound to a specific IDE.  I would like to
> > use PythonWin on NT, if I can, or do the journaling in batch mode.
> >
> > My plan is to implement a class/module/whatever that I can used to
> > overwrite sys.stdout.  This class/module/whatever would write data to a
> > file and send it to the old stdout.  My mail questions are these:
> >
> >     Does such a class/module/whatever exist?
> >     If not, does anyone have an idea where I can start for info?  As a
> > last resort, I will start wading through IDLE and PythonWin, but I would
> > like some ideas first.
> >
> > Thanks!
> >
> > Jonathan Polley
> > jwpolley(a)collins(d)rockwell(d)com



From dsh8290 at rit.edu  Sat Dec 30 23:53:33 2000
From: dsh8290 at rit.edu (D-Man)
Date: Sat, 30 Dec 2000 23:53:33 -0500
Subject: identifying user from Win9x?
In-Reply-To: <92ibvf$if2$1@nntp6.u.washington.edu>; from fpm@u.washington.edu on Fri, Dec 29, 2000 at 03:52:15PM +0000
References: <92ibvf$if2$1@nntp6.u.washington.edu>
Message-ID: <20001230235333.D19500@westsidecnc.rh.rit.edu>

I don't know anything about Novell, but MS Windows doesn't have a
concept of users.  Everyone is "root" (so to speak).  Win9[58] have
"users" in the sense of having a different Desktop and Start Menu, but
that is all they do.  If you are using NT it is a different story.  NT
actually does have different users with varying permissions.  I don't
know how to extract the username, but it does exist to be extracted.

HTH,
-D



From ngps at madcap.dyndns.org  Fri Dec 22 09:34:05 2000
From: ngps at madcap.dyndns.org (Ng Pheng Siong)
Date: 22 Dec 2000 14:34:05 GMT
Subject: How to make a GUID with plain python
References: <3a413a17.3416402@192.168.0.1>
Message-ID: <91voot$cs5$1@violet.singnet.com.sg>

According to Georg Gogo. BERNHARD <gogo at bluedynamics.com>:
> I am looking for a way to create a global unique identifier / UUID
> with python...

Here's an expired I-D describing "the UUID formats and various ways
of generating them".

    http://www.shore.net/~rsalz/uuid-i-d.txt


Happy hols!


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



From jerry_spicklemire at my-deja.com  Wed Dec 27 11:28:35 2000
From: jerry_spicklemire at my-deja.com (jerry_spicklemire at my-deja.com)
Date: Wed, 27 Dec 2000 16:28:35 GMT
Subject: Python vs PHP (& Java?)
References: <mailman.977897649.4877.python-list@python.org> <92d22f$cnv$1@nnrp1.deja.com>
Message-ID: <92d5bc$ffm$1@nnrp1.deja.com>

a better link for Quixote:
 http://www.mems-exchange.org/software/python/quixote/


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


From shindich at my-deja.com  Fri Dec  8 20:42:15 2000
From: shindich at my-deja.com (shindich at my-deja.com)
Date: Sat, 09 Dec 2000 01:42:15 GMT
Subject: Python wrapper for C++ STL?
References: <3A31760E.D157EB54@erols.com> <npfY5.1850$x6.2577114@news2.rdc2.tx.home.com>
Message-ID: <90s2lm$ig1$1@nnrp1.deja.com>

In article <npfY5.1850$x6.2577114 at news2.rdc2.tx.home.com>,
  "Rainer Deyke" <root at rainerdeyke.com> wrote:
> "Edward C. Jones" <edcjones at erols.com> wrote in message
> news:3A31760E.D157EB54 at erols.com...
> > Has anyone written a Python wrapper for the C++ Standard Template
> > Library?
>
> Why would anybody want to?  Python provides most STL fnctionality
natively.
> Such a wrapper would need to constantly translate C++ exceptions to
Python
> exception and vice versa, making the wrapper be more trouble than it's
> worth.
>
> --
> Rainer Deyke (root at rainerdeyke.com)
> Shareware computer games           -           http://rainerdeyke.com
> "In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor
>
>
STL has other things that are of great use to C++ developers -- sets
and a bunch of generic algorithms to name a few. But STL looses its
beauty without the type safety. And Python is not type safe. I guess
the only feature that I really miss is std::set. But a set can be
implemented as a dictionary that uses the same object as both the key
and the value.


Sent via Deja.com http://www.deja.com/
Before you buy.


From jurjanm at c.ict.om.org  Fri Dec  8 13:54:03 2000
From: jurjanm at c.ict.om.org (Jurjan-Paul Medema)
Date: Fri, 8 Dec 2000 18:54:03 -0000
Subject: Namespace model
In-Reply-To: <976202030.38510@egroups.com>
Message-ID: <200012081854.SAA03180@c.ict.om.org>

Hello all,

I like many features of Python's unique language design a lot 
(since I first started studying and using it last summer). But what 
bothers me, is the whole Namespace model that is so unlike a 
traditional scope model. Fact is that I am not happy with it. I am 
sure there has been a lot of discussion about this during the last 
10 years, and therefore I hope that somebody can clarify it to me.

Especially in functional programming it is common place to nest 
functions. The FAQ answers the question how a nested function 
can access a parameter to the parent function by having the same 
parameter repeated as a default parameter (FAQ4.5). This I find at 
least more acceptable than the alternative that requires the 
introduction of a completely new object class, because to me that 
is too remote from the original idea. However this becomes the only 
possible way of getting around things when you want the nested 
function to be recursive (again: this is common place in functional 
programming). I have already encountered a situation where I 
needed to pass a function as a parameter to a standard Python 
Module, where I needed recursion with more than one parametere, 
but also had to comply with a set function signature. It took quite 
some redesign to come up with a working OO-model, that turned 
out to be bigger than I thought should be necessary. 

I am not going to bother you with that example, but this is an 
abstract version:

# 'operation_A', 'operation_B', 'operation_C' and 'operation_D'
# can be anything in this example.
# (At least) 'operation_C' returns a pair of values.

def foo(x) :

	def bar(y, i) :
		if y : return operation_A(y, i)
# the following 'bar' is not found in the namespace.
           else : return operation_B( bar( operation_C(y, i) ) )  

	z = bar(x, 0)
	return operation_D (x, z)

Here Python simply lacks the intuitivity that it promises. I like OO, 
but I also like to keep functions together that really belong together 
(on a different level than a class level obviously).

I would like to ask Guido, but if anybody has a good idea, please 
explain to me why Python was designed with the Namespace 
model instead of a normal scope model!!
It can't be that it's jsut because this keeps the interpreter/compiler 
simpler or is that all that it is?? Can we expect this to be resolved 
in future versions of Python or is this how it always will be...?

Seriously, I don't think a language that claims to be designed for 
educational purposes should contain something this inconsistent: 
you either support function nesting or you don't, but this is just half-
hearted.
Please clarify, somebody!
 
And yes, I honestly do like Python a lot, so I hope I do not come 
across as grumbling too much... 

Thanks,
Jurjan-Paul Medema



From neilh at scintilla.org  Thu Dec 28 17:40:53 2000
From: neilh at scintilla.org (Neil Hodgson)
Date: Thu, 28 Dec 2000 22:40:53 GMT
Subject: 8.3 peculiarity if run w/i console
References: <92g4dn$5ug$1@newsreaderm1.core.theplanet.net>
Message-ID: <VjP26.36442$xW4.287502@news-server.bigpond.net.au>

Franz GEIGER:

> If I run a script within the Python 2.0 console, the names of the files
that
> are made by that script are all "abbreviated", e.g.:
> ALongFileName.txt becomes ALongF~1.TXT. If I run the script within
> PythonWin, the names are o.k.

   I get the same behaviour from console and PythonWin on W2K with:
f = open("ALongFileName.txt", "w")
f.close()

   What is it that is saying the name is abbreviated? On Windows, all file
names are stored in two forms - a long form and an abbreviated form. The
abbreviated form is for use by old software which can't cope with the long
names. If you do:
dir /x
   then you will see both forms.

   Neil




From joconnor at cybermesa.com  Tue Dec 12 13:58:53 2000
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Tue, 12 Dec 2000 18:58:53 +0000
Subject: Smalltalk and Python
Message-ID: <3A36756D.EDF969EF@cybermesa.com>

> >So I was wondering if any of you Smalltalkers had tried Python and could
> >compare the experience to working with Smalltalk .
>
> I think of Python as a very advanced scripting language. Much cleaner
> and more regular than Perl, not as powerful as Smalltalk. It is a great
> language and should appeal to many Smalltalkers for smaller tasks.

Complete agreement.  I'm a hardcore Smalltalk developer and use Python a
lot for those small scripting projects and other situations where
Smalltalk is not the right choice

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

"God himself plays the bass strings first when He tunes the soul"



From - at -.-  Thu Dec  7 13:03:21 2000
From: - at -.- (- at -.-)
Date: 7 Dec 2000 18:03:21 GMT
Subject: MANY EASY, GREAT-PAYING JOBS AVAILABLE!!
Message-ID: <90ojd9$6e7$190@pluto.ucc.uno.edu>

HOME WORKERS NEEDED!

Earn up to $500 easy (or more) per week, performing simple assembly or
clerical work in your home.

Dear Future Home-Worker:

Would you like to have a higher income each week? an extra $500 or more? Are
you tired of trying to make ends meet on your present income? Do you find
yourself envying your friends and neighbors because of things they have,
like new cars, TV's, home entertainment centers, or all the right clothes
that you just can't seem to afford? Many people just like you, for one
reason or another, are choosing to take advantage of all their spare time to
make more money -- EXTRA money. With the economy the way it is today, a
family cannot make it on one paycheck a week. Whether you are a mother
needing to stay home with your children, or a single man or woman needing a
bigger and better income, we all have one thing in common:
WE ALL NEED MORE MONEY!

We'll Show You How to Receive that Extra Income

NOW YOU can make money, by working for companies from the privacy and
comfort of your own home, working at your own pace, at that! Manufacturing
and Distributing companies supply you with the materials and easy-to-follow
instructions, and pay you for the completed works. They work across Canada
and the United States. Everything is done through the mail, and these
companies will even pay shipping and postage!

These jobs are real and honest - you can earn an honest day's wage for an
honest day's work! And it's so easy to get! You will be hired in your own
home through the mail. NO RESUMES, EXPERIENCE, SELLING, OR RUNNING FOR
INTERVIEWS. Work for several companies at one time too! Let the whole family
pitch in. Be your own boss and choose your own hours. You must be honest and
trustworthy.

Wouldn't it be nice if you could find a job that let you work from your own
home, in your spare time? Where there's no need to rearrange yoiur present
schedule and miss those pleasures in life you so much enjoy?  And best of
all... you'll receive steady paychecks each week. Your time is worth much
more than just $7.00 an hour, and your family life need not suffer with you
being away from home. With no boss looking over your shoulder, you'll feel
like you're being paid for doing a hobby. While doing the kind of work you
enjoy, you can watch your favorite TV shows and still make money. The whole
family can get involved for even greater income.

YOU CHOOSE the Job you enjoy

We have all had hobbies in the past we wish we coiuld have been paid for.
Now there's a way we can earn money each week by doing these hobbies. After
extensive research, we have developed the Home Employment Directory - a
comprehensive guide that lists over a hundred companies needing your help,
and pay TOP DOLLARS. Many companies actually train you to work for them.

Here are just a few examples (there are MANY more):

Arts & Crafts
Computer Work
Data Entry
Hand Painting
Mail Processing
Newspaper Clipping (very easy and quickly profitable)
Recording Video Tapes
Sewing & Needlework
Survey Work
Typing

Assembly & Manufacturing of:
Beaded Earrings, Cloth Flowers, Jewelry, Plant hangers, Stuffed Animals,
Tote Bags, Toys & Gifts, and Wooden Crafts


NO SELLING REQUIRED

The companies offering home assembly do their own selling. All they want you
to do is follow their easy instructions, and they'll pay you for the
completed products. You simply assemble their products or complete their
home-work projects according to their instructions, and you'll get paid upon
completion and return of the finished goods to the company. They will do the
rest!

Since people have a wider range of interest, the Home Employment directory
also contains a large variety of home-business and home-sales opportunities,
as well as regular work. Positions are identified by category. If you don't
want a position that requires any selling, simply ignore it and go on to
those you are interested in. There are MANY to choose from!


Below are just a few very brief descriptions of the type of assembly or
clerical work available. There's an even greater variety of opportunities
available, too many to list an example of each:


- One agency offers typing, hand-addressing, and other easy clerical work
    you can do from home to earn up to $500 a week!

- Earn up to $495 per week assembling beautiful and distinctive earrings for
    a company. They will provide easy and detailed instructions
    - No selling involved!

- Sew beautiful golf club covers for a growing company and earn up to
    $421.50 weekly.

- Make holiday gifts and decorations year round for a company that will pay
    up to $450 weekly. Assemble oven mitts, pillows, ornaments, etc.

- One company will get you started in earning $625 or more per week by
    processing their mail!

- Assemble clown dolls for a company that will pay you $750 for your
    completed work!

- All you need is a sewing machine to earn up to $350 a week by sewing nylon
    sports bags!

About Our Directory:

The Home Employment Directory includes the name and address of each company,
with a full description of the assembly or clerical work offered. Our
listings include a large assortment of companies and exciting home-work
opportunities, the abve is just a very small sample of what's available. Our
directory is a comprehensive, up-to-date publication containing infomation
on the most profitable opportunities available. Most of the assembly
opportunities require no special training, skills, or equipment - just a
sincere desire to get started. Others require basic sewing, typing, writing,
computer skills, etc. Many companies will train you to work with them. They
will provide you with instructions that are easy to understand, and projects
that can be mastered quickly. All the opportunities are easy to get started
in, and can be done either part-time or full-time for an excellent income!

Now, It's All up to YOU!

These companies NEED YOU! They want you to be able to earn a large income
from home, and pay you handsomely for your efforts. What more could you ask
for? The only way on earth that these jobs will work for you is... if you do
it! For your sake and your family's - DON'T PASS UP THIS GOLDEN OPPORTUNITY!
Some day soon you'll probably look back andconsider this to be the best
investment you've ever made. It is our sincere hope that you'll respond
today and get started making money from your own home NOW!

Similar guides are sold by others for $40 or $50 or more.. sometimes even as
much as $100 or $200... but our Home Employment Directory can be yours for
only $20. That includes shipping and handling charges for first class mail.
(Add $5 for a Priority Mail rush, or $15 for an Express Mail rush)

Don't Delay - Don't let someone else take YOUR work-at-home job away from you!

                  PRINT THIS OUT AND MAIL IT IN WITH PAYMENT                  
------------------------------------------------------------------------------
                     Home Employment Directory Order Form


I'm ready to start making money from home now!
Please send me a copy of the Home Employment Directory to this address:


   NAME ____________________________________________________________________

Address ____________________________________________________________________

   City _______________________ State/Prov _________ Postal Code____________

(check one below)

___ Send it to me via FIRST CLASS MAIL. (usually 2-3 days, sometimes 5-7 days)

___ RUSH it to me via PRIORITY MAIL. (usually 2-3 days, rarely 4-5 days)
    I'VE ENCLOSED AN ADDITIONAL $5 FOR THE PRIORITY MAIL RUSH CHARGE.

___ RUSH it to me via Express Mail. (most places overnight, sometimes 2 days)
    I'VE ENCLOSED AN ADDITIONAL $15 FOR THE EXPRESS MAIL RUSH CHARGE.

NOTE TO CANADIANS -- Use the First Class Mail option, unless you have an
address within the states where you can receive mail. International Express
and Global Priority seem to have a way of sitting in Customs warehouse for
anywhere from a week to a month. Also, be sure to use an international postal
money order, made out in US Dollars. Foreign checks will NOT be accepted
because of the additional $15 charge for processing them.

NOTE TO EVERYONE
-- Send a money order for fastest delivery. They will be shipped out by the
        very next day.
-- ALL Check orders will be held for 2 weeks, regardless of whether a "Rush"
        option has been specified.
-- If nothing has been checked above for an order, it will be shipped out
        First Class Mail.

MAKE CHECKS OR MONEY ORDERS PAYABLE TO:  Brent Cornwell

Include on the bottom of the check (in the memo field), or on the line below
the address on the money order, the following text: 

        Home Employment Directory Order

Mail the payment and this orderform to:

Brent Cornwell
GHT Services - Home-Based Employment
PO Box 68
Hahnville, LA 70057


From andrew.spambgone.chatham at duke.edu  Mon Dec 11 21:43:52 2000
From: andrew.spambgone.chatham at duke.edu (Andrew H. Chatham)
Date: 12 Dec 2000 02:43:52 GMT
Subject: function pointer
References: <3A358485.FC22A29@bioeng.ucsd.edu>
Message-ID: <slrn93b4bd.8bv.andrew.spambgone.chatham@fizgig.dorm.duke.edu>

You could use a CObject, which is what I've done before, though it has
limitations. A CObject is really just a Python object that holds a C pointer.
See http://python.org/doc/current/api/cObjects.html . Of course, doing it this
way, you'd have to already know what kind of arguments and return values the
function takes and there's also the possibility that someone could pass a bad
pointer if they had a CObject from another module, since you cast it to a "void
*" to store and retrieve it from the CObject. So if you already know what kind
of function it will be a pointer to and don't expect people to go out of their
way to break your module, a CObject would be a good choice.

Andrew

On Mon, 11 Dec 2000 17:51:01 -0800, Curtis Jensen <cjensen at bioeng.ucsd.edu>
wrote:
>Is there a way to pass a pointer to a function from a module in a shared
>library?  e.g. We have a compiled C module that has the function "foo" in it
>(foo is not wrapped, it's straight C code).  We have another copiled C module.
>We want "foo" to be passed in to a function in the second module as a
>paramater.  Is this possible?  Hope this makes sence.
>
>-- Curtis Jensen cjensen at bioeng.ucsd.edu http://www-bioeng.ucsd.edu/~cjensen/
>FAX (425) 740-1451


-- 
Andrew Chatham
Duke University '02
andrew.chatham at duke.edu


From hinsen at cnrs-orleans.fr  Sun Dec 10 12:29:07 2000
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 10 Dec 2000 18:29:07 +0100
Subject: Parallel Computing
References: <3A300B1D.223630F3@bioeng.ucsd.edu> <874s0fpzr2.fsf@lisboa.ifm.uni-kiel.de> <90qgmt$1ko$1@news.nuri.net>
Message-ID: <m3g0jwqj6k.fsf@chinon.cnrs-orleans.fr>

"June Kim" <junaftnoon at nospamplzyahoo.com> writes:

> Is "Python Parallelization Project" still alive? The link to the MPI
> interface seems broken. Is there anyother place where I can get it?
> 
> http://starship.python.net/crew/hinsen/parallel.html

I can only confirm that the link is broken; I don't have any more
recent information. However, you can try the MPI interface in
ScientificPython, which is available at

    ftp://dirac.cnrs-orleans.fr/pub/ScientificPython/

-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.56.24
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------


From thaiwey at my-deja.com  Wed Dec  6 11:30:45 2000
From: thaiwey at my-deja.com (thaiwey at my-deja.com)
Date: Wed, 06 Dec 2000 16:30:45 GMT
Subject: SSL socket in 2.0
Message-ID: <90lpjm$edp$1@nnrp1.deja.com>

hi,

i have been trying to compile the socket module with SSL support on my
nt without much success. i was using the pcbuild.dsw in MSVC 6.0. can
anyone advice me as to how i can set the necessary flags in the project
options/settings to turn on SSL.

i have already compiled openssl-0.9.6 for my machine. i have also added
"USE_SSL" in the preprocessor definitions, paths to openssl in the
include directories, library path to openssl, and "libeay32.lib
ssleay32.lib" to the Object\library modules section. when compiling, i
keep getting the following errors:

Compiling...
socketmodule.c
G:\Python-2.0\Modules\socketmodule.c(2112) : error C2099: initializer
is not a constant
G:\Python-2.0\Modules\socketmodule.c(2114) : warning C4047:
'initializing' : 'int ' differs in levels of indirection from 'char [4]'
G:\Python-2.0\Modules\socketmodule.c(2115) : warning C4047:
'initializing' : 'char *' differs in levels of indirection from
'unsigned int '
G:\Python-2.0\Modules\socketmodule.c(2118) : warning C4047:
'initializing' : 'int ' differs in levels of indirection from 'void
(__cdecl *)(struct _object *)'
G:\Python-2.0\Modules\socketmodule.c(2120) : warning C4047:
'initializing' : 'int (__cdecl *)(struct _object *,struct _iobuf *,int
)' differs in levels of indirection from 'struct _object *(__cdecl *)
(struct _object *,char *)'
Error executing cl.exe.

thanks.

tw


Sent via Deja.com http://www.deja.com/
Before you buy.


From cnelson9 at gte.net  Tue Dec 19 00:54:47 2000
From: cnelson9 at gte.net (Clifford J. Nelson)
Date: Tue, 19 Dec 2000 05:54:47 GMT
Subject: Python in "math through programming" curriculum
References: <2g2t3t06r2agrrjmfnr41eu4lc2u39a8qi@4ax.com>
Message-ID: <3A3EFF47.67DE69CF@gte.net>


Kirby Urner wrote:
> 
> Repost from edu-sig at python.org elist -- one typo/URL fixed.
> 
> =============================
> 
> List-Archive: <http://www.python.org/pipermail/edu-sig/>
> Date: Mon, 18 Dec 2000 01:08:01 -0800
> 
[snip]
> 
> Awhile back, Tim Peters recommended 'Concrete Mathematics'
> (Knuth a co-author) as a good example of a "math through
> programming" text book.  I subsequently bought it, and
> had to agree.  So how do we "lower a ladder" to middle
> and high schoolers starting from there?  I think Python
> is part of the answer.
>

I don't think Python should be a standard at all, because of
my experience with it on the Mac. Keep the options open for
computer programming languages.

Here are two:

Clean at: http://www.cs.kun.nl/~clean/

Caml at:  http://pauillac.inria.fr/caml/

Both of those languages are OOP too, and easier to use than Python.

  Cliff Nelson
[snip]


From junaftnoon at nospamplzyahoo.com  Tue Dec 19 03:51:46 2000
From: junaftnoon at nospamplzyahoo.com (June Kim)
Date: Tue, 19 Dec 2000 17:51:46 +0900
Subject: tkinter event binding to mouse double click
Message-ID: <91n7l6$7ij$1@news.nuri.net>

I'm using Python 2.0 on Win32.
In the code following, the binding doesn't seem to work; it can't detect
mouse double-clicking at all. (Maybe a platform-specific problem, I reckon)


from Tkinter import *

root = Tk()

def callback(event):
    print "clicked at", event.x, event.y

frame = Frame(root, width=100, height=100)
frame.bind("<Double-Button-1>", callback)
frame.pack()

root.mainloop()



From Norman_Shelley-RRDN60 at email.sps.mot.com  Thu Dec 21 18:04:28 2000
From: Norman_Shelley-RRDN60 at email.sps.mot.com (Norman Shelley)
Date: Thu, 21 Dec 2000 16:04:28 -0700
Subject: ?Why is 0e-6 invalid?
Message-ID: <3A428C7C.5CFBB181@email.sps.mot.com>

Why is 0e-6 invalid and 0*.000001 is valid?

They are equivalent and a "normal" mathematically trained individual
sees them for what they are immediately.

(I have a trivial engineering notation to number convertor that takes
"0u" or "5u" and converts them to "0e-6" and "5e-6" and then evaluates
them. The later works fine, the former fails. Granted I could check for
this condition, but it seems that the Python evaluator should be able to
work on this boundary case as it still consistent.)


>>> 0*.000001
0.0
>>> 1e-6
1e-06
>>> 0e-6
  File "<stdin>", line 1
    0e-6
     ^
SyntaxError: invalid syntax




From syver at NOSPAMcyberwatcher.com  Tue Dec 12 21:12:59 2000
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Wed, 13 Dec 2000 03:12:59 +0100
Subject: reloading modules question
References: <90scdf$hpe$1@troll.powertech.no> <ouk43t494isd9m8fk92udjjbg2pb6crq7q@4ax.com> <910om3$2vk$1@troll.powertech.no> <kg9a3t0s6ah2c7t3pf94o3vl4ls2hespuc@4ax.com>
Message-ID: <916m48$648$1@troll.powertech.no>

<dale>
> Starting and stopping the web site has always worked for us. (Windows 2000
Server)
</dale>

You mean something like net stop w3svc? Because just stopping the default
web site in inetmgr doesn't work for me on Win 2000 server. Unloading the
application works however.





From fredrik at effbot.org  Tue Dec 12 12:56:17 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 12 Dec 2000 17:56:17 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl>
Message-ID: <5FtZ5.3570$Kd1.415190@newsb.telia.net>

Remco Gerlich wrote:
> There is no real dichotomy here. Some types have methods that can change
> the type, other types don't.

careful: you meant "can change the contents of objects of that type"

(but you already know that, of course).

</F>




From kuntsi at cc.helsinki.fi.spam.no  Wed Dec 27 23:11:17 2000
From: kuntsi at cc.helsinki.fi.spam.no (Antti Kuntsi)
Date: 28 Dec 2000 04:11:17 GMT
Subject: sniffing btw my web browser and a server
References: <92aush$rhc$1@news.nuri.net>
Message-ID: <slrn94lfb4.gj1.kuntsi@myntti.helsinki.fi>

June Kim <junaftnoon at nospamplzyahoo.com> wrote:
> I'm doing some sort of web client programming, and have to observe the
> messages that my browser and web-servers transmit, being tired of guessing
> what the messages'd be and looking up the references and the specifications.
> Anyone has done it in Python or knows how it could be done?

Writing a simple HTTP-proxy in Python that dumps all the data (or just the
HTTP-headers for both requests and replies, if that is enough) is quite easy:

-create a INET-socket, and start accepting connections.
-When you get a connection, read the request, and print the data to a file
 or stdout
-Open a socket to the server (you get this from the first line of the request,
 as the request made by browser to your HTTP-proxy tells you this)
-write the request to the server
-read the response from the server, dump/print the data, and send it to the
 browser.
-go back to accepting connections.

To make it fancy, you can make it multiplexed or threaded, but it's not
absolutely necessary.

Do note, that you have to
1) change the first line of the request when you make the req to the server
   i.e. from HTTP-proxy-request to HTTP-server-request
or
2) open the connection to a real HTTP-proxy instead of the server, and let it
   take care of figuring out the web-server's address and port.

.antti
-- 
         /"\                           |    iki.
         \ /     ASCII Ribbon Campaign |    fi/
          X      Against HTML Mail     |    mickut
         / \


From garryknight at bigfoot.com  Sun Dec 24 07:11:25 2000
From: garryknight at bigfoot.com (Garry Knight)
Date: Sun, 24 Dec 2000 12:11:25 +0000
Subject: Some Python 2.1 ideas
References: <mailman.977590932.19335.python-list@python.org> <Cw616.23054$qN6.3308205@typhoon.we.rr.com> <3A452881.7686E717@uswest.net> <%vb16.187233$U46.5801214@news1.sttls1.wa.home.com>
Message-ID: <924p5d$9bi$1@taliesin.netcom.net.uk>

In article <%vb16.187233$U46.5801214 at news1.sttls1.wa.home.com> "Greg
Jorgensen" <gregj at pobox.com> wrote:

> How should these be handled?
> 
>   "foo\n\n\n"

chomp  => "foo\n\n"
chomps => "foo"

-- 
Garry Knight
garryknight at bigfoot.com


From moshez at zadka.site.co.il  Tue Dec 26 23:11:13 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Wed, 27 Dec 2000 06:11:13 +0200 (IST)
Subject: ~Python ?
In-Reply-To: <20001226181033.E25820@xs4all.nl>
References: <20001226181033.E25820@xs4all.nl>, <3a4833f8.346848@news.t-online.de> <slrn94hd3j.1br.jp@localhost.localdomain>
Message-ID: <20001227041113.E6158A84F@darjeeling.zadka.site.co.il>

On Tue, 26 Dec 2000, Thomas Wouters <thomas at xs4all.net> wrote:

> or rm ./~
> or rm '~'
> or rm "~"
> or rm /full/path/to/~
> 
> There's more than one way to do it :)

To bring this back on topic from the depths you dragged us to <wink>

% python
>>> import os
>>> os.unlink("~")

-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From ollir at linuxfreak.com  Mon Dec 25 03:43:45 2000
From: ollir at linuxfreak.com (Olli Rajala)
Date: Mon, 25 Dec 2000 08:43:45 GMT
Subject: A Suggestion for Python Colon Syntax
References: <mailman.977446446.16022.python-list@python.org> <9207a3$av6$1@hecate.umd.edu>
Message-ID: <3a45eb13.10748563@uutiset.nic.fi>

On 22 Dec 2000 18:42:11 GMT, William Djaja Tjokroaminata
<billtj at y.glue.umd.edu> wrote:
>Exactly.  I just regretted the initial study that used newbies and they
>preferred colons.  At the present time, correct me if I am wrong, I think
>it is still rare that Python is one's first programming language.  People
>who start to use Python may already use some other popular programming
>languages.  

I've tried many languages before Python: C/C++, Java, Perl, Visual
Basic (Some people says, that it's not programming language, and maybe
they're right. =), Delphi, Pascal and Php.  But I didn't learned them
further, because I didn't have time and I didn't feel that they're my
thing.  Perl and Php made an exception.  I like Perl, and it's so
handy in some tasks, but Perl is Perl. =)  Php is very handy when
static webpages are not enough for you.  

Then someday I found "How To Become A Hacker FAQ" by Eric S. Raymond
and it says that you should first learn Python.  And now I'm stuck
with it.  I think that someday when I'll be the real hacker I can say
that Python was the first language that I learned.

I hope you understand what I tried to say, but I can't guarantee it,
because my English isn't very good. 

-- 
Olli Rajala "Quite normal guy"

Bacgkround information about Linux
http://gamma.nic.fi/~h.rajala/linkit/links.php?linux


From donn at oz.net  Mon Dec 18 01:41:43 2000
From: donn at oz.net (Donn Cave)
Date: 18 Dec 2000 06:41:43 GMT
Subject: Interrupting a thread...
References: <91hqgk$pu0$1@bob.news.rcn.net>
Message-ID: <91kbj7$3q2$0@216.39.151.169>

Quoth "David Allen" <s2mdalle at titan.vcu.edu>:
| I'm writing a program which does some networking,
| and I have a thread that is devoted to grabbing 
| things off of the network so that the program won't
| hang while it is reading data.
|
| The problem is, I also occasionally need to interrupt
| this data fetching operation, so I need a way to
| either raise an exception in the other thread, or
| otherwise signal to it to stop what it's doing.
| Since the download thread calls other functions
| that are not sensitive to the data in the original
| object, I can't check to see if some value has 
| changed each time I'm reading in a loop.
|
| Ideally, I'd like thread 1 to raise an exception
| that would be propagated in thread 2, not thread 1.
| Is this possible?  What other techniques might work?

No, and for me it would not be all that ideal either.
Exceptions don't interrupt anything, they're more of
a flow control mechanism.

I don't understand paragraph two, where functions
are not sensitive to the original object (what
original object?), so I'll guess you're asking a
question I have an answer for.  You actually have
a socket, and you want a thread to read data from
the socket as well as receive some other inputs.

The select() function allows you to wait for
events on multiple sockets/pipes.  In order to
take advantage of that, you need to create a
pipe or another socket, dedicated to communication
between the main thread and the socket thread.
The socket thread selects on the socket and the
read end of the pipe.  The main thread writes
to the pipe when it has something to say to the
socket thread.  That won't interrupt an actual
data transfer, but I don't think you really want
to do that.

	Donn Cave, donn at oz.net


From thomas at xs4all.net  Sat Dec 23 12:08:19 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Sat, 23 Dec 2000 18:08:19 +0100
Subject: <> vs != was Re: A Suggestion for Python Colon Syntax
In-Reply-To: <LNBBLJKPBEHFEDALKOLCKEDEIFAA.tim.one@home.com>; from tim.one@home.com on Sat, Dec 23, 2000 at 12:15:25AM -0500
References: <2kM06.120$Lm2.7518@news.clarkson.edu> <LNBBLJKPBEHFEDALKOLCKEDEIFAA.tim.one@home.com>
Message-ID: <20001223180819.E5515@xs4all.nl>

On Sat, Dec 23, 2000 at 12:15:25AM -0500, Tim Peters wrote:

> Best guess is that "<>" will go away in P3K.  I jokingly told Guido it was
> clearly wrong for an inequality operator to look so symmetric, and he forgot
> to laugh <wink>.

Well, excuse me, but it's clearly wrong for an equality test operator to
look so much like an augmented assignment statement ! I definately prefer
'<>' myself, anyway, I just wish I could use it in C without turning towards
preprocessor macros :)

+=-*=-/=-!=-ly y'rs,

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From wware at world.std.com  Mon Dec 25 23:43:34 2000
From: wware at world.std.com (Will Ware)
Date: Tue, 26 Dec 2000 04:43:34 GMT
Subject: Season's greetings (was Re: Some Python 2.1 ideas)
References: <mailman.977765348.13376.python-list@python.org> <928otj02l39@news1.newsguy.com>
Message-ID: <G65rsM.4K5@world.std.com>

Alex Martelli (aleaxit at yahoo.com) wrote:
> One thing I wasn't able to easily answer -- what's the
> *easiest* way for a Python beginner to do function-plots,
> with nicely labeled graphs, minor/major ticks, color,
> auto-scaled, &c &c?

Perhaps there is a Windows port of gnuplot? That's what I use
on Linux; I've never tried doing graphs on Windows (well once
or twice in Excel but never from Python).

-- 
import string,time,os;print string.join((lambda x:x[:10]+x[8:])(map(
lambda x:string.center("*"*(lambda x:((x<24) ### Seasons Greetings, Will Ware
*(x-3))+3)(x),24),range(1,28, 2))),"\n") ################ wware at world.std.com


From billtj at y.glue.umd.edu  Thu Dec 21 12:43:29 2000
From: billtj at y.glue.umd.edu (William Djaja Tjokroaminata)
Date: 21 Dec 2000 17:43:29 GMT
Subject: A Suggestion for Python Colon Syntax
References: <91tc8h$2pp$1@hecate.umd.edu> <W2r06.1076$AH6.86249@newsc.telia.net>
Message-ID: <91tfg1$282$1@hecate.umd.edu>

Fredrik Lundh <fredrik at effbot.org> wrote:
: "William Djaja Tjokroaminata" wrote:
:> However, a lot of time when I ran my Python program, I got syntax
:> error, because I forgot to put colon (:) after my if (or for, etc.) clause.
:> My question is, is the colon really necessary, syntactically, in Python?

: Short answer:  The colon isn't there for the computer, it's
: there for human readers.  Just like the colons in your post:

Well, the colons in my post is because I am writing in English.  But
Python does not have semicolon etc, just to make it "pretty", but they do
have consequences if they are omitted.  I personally don't see how the
colon help me more in distinguishing the clause from the body of the
clause when reading a Python code, because the indentation is supposed to
be the marker.

:> My suggestion is as follows:
:>
:> 1) If the if (or for, etc.) clause is followed by a block, then the colon
:> is optional (for backward compatibility):
:>
:>    if condition is true
:>        statement_1
:>        statement_2
:>        ....

: (Using colons and indentations in the same way as in other written
: languages is just one of many ideas Python has borrowed from the
: ABC language: http://www.cwi.nl/~steven/abc )

Well, I don't know the rest of the people who use Python, but I
personally don't know anything about the ABC language.  If a feature is
borrowed from C or Java or Perl, I think it is good.  But a feature that
seems to add nothing from an obscure language to me is questionable.

: :::

: Longer answer: ... nah, I'll leave that to the "while true" crowd.
: I'm sure they'll keep you entertained until the end of next year!

I am waiting, everyone....

: :::

: Happy holidays, everyone!

: </F>


Bill



From ljohnson at resgen.com  Wed Dec  6 10:26:13 2000
From: ljohnson at resgen.com (Lyle Johnson)
Date: Wed, 6 Dec 2000 09:26:13 -0600
Subject: How can i use a DLL in python ?
References: <t2sjn0aji5dkef@corp.supernews.com>
Message-ID: <t2smf8njad8g2a@corp.supernews.com>

You have at least two options that I'm aware of. One is to use Sam Rushing's
"calldll" extension that allows you to call Win32 DLL functions directly
from Python. I've never used this and so I don't know what restrictions it
might have. Here's a link:

    http://www.nightmare.com/software.html

Another option (which I've used a lot) is to create Python wrappers for
those C/C++ functions using SWIG:

    http://swig.sourceforge.net

I'm not familiar with the Adobe toolkit in question, but if it's API is
slanted towards C/C++ programmers then the same functions may prove to be
awkward for use in Python programs. Using SWIG you could cook up some more
Python-esque interfaces to the same functionality without too much work.

Hope this helps,

Lyle


"^Socrates^" <joseamartin at venmex.com> wrote in message
news:t2sjn0aji5dkef at corp.supernews.com...
> The problem is that.
>
> I am trying to port the fdftk.dll  and fdftk.lib from adobe acrobat to
work
> in python
> with it.
>
> Adobe was made a toolkit for Java,C++,and Perl , but currently no version
> for Python.
> I was emailed to them and request the Python Version, but a have no
response
> from them.
>
> Then I need to do This by myself !.
>
> How can use this Dll and .lib from Python ?
> Can i use this modules  ?
>
> Thanks...
>
> Jose Antonio martin
> joseamartin at venmex.com
> joseamartin at cantv.net
>
>
>
>




From MarkH at ActiveState.com  Thu Dec 21 02:55:00 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Thu, 21 Dec 2000 07:55:00 GMT
Subject: Fatal Python error: PyThreadState_Get
References: <91rcjo$2dtl$1@news2.ottawa.cyberus.ca>
Message-ID: <3A41B623.20502@ActiveState.com>

Gordon Williams wrote:

> Fatal Python error: PyThreadState_Get: no current thread
> 
> abnormal program termination
> 
> I'm sending and receiving information in one thread via a serial port.  In
> the 1.5.2 code I was using the Marshal serial module.  As that only works
> with 1.5.2, I have now converted that to use windows native stuff using Mark
> Hamond's work.

It means that something using callbacks isn't playing ball.  Can you 
narrow it down and send me a sample - these are easy to fix once you 
know what function is causing it.

Mark.

> 
> There was a thread earlier this month about this same problem.  I think that
> they were also using W95.  It was unclear whether it was resolved or how it
> was resolved.
> 
> Before I start tracking the problem down, is this a known problem?  What
> causes it? Any pointers on where I should look first?
> 
> Regards,
> 
> Gordon Williams




From Armin_member at newsguy.com  Sun Dec 31 05:26:57 2000
From: Armin_member at newsguy.com (Armin Steinhoff)
Date: 31 Dec 2000 02:26:57 -0800
Subject: SMB Module
Message-ID: <92n1lh02evv@drn.newsguy.com>

In article <slrn94trot.b73.mballen at angus.foo.net>, mballen at erols.com says...
>
>I frequent the Samba mailing list. A couple of days ago Richard Sharp
>put together an smbclient.so library of client functions for SMB. SMB
>is Microsoft's native file and print sharing protocol. Someone wrote a
>Perl module using it. I just thought one of you guys might want to do the
>same for Python. Seems only natrual since there's a nice little library
>all packaged up with open(), read(), write(), ...etc all setup easy like.

Makes no sense, because a SMB client mapps normaly a M$ volume to a directory
tree (UNIX) or a virtual drive (M$) ... I don't see in that case any tasks for
Python.

Armin

Python in real-time: http://sourceforge.net/projects/pyqnx



From qrczak at knm.org.pl  Fri Dec 29 04:20:03 2000
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 29 Dec 2000 09:20:03 GMT
Subject: unix pipes + python
References: <m4iu27odrr4.fsf@macquarie.com.au>
Message-ID: <slrn94olq3.s27.qrczak@qrnik.zagroda>

29 Dec 2000 10:56:47 +1100, Timothy Docker <timd at macquarie.com.au> pisze:

> This script behavious almost as expected. I end up with the text
> echoed back, but the 'cat' subprocess is left running - I was
> expecting it to exit when the write side of the pipe was closed in the 
> parent, but it is still reading from fd 0.

Close the write side in the child, just after forking. As long as
someone has the pipe open for writing, cat does not see the end
of file (even if someone == cat itself).

I would also close the read side in the parent.

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


From root at rainerdeyke.com  Fri Dec  8 22:25:59 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Sat, 09 Dec 2000 03:25:59 GMT
Subject: Python wrapper for C++ STL?
References: <3A31760E.D157EB54@erols.com> <npfY5.1850$x6.2577114@news2.rdc2.tx.home.com> <90s2lm$ig1$1@nnrp1.deja.com>
Message-ID: <bDhY5.1999$x6.2840573@news2.rdc2.tx.home.com>

<shindich at my-deja.com> wrote in message news:90s2lm$ig1$1 at nnrp1.deja.com...
> In article <npfY5.1850$x6.2577114 at news2.rdc2.tx.home.com>,
>   "Rainer Deyke" <root at rainerdeyke.com> wrote:
> > "Edward C. Jones" <edcjones at erols.com> wrote in message
> > news:3A31760E.D157EB54 at erols.com...
> > > Has anyone written a Python wrapper for the C++ Standard Template
> > > Library?
> >
> > Why would anybody want to?  Python provides most STL fnctionality
> natively.
> > Such a wrapper would need to constantly translate C++ exceptions to
> Python
> > exception and vice versa, making the wrapper be more trouble than it's
> > worth.

> STL has other things that are of great use to C++ developers -- sets
> and a bunch of generic algorithms to name a few. But STL looses its
> beauty without the type safety. And Python is not type safe. I guess
> the only feature that I really miss is std::set. But a set can be
> implemented as a dictionary that uses the same object as both the key
> and the value.

I tend to use a dictionary where all values are None for sets.

The STL does have some nice features which are missing from Python.  For
example, C++ sets and maps are automatically sorted.  However, I feel that
it would be easier to write a Python extension that provides these features
from scratch than to write a wrapper for the STL.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From gregj at pobox.com  Sat Dec 30 20:49:38 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Sun, 31 Dec 2000 01:49:38 GMT
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <05C6FUhLDNUU-pn2-tuTnzOCokxem@bill>
Message-ID: <Sgw36.208288$U46.6385684@news1.sttls1.wa.home.com>

I don't have any problem with Python's flow control statements, but it seems
a lot of people would like something equivalent to C's "do ... while" (or
"do ... until" if that turns you on). Logically the "while ... if ... break"
construction can handle anything, but the idiom stretch a bit thin. When
reading well-written C, "while (1)" or "for (;;)" means "loop an unknown
number of iterations," and "do ... while" means "loop at least one
iteration." If Python had a "do ... while" loop, some programmers could
write clearer code. It's not world peace, but it's something.

Some people don't like using the "break" statement to get out of a
loop--perhaps bad memories of BASIC and unpredictable loop index values.
They'll have to get over it. Python does have the (somewhat unusual) "else"
statement that can attach to "for" and "while" loops, in case the programmer
must distinguish between the loop running to completion or terminating by
"break." I haven't needed the "while ... else" (or "for ... else")
construction, probably because C doesn't have it and I think in C.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From prashanth.mundkur at kla-tencor.com  Wed Dec 13 15:13:31 2000
From: prashanth.mundkur at kla-tencor.com (Prashanth Mundkur)
Date: 13 Dec 2000 21:13:31 +0100
Subject: unwrapping argument tuples
Message-ID: <3A37D870.F3D01433@kla-tencor.com>

is there a way to unwrap a tuple argument to a function and
match it to the parameters?

that is, if

>>>lst = [1, 2, 3]

>>>def func (p1,p2,p3):
       return p1+p2+p3

then I want to apply func somewhat as

>>>func(lst)
6

such that the elements of lst are unwrapped and and matched to p1, p2,
p3

tx for any tips.

--prashanth



-- 
Posted from [206.67.220.38] 
via Mailgate.ORG Server - http://www.Mailgate.ORG


From wmistar at cyclones.com  Mon Dec 25 01:08:04 2000
From: wmistar at cyclones.com (wmistar at cyclones.com)
Date: Mon, 25 Dec 2000 06:08:04 GMT
Subject: Problem passing COM object
Message-ID: <926o83$39e$1@nnrp1.deja.com>

Hi,

I'm trying to call the following COM interface: (generated by MakePy)

	def AutoLogIn(self, UserName=defaultNamedNotOptArg,
Password=defaultNamedNotOptArg, databaseDir=defaultNamedNotOptArg,
ServerName=defaultNamedNotOptArg, forceLogin=defaultNamedNotOptArg,
appObject=defaultNamedNotOptArg):
		return self._oleobj_.InvokeTypes(0x2, LCID, 1, (3, 0),
((8, 0), (8, 0), (8, 0), (8, 0), (11, 0), (16393, 0)),UserName,
Password, databaseDir, ServerName, forceLogin, appObject)

Notice the 7th parameter 'appObject'.  When I pass an object that
was instantiated with:
tr = win32com.client.Dispatch('TRDatabase.Application')

The error comes out as follows:
>>> dbh = trdb.AutoLogin("", "", "", "", 0, tr)
Traceback (innermost last):
  File "<interactive input>", line 1, in ?
  File "<COMObject TRDatabase.Database>", line 2, in AutoLogin
com_error: (-2147352571, 'Type mismatch.', None, 6)

The tr is confirmed to be a COM object by doing this:
>>> tr
<COMObject TRDatabase.Application>

So why the Type mismatch error? Do I have to do something special
to pass around Application COM objects?

In VB, the same call is sucessfully done by doing this:
Dim mtr As TRDatabase.Application

dbh = trdb.AutoLogin("", "", "", "", bForceLogin, mtr)

Please help,
-Will--


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


From jcm at bigskytel.com  Sat Dec 23 06:01:10 2000
From: jcm at bigskytel.com (David Porter)
Date: Sat, 23 Dec 2000 04:01:10 -0700
Subject: Python in "math through programming" curriculum
In-Reply-To: <3A445754.E88D0A01@gte.net>; from cnelson9@gte.net on Sat, Dec 23, 2000 at 07:11:46AM +0000
References: <2g2t3t06r2agrrjmfnr41eu4lc2u39a8qi@4ax.com> <3A3EFF47.67DE69CF@gte.net> <kj624toonmgugtrgrpkea9b01a56hgogam@4ax.com> <3A413A14.54DF58D1@gte.net> <laj24tcfio6ldfqktnehkck93j7v6iciiv@4ax.com> <3A445754.E88D0A01@gte.net>
Message-ID: <20001223040110.A25773@bigskytel.com>

* Clifford J. Nelson <cnelson9 at gte.net>:
> 
> Of course, when I say "look at them", I mean download the
> version that runs on your computer, read all the
> documentation, and try programming with them for a year or
> two like I have. Then you won't have to ask so many simple questions.

A year or two? Then they already lose to Python. 


David

-- 
        You correct it. 



From quinn at lira.ugcs.caltech.edu  Thu Dec  7 00:34:14 2000
From: quinn at lira.ugcs.caltech.edu (Quinn Dunkan)
Date: 7 Dec 2000 05:34:14 GMT
Subject: Bug: Readline getting stuck on Linux and Solaris
References: <lxk8zqwupy2.fsf@ioasun3.epfl.ch> <90m53k$ovh$1@nnrp1.deja.com> <m3u28h48ja.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <slrn92u8al.76b.quinn@lira.ugcs.caltech.edu>

On 06 Dec 2000 20:08:41 +0000, Michael Hudson <mwh21 at cam.ac.uk> wrote:
>sragsdale at my-deja.com writes:
>
>> In article <lxk8zqwupy2.fsf at ioasun3.epfl.ch>,
>>   Jan Kybic <Jan.Kybic at epfl.ch> wrote:
>> > Hi,
>> > 	I am using Python 2.0, I downloaded it in the RPM form and I
>> > also compiled it from source on both Linux RedHat 6.2 and Solaris. In
>> > both cases, when I have the readline module enabled, the keyboard
>> > input gets stuck.
>> 
>> I'm having similar problems with Python 2.0 hanging in Irix.  What's
>> most annoying is that this happens on the Python command line when you
>> hit control-C.  In Python 1.5.2 this is a KeyboardInterrupt event.  In
>> Python 2.0 this hangs the interpreter which starts using 100% of the
>> processor.  I can't kill it with more control-C's.  If I disable the
>> readline (4.1) module this problem goes away.  This problem does not
>> happen in Linux 2.2.15.  I can't find this bug listed on the sourceforge
>> bug list.
>
>Hmm.  This *used* to happen on Linux (with slightly different
>symptoms; control-C would kill the interpreter, not hang it), but the
>bug got fixed.  A couple of things to try: does control-C do what it
>should when you're not sitting at the ">>> " prompt?  E.g. if you type
>
>>>> while 1:
>...    pass
>...
>
>and whack control-C, what happens?
>
>The other thing is to try building without threads.
>
>The problem with Linux was a sigaction/signal mismatch - but I don't
>know a lot about signal handling and nothing at all about IRIX, so I
>can't say if I'd expect trouble here....

Strange.. readline has *never* worked right for me, on IRIX or Linux, on any
version of python (1.4, 1.5.2, 2.0), and any version of readline (3.0, 4.0,
4.1), if python is compiled with threads.  IRIX (6.4, 6.5, 6.5.2) hangs, Linux
(2.0.36, 2.2.12, 2.2.17) exits.

I gave up trying to get it to work a long time ago, and compiled a seperate
python with no threads for when it's important.

I just tried your 'while 1: pass' thing, and the behaviour is the same there
too.


From sholden at holdenweb.com  Mon Dec 18 09:51:52 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 18 Dec 2000 09:51:52 -0500
Subject: Alarms in threads - how to get/mimic?
References: <Pine.LNX.4.21.0012161749380.7913-100000@traal.eecs.berkeley.edu> <ltd7erk8xd.fsf@asfast.com>
Message-ID: <Ryp%5.39166$n51.703086@e420r-atl1.usenetserver.com>

Lloyd Zusman <ljz at asfast.com> wrote in message
news:ltd7erk8xd.fsf at asfast.com...
> Luca de Alfaro <dealfaro at eecs.berkeley.edu> writes:
>
> > [ ... ]
> >
> > What is not clear to me is the following.  When there is a timeout,
> > the exception Timeout is not caught by httplib.  Hence, if the thread
> > that called httplib just continues its job giving up on the
> > connection, httplib will not have a chance to do any clean-up, such as
> > for example releasing/closing the socket.
> >
> > Hence, it seems to me that in order to abort a web page load cleanly
> > upon timeout, I also have to modify the httplib module to handle the
> > exception Timeout, and do clean-up work if that happens.
> >
> > Is my understanding correct?
>
> Not entirely.  When using timeoutsocket, all you have to do is start
> catching timeout exceptions when you make the various method calls
> within httplib
[snip code]

Basically, if the code you make use of traps the errors raised by
timeoutsocket and re-raises them after it's done whatever cleanup it needs
to then there should be no problem.  However, there MAY be problems if the
timeoutsocket error is not caught by the higher-level protocol code, and
makes it to the outer code which called it.  In that case there is a
possibility that necessary clean-up may not be performed.
>
> It's recommended in Python that exception handling be done whenever
> it's appropriate; therefore, by wrapping your httplib calls in this
> same manner, you'll remain true to the Pythonic Way, and you won't
> have to re-write any httplib code when using timeoutsocket.
>
> Note that the timeoutsocket.setDefaultSocketTimeout() method call
> should be made before making any httplib calls.  If you want to use
> different timeout values when making different httplib calls (or other
> calls socket-based methods), you'll have to re-invoke the
> setDefaultSocketTimeout() method before each call that uses a changed
> timeout value.
>
Note also that you can set the timeout on each socket individually.
setDefaultSocketTimeout is usually used to ensure that connect() attempts in
upper-layer protocols (which happen before you have a socket object in your
application to manipulate) don't hang forever.

regards
 Steve







From fredrik at effbot.org  Mon Dec 11 12:51:41 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Mon, 11 Dec 2000 17:51:41 GMT
Subject: PIL binary for Python 2.0 for Windows ?
References: <3a2f472f$0$235@hades.is.co.za> <KZOX5.2015$Kd1.294314@newsb.telia.net> <IlSX5.24763$R77.1876538@nnrp4.clara.net> <9JnY5.2716$O%.428568@newsc.telia.net> <Xns9005AB80Aseed7@news1.zwoll1.ov.nl.home.com>
Message-ID: <Nu8Z5.3452$Kd1.391383@newsb.telia.net>

"ingo" wrote:
> >also see the updated pil page:
> >    http://effbot.org/pil
>
> When I use these and run the viewer.py demo from 1.1.1 scource I get an
> error:
>
>     import _imagingtk
> ImportError: No module named _imagingtk
>
> Did I miss something?

I didn't ship an _imagingtk module, since it's dependent
on both the Python version you're using and the Tcl/Tk
version...

pythonware.com will release a PythonWare PY200 kit any
day soon, which includes versions of Python, Tkinter, PIL
(etc) that are known to work together...

</F>




From pearu at ioc.ee  Wed Dec 27 15:27:46 2000
From: pearu at ioc.ee (Pearu Peterson)
Date: Wed, 27 Dec 2000 22:27:46 +0200
Subject: Python 2.0 parser module bug(?) and a Q
Message-ID: <Pine.LNX.4.10.10012272204060.330-100000@kev.ioc.ee>

Hi,

I have found that parser module may have a bug in its functions tuple2ast
or suite.
This is illustrated in the following python session:

>>> import parser
>>> parser.tuple2ast((parser.suite('import sys,math')).totuple())
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
parser.ParserError: Expected node type 280, got 279.
>>> parser.tuple2ast((parser.suite('import sys\nimport math')).totuple())
<ast object at 0x810e978>

I am using Python 2.0.

And my question is (not really related to above):

  Is there inverse function for parser.suite? 

For example, if I create an ast object

	ast = parser.suite('x=5\ny=x+2')

then I'll transform it to tuple:

	t = ast.totuple()

so that I can analyze it, create a new ast-tuple `t2' from it, may be a
bit changed, but then using

	ast2 = parser.tuple2ast(t2)

I get another ast object. In order to view it, I would like to
transform it to a string but I could not find any method/function for
that.

I would be satisfied if `t2' could be transformed to the corresponding
string. Has anyone implemented a function for that?


Thanks,
	Pearu




From rturpin at my-deja.com  Mon Dec 25 11:25:41 2000
From: rturpin at my-deja.com (rturpin at my-deja.com)
Date: Mon, 25 Dec 2000 16:25:41 GMT
Subject: One Python 2.1 idea
References: <mailman.977590932.19335.python-list@python.org> <9237md$orm$1@nnrp1.deja.com> <924bjt0e3n@news1.newsguy.com> <925baq$5sj$1@nnrp1.deja.com> <925tak01tpu@news1.newsguy.com> <926ici$vi9$1@nnrp1.deja.com> <92722s0l1f@news1.newsguy.com>
Message-ID: <927se2$q7n$1@nnrp1.deja.com>

I hope to write more about this in another post, but for
now, just a few short comments.

In article <92722s0l1f at news1.newsguy.com>,
  "Alex Martelli" <aleaxit at yahoo.com> wrote:
> C++ offers lots of very-low-level constructs AS WELL
> AS pretty high-level ones, and, as I said, the low-level
> parts keep coming up then and again even when one is
> trying to work more abstractly ..

To me, this is a BIG reason for preferring Python to
C++. Yes, you can implement any abstraction in C++.
Of course. But they are ADDITIONS to the language. You
cannot decipher a C++ program without also knowing the
lower level abstractions. Adding higher level
abstractions makes the language larger, more complex,
harder to maintain, and less productive. Once again,
I'm going to push the notion that the practical power
of a language is its level of abstraction divided by
its complexity. The reason a lower level language like
C++ will never provide the productivity boost of a
higher level language like Python is precisely that
it has to become more complex as it builds up to the
same level of abstraction.

> RDBMS enjoy a crucial difference: the low-level
> information on how best to optimize (e.g.) a join
> is in full possession of the RDBMS; it does not so
> much depend on what SQL queries I'm using now
> or in the future, but on the physical arrangement of
> tables, indices, &c, which the DB 'knows'. ..

This is even more true for optimizing compilers. They
will recognize access patterns that you don't, and
in contexts where they are not at all obvious. If you
think you can do nearly as a good a job at this as
even a SIMPLE data and control flow analysis, I think
you are over-estimating your abilities, and under-
estimating the benefits of automatic analysis.

> I _don't_ like programs which try to read my mind
> and "do what's best for me whether I want it or not".

Let's be clear that we are talking only about FUNCTIONALLY
IDENTICAL implementations of language abstractions. Are
you really arguing that Python programmers should be
fully cognizant of how each abstraction is implemented?
When you write Python, you keep present in your mind the
memory management strategy for sequences, the lookup,
insertion, slice, and deletion algorithms for strings,
the find, replace, and add algorithms for dictionaries,
the pre-allocation strategy for constant methods and
data members, how regular expressions are compiled, and
the garbage collection strategy for everything? And more,
you keep track of how this changes from point release to
point release? If so, I think C++ might be a better
language for you!

> .. IF I have a semantic constraint about there being
> no deletes on this object, it's a CRUCIAL part of my
> abstraction for it -- I want to state it clearly and
> readably, and let compiler and human readers make of
> my statement what they wish.

But this is NOT what we are talking about. IF it is
important functionally to some list object that there are
no deletes from it, if deleting from it is an application
error, then you SHOULD create a new type, and build in
this constraint, EVEN IF it imposes a performance penalty
to ensure this constraint. But that is NOT the issue. The
issue is the much more common case where it just so
happens that there are no deletes from a list, in the
code as it appears now, and this can be used for
performance optimization. It is NOT part of the
application's semantics, but just a happenstance of
current code.

> .. As the program designer, I *know* whether I intend
> deletes to be 'forbidden' on the object, or if it just
> happens accidentally that there are none YET; ..

And in the far more common case where there are none YET,
that fact can and should be used to improve performance,
until it is no longer easily determined. And there are
MANY patterns like this. Are all insertions at the end?
Are all accesses sequential? Etc. These are all patterns
that can be used for optimization. Note the push for a
special construct for iteration. None of these are
intended as explicit semantics of the application. You
now write:

     for i in m.keys():

Tomorrow, you fix a bug or add functionality, and write:

     tempL = m.keys()
     for i in tempL:
         ..
     for i in tempL:
         ..

Both can be optimized. In neither case is "list is use
only for iteration" nor "list is used for two iterations"
ever a part of what you are trying to achieve in the
application. And if you introduce tempL in the second
version only for performance's sake, that means (a) you're
making assumptions about how byte code is generated that
may fail in version 3.1, (b) you're having to think about
things at a level different from the innate Python
abstractions, and you're productivity is thereby lessened,
and (c) Python 2.1 has failed to provide you adequate
performance. My own preference is to think about things
in terms of Python abstractions, write clear and readable
code, and assume that the underlying abstractions are
implemented efficiently. The last is a bit of a
counter-factual, which is why we are having this
discussion.

Russell


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


From sragsdale at my-deja.com  Wed Dec  6 14:47:03 2000
From: sragsdale at my-deja.com (sragsdale at my-deja.com)
Date: Wed, 06 Dec 2000 19:47:03 GMT
Subject: Bug: Readline getting stuck on Linux and Solaris
References: <lxk8zqwupy2.fsf@ioasun3.epfl.ch>
Message-ID: <90m53k$ovh$1@nnrp1.deja.com>

In article <lxk8zqwupy2.fsf at ioasun3.epfl.ch>,
  Jan Kybic <Jan.Kybic at epfl.ch> wrote:
> Hi,
> 	I am using Python 2.0, I downloaded it in the RPM form and I
> also compiled it from source on both Linux RedHat 6.2 and Solaris. In
> both cases, when I have the readline module enabled, the keyboard
> input gets stuck.

I'm having similar problems with Python 2.0 hanging in Irix.  What's
most annoying is that this happens on the Python command line when you
hit control-C.  In Python 1.5.2 this is a KeyboardInterrupt event.  In
Python 2.0 this hangs the interpreter which starts using 100% of the
processor.  I can't kill it with more control-C's.  If I disable the
readline (4.1) module this problem goes away.  This problem does not
happen in Linux 2.2.15.  I can't find this bug listed on the sourceforge
bug list.


Sent via Deja.com http://www.deja.com/
Before you buy.


From tjreedy at udel.edu  Wed Dec 20 12:47:09 2000
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 20 Dec 2000 12:47:09 -0500
Subject: Python's scope rules
References: <3A40813B.C04D21C@et.dtu.dk>
Message-ID: <91s38r$sjd$1@news.udel.edu>

"Brian Elmegaard" <be at et.dtu.dk> wrote in message
news:3A40813B.C04D21C at et.dtu.dk...
> I am not a computer scientist, just an engineer knowing how to program in
> Fortran, Pascal,... So I after reading some of the python material on the
web, I
> decided to give it a try and is now using the language some, finding it
fun and
> easy.

I'm a statistician who discovered the same thing about four years ago.

> On usenet I have now learned from skilled scientists,

By their own evaluation? Think instead 'CS theologians' or even 'religious
fanatics'.

> that Python has some  deficiencies regarding scope rules, and I am not
capable
> of telling them why it  has not (or that it has).

Deficiency is in the eye of the beholder.  This religious discussion has
been going on for at least since I started listening; probably has since
the beginning of Python; and will probably continue until Python dies.
Consider not getting too involved.

> In Aaron Watters 'The wwww of python' I have read that Python uses
lexical
> scoping with a convenient modularity. But, other people tell me it does
not.
> Some say it uses dynamic scoping, some say it uses it own special
> 'local-global-builtin' scope. What is right?

I have discovered that while good computer scientists may develop a
definsible terminology that they use consistently each unto themselves,
they do not always agree.  This seems to one of those areas where language
is not settled.

> The above mainly is a theoretical question.

A question like 'what is the optimal asymptotic efficiency of a sorting
algorithm based on comparisions' has an answer: n log n.  The scope
discussion does not seem to.

> A more practical example which I  agree seems a bit odd is:
> >>> x=24
> >>> def foo():
> ...     y=2
> ...     print x+y
> ...
> >>> def bar():
> ...     x=11
> ...     foo()
> ...
> >>> bar()
> 26

When a function encounters a name that is not in its local scope, where
should it look to find a value to substitute (given that there are no
explicit pointers)?  There are three general answers: nowhere (raise an
exception or abort); somewhere in the definition environment; or somewhere
in the usage environment.  Python's current answer is in the namespace of
the defining module (and then in the builtin namespace attached to that
module).

The only things I find 'odd' (and definitely confusing at first) is the use
of 'global' for the module namespace (and I suppose there is some history
here) and the sometimes claim that there are two rather than three possible
scopes.

For odd behaviour, consider the following possibility:

plane_geometry.py:
_dimension = 2
def SegSectTri2d(line_segment, triangle):
  "does line_segment intersect triangle in xy-plane"
  # uses _dimension and depends on it being 2
...
space_geometry.py
import plane_geometry
_dimension = 3
def SegSectTri3d(line_segment, triangle):
  "does line_segment intersect triangle in xyz-space"
  # uses _dimension and depends on it being 3
  # uses plane_geometry.SegSectTri2d() for part of its calculation

Now, I would be disappointed if the plane geometry functions did not work
when called from space geometry functions because the two modules happened
to use the same spelling for the private dimension constant.

Terry J. Reedy





From dale at out-think.NOSPAMco.uk  Wed Dec 13 13:35:33 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Wed, 13 Dec 2000 18:35:33 +0000
Subject: Source Code Management
References: <fb4f3tsgffnmvme64aol1cn397te7ln349@4ax.com> <918711$meh$1@reader1.imaginet.fr>
Message-ID: <prff3tc59qtv0qr9afj16dali3jv789c7u@4ax.com>

"Gilles Lenfant" <glenfant at equod.com.nospam> wrote:

>Have a look to cvs at http://www.gnu.org/software/cvs/cvs.html
>http://www.wincvs.org
>http://www.cvshome.org/
>
>The best of the open source version control tool.
>There are python wrappers to CVS somewhere, I don't remember where.
>

I've just dug CVS out and downloaded it but I discover that the NT version appears to require a
mounted volume.

That's a disappointment.


--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants


From sandj.williams at gte.net  Wed Dec 20 18:57:41 2000
From: sandj.williams at gte.net (Steve Williams)
Date: Wed, 20 Dec 2000 23:57:41 GMT
Subject: ANNOUNCE: soundex.py
References: <mailman.977338109.32467.python-list@python.org>
Message-ID: <3A414897.D82C953C@gte.net>

Skip Montanaro wrote:

> Python 2.0 no longer ships with a soundex module.  Sometime ago, Tim Peters
> and Fred Drake each cooked up replacements written in Python.  I merged them
> together into a single module which is available from
>
>     http://musi-cal.mojam.com/~skip/python/
>
> If you have any questions or comments on the module, please send them my
> way.
>

Soundex routines traditionally return a fixed number of characters--the NDIGITS
in your routine.  That's 'cause the system was developed before computers.

I've found you can good results with a variable length soundex string--the more
information you give the routine (first names, middle names, prefixes and
suffixes) the better/smaller the result set.

Store the full soundex key as a varchar in your database and use the SQL LIKE
statement to do the retrieval.

This is particularly useful with one syllable surnames--you really need to add
more to the name to get anything useful.  (Mao Tse-Tung == M000 vs. M32352, you
be the judge).

For example,
    print get_soundex('van')
V5
    print get_soundex('van der tamp')
V536351
    print get_soundex('van der tamp, albert')
V53635141632
    print get_soundex('van der tamp, albert c, lieutenant colonel, Phd')
V5363514163243553245413   <== this is the full key stored as a varchar in your
database

So a retrieval like V53635141632435532454% will return all the lieutenant
colonel albert c van der tamps in your database, whether they have PhDs or not.




From zope at TheWebsons.com  Wed Dec 27 01:05:50 2000
From: zope at TheWebsons.com (Ben Ocean)
Date: Tue, 26 Dec 2000 22:05:50 -0800
Subject: Python vs PHP (& Java?)
Message-ID: <LOBBJMHOIFOLFIPFANHJAEMICFAA.zope@TheWebsons.com>

Hi;
I?m an intermediate-level programmer that?s new to Python (via Zope) and I
like it a lot. However, it appears to me that, while Python is a good
replacement for Perl, it cannot replace much of the functionality of PHP,
which can be scripted into HTML (whereas Python, it appears, must be
called). So, I have to wonder about the speed of Python vs. PHP for
Internet-based applications. Python has the advantage of byte-code and
compilation, but it?s server-side. PHP might not be as fast on the server,
but it doesn?t have to make as many trips back and forth between server and
client. Plus, I?m curious as to why Python wasn?t written in such a way that
it could be written directly into HTML documents like PHP. On a similar
subject, someone once informed me that Java was by far the best language for
scripting communication programs over the Internet (like chat, IRC, etc.).
How does Python compare? Your comments are appreciated.
BenO
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20001226/e5c37baf/attachment.html>

From gherman at darwin.in-berlin.de  Sun Dec 17 07:57:33 2000
From: gherman at darwin.in-berlin.de (Dinu C. Gherman)
Date: Sun, 17 Dec 2000 13:57:33 +0100
Subject: Singleton vs Proxies DP (was Re: Solution: Direct access to Printer 
 I/O lines)
References: <917k9b$9i9$1@nnrp1.deja.com> <slrn93hrjn.2vp.ssthapa@localhost.localdomain> <91cnp3$g59$1@nnrp1.deja.com> <91d23f0jk7@news1.newsguy.com> <91gfnp$7r4$1@animus.fel.iae.nl> <91gkh801e49@news2.newsguy.com>
Message-ID: <3A3CB83D.45C8775E@darwin.in-berlin.de>

Alex Martelli wrote:
> 
> "Carel Fellinger" <cfelling at iae.nl> wrote:
> >
> > And in an other thread you wet my appetite and direct me here!
> > So let us hear that rant, mister, and enlight us mere mortals
> > who still think that there is legitimate use of this singleton
> > pattern.
> 
> [...]
> 
> Within the component that implements the singleton, it
> does not matter how and where you place the state, and
> the code that knows about that state's representation and
> accesses and modifies it directly.  You may choose to
> have that state as instance-variables of an object, that
> code in its methods, and ensure a single instantiation of
> that object, but that's a somewhat roundabout way for
> languages which allow per-class (C++) or per-module
> (Python) functions and data.  But it's a minor decision, of
> implementation more than of design; what matters is the
> interface you choose to expose from that component.
> 
> And _should_ that interface be one of a single object that
> gets instantiated once, so it must be accessed through
> some single 'give me a reference to Mr. Singleton please'
> function?  I do not see why that could ever be the ideal
> interface.  There are all sorts of issues that give slight
> problems, and I don't see any compensating benefit when
> compared with the following approach.
> 
> Why not have client-code ask for objects normally,
> use them, then drop them when it's done?  If client
> code asks for 'another object' -- why not hand it out?
> 
> It doesn't matter if the 'handed out' objects have
> different *identities*... as long as they share *state
> and behavior*, no?  I forget on what name this simple
> pattern was originally published in "C++ Report",
> but let me call it 'lightweight proxies' (LWP, versus S
> for the singleton pattern).  It's just one of many, many
> examples of the benefits that can often be had by
> splitting notions of id and state in OO programming.
> 
> [...]

The GOF book defines a Singleton rather clearly as a
*unique instance* of a class (while it does not exclude
that a limited number > 1 of instances might also be
useful, sometimes). IMHO, a unique identity is part of
the pattern as you will often have the case where the
*usage* of a Singleton, i.e. the time windows in which
it is used by multiple clients do overlap. In fact, I
think this is the regular case for this pattern, while
you seem to argue in favour of use-cases where clients
immediately drop their handle to that one or *same*
instance and there is no overlap.

Of course, you're free to implement the pattern in what-
ever way you like, but if, as you seem to do, define your
Singleton pattern as multiple instances sharing behaviour
*and* state, i.e. while allowing for different identities
as well (in the general case) I'm tempted to suggest, may-
be, a different name, like Clone, Replicant or something
similar. A Proxy, at least to me, always means a place-
holder for some *other* object, like "intelligent poin-
ters" or the like. Well, if you add a Decorator to your
pool of Clones you might get a Singleton again, but then,
why would I do that? Doesn't a large pool of Clones mean
a waste of memory?

Anyway, it probably all depends on the the problem you
want to solve and the forces being at work, as well as 
your personal preferences... 

Glad-to-see-more-pattern-threads-again'ly,

Dinu

-- 
Dinu C. Gherman
................................................................
"The only possible values [for quality] are 'excellent' and 'in-
sanely excellent', depending on whether lives are at stake or 
not. Otherwise you don't enjoy your work, you don't work well, 
and the project goes down the drain." 
                    (Kent Beck, "Extreme Programming Explained")


From christophertavares at earthlink.net  Fri Dec 29 14:35:25 2000
From: christophertavares at earthlink.net (Christopher Tavares)
Date: Fri, 29 Dec 2000 19:35:25 GMT
Subject: identifying user from Win9x?
References: <92ibvf$if2$1@nntp6.u.washington.edu>
Message-ID: <1I536.42046$3B5.1723486@newsread2.prod.itd.earthlink.net>

"Frank Miles" <fpm at u.washington.edu> wrote in message
news:92ibvf$if2$1 at nntp6.u.washington.edu...
> I have a set of python routines that are used to coordinate and
> account for various activities in my group.  I would like to be
> able to tailor how these routines work, and what options are first
> presented to each user.
>
> There are ways to determine the user ID in Unix-based systems (os.getuid).
> I haven't yet seen anything comparable for Windows.  Possible source
> of information: it is SOP for each user to login to the (Novell-based)
> network before doing anything else.  Is there some way to extract this
> information from Windows, or failing that, a Novell network/server?
>
> TIA...
>
> -frank
>

The Win32 API call that you want is WNetGetUser. I would expect that it's
exposed from the win32net module in the python win32 extensions. I can't be
sure because I don't have python installed on the machine in front of me at
the moment.

-Chris





From ulf.engstrom at b2b-link.com  Thu Dec  7 04:52:24 2000
From: ulf.engstrom at b2b-link.com (=?iso-8859-1?Q?Ulf_Engstr=F6m?=)
Date: Thu, 7 Dec 2000 09:52:24 -0000
Subject: Cookies seem to disappear
References: <wnCX5.218$4b3.42696@news4.mia>
Message-ID: <013901c06033$6aa3eb60$174b9fc3@no>

The browser will only save the cookie if you specify where the cookie later
will be used, that means, you have to set the Domain of the cookie:
c['state']['Domain'] = 'www.yourplace.com'
Also, not all (old) webservers handle max-age correctly, and if you still
have problems after adding domain to your cookie, maybe just try Expires
instead of max-age.
Cookies explained: http://home.netscape.com/newsref/std/cookie_spec.html
Regards
Ulf

Dennis:
> Hi,
>
> I am trying to use SmartCookie class to store state.  Everything works
> perfectly until I exit my browser.  The next time I try to view the page
> my state has been reset to the default values.  I set the 'max-age'
> value like  so:
>
> c = Cookie.SmartCookie()
> c['state'] = stateDictionary
> c['state']['max-age'] = '3600'
> print c.output()
> print
>
> I've tried Netscape on WinNT and Linux and IE on WinNT, they all exhibit
the
> same behaviour.  The thing that is throwing me is that everything works
until I
> exit my browser.  It is as if I'm not setting the max-age correctly but
> how difficult can that be?
>
> Thanks for your help.
>
> --dennism
> --
> http://www.python.org/mailman/listinfo/python-list




From harisri at my-deja.com  Tue Dec  5 00:51:13 2000
From: harisri at my-deja.com (harisri at my-deja.com)
Date: Tue, 05 Dec 2000 05:51:13 GMT
Subject: Newbie question about 'cmpfiles'
Message-ID: <90hvof$crl$1@nnrp1.deja.com>

Hello,

I am a newbie user to python 2.0

Can someone explain 'filecmp' module - 'cmpfiles' function with a simple
example?

I am trying to compare files under 2 different directories, but unable to use
this function.

The 'Python Reference Library' does explain the function, but I am unable to
understand the syntax. What does "common" mean in "cmpfiles(dir1, dir2,
common)"?

Thanks in advance,
Hari.


Sent via Deja.com http://www.deja.com/
Before you buy.


From aahz at panix.com  Sun Dec 17 09:45:36 2000
From: aahz at panix.com (Aahz Maruch)
Date: 17 Dec 2000 06:45:36 -0800
Subject: Alarms in threads - how to get/mimic?
References: <Pine.LNX.4.21.0012161749380.7913-100000@traal.eecs.berkeley.edu>
Message-ID: <91ijig$50b$1@panix6.panix.com>

In article <Pine.LNX.4.21.0012161749380.7913-100000 at traal.eecs.berkeley.edu>,
Luca de Alfaro  <dealfaro at eecs.berkeley.edu> wrote:
>
>What is not clear to me is the following.  When there is a timeout,
>the exception Timeout is not caught by httplib.  Hence, if the thread
>that called httplib just continues its job giving up on the
>connection, httplib will not have a chance to do any clean-up, such as
>for example releasing/closing the socket. 
>
>Hence, it seems to me that in order to abort a web page load cleanly
>upon timeout, I also have to modify the httplib module to handle the
>exception Timeout, and do clean-up work if that happens. 

It depends what you mean by "cleanly".  You don't need to worry about
the loose socket taking up memory, generally speaking; objects clean
themselves up when they go out of scope (Python's exception model does
scope-based cleanup (unlike C++'s, if I understand C++ correctly).  In
addition, if you're using httplib directly (instead of urllib), it's
fairly easy to go poking around in the internals of HTTP() if you take a
reference to it in your code before trying to do a retrieval (the object
won't go out out of scope as long as there's a reference to it).
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

"Using LOCs as a measure of programmer productivity is like estimating the 
productivity of a welder by the amount of acetylene used."  


From fgleeson at my-deja.com  Sat Dec 30 15:30:02 2000
From: fgleeson at my-deja.com (fgleeson at my-deja.com)
Date: Sat, 30 Dec 2000 20:30:02 GMT
Subject: Komodo
References: <B672A66F.72CC%bobhicks@adelphia.net>
Message-ID: <92lgk5$qeq$1@nnrp1.deja.com>

In article <B672A66F.72CC%bobhicks at adelphia.net>,
  Robert L Hicks <bobhicks at adelphia.net> wrote:
> Has anyone tried Komodo from ActiveState yet? What are you thoughts on it? I
> see that a Linux beta is out now as well.
>
> Will Komodo for Linux have to have the ActiveState versions of Perl and
> Python to work?
>
> Bob

You can also debug perl (and python?) on linux/unix with ddd (Data Display
Debugger) the command line is ddd -perl yourscript.pl -youroptions

Incidentally, how are we supposed to pass in command line arguments to our
perl scripts? Is there anyway to do that on the local machine?


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


From removethis_kranki2 at earthlink.net  Fri Dec 22 08:44:23 2000
From: removethis_kranki2 at earthlink.net (Bob W.)
Date: Fri, 22 Dec 2000 13:44:23 GMT
Subject: Newbie Question
Message-ID: <05m64tkqev0aavea80ttjse1sj2gd447hg@4ax.com>

I am trying to write the following class, but it fails in the Run()
method reference to __CheckFile and I am not sure how to get it to
work.  Thanks in advance for your help!

class RmwExpireCheck:
    def __init__( self, szFileName, szDir ):
        self.szFileName = szFileName
        FileStatList = os.stat( szFileName )
        if FileStatList:
            self.iCheckDate = FileStatList[8]
        self.szDir = szDir
        self.szFileList = []

    def __getitem__(self, i):
        line = self.szFileList[i]
        if line:
            return line
        else:
            raise IndexError

    def __CheckFile( self, szDirName, szFileNames ):
        for szFileName in szFileNames:
            szPathName = os.path.join( szDirName, szFileName )
            if os.path.isfile( szPathName ):
                iFileTime = os.stat( szPathName )[8]
                if iFileTime > self.iCheckDate:
                    self.szFileList.append( szPathName )
        
    def Run( self ):
        os.path.walk( self.szDir, RmwExpireCheck.__CheckFile, self )
        




From syver at NOSPAMcyberwatcher.com  Fri Dec  8 06:58:53 2000
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Fri, 8 Dec 2000 12:58:53 +0100
Subject: SSL how to?
Message-ID: <yZ3Y5.3016$9B.89007@news1.oke.nextra.no>

Anyone know if SSL is painless to use under python. Does httplib support it?




From nospam at caltech.edu  Fri Dec  8 03:39:42 2000
From: nospam at caltech.edu (Nathaniel Gray)
Date: Fri, 08 Dec 2000 00:39:42 -0800
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu> <90msh0$rca$1@panix2.panix.com>
Message-ID: <3A309E4E.A8047C5@caltech.edu>

Aahz Maruch wrote:
> 
> In article <3A2EE2AF.AD8A51E2 at caltech.edu>,
> Nathan Gray  <nospam at caltech.edu> wrote:
> >
> >What a rude response to a polite and legitimate question.
> 
> Well, you're one-third right on your adjectives.  Rude, yes; polite and
> legitimate, no.  Perhaps you haven't been on Usenet long enough (or in
> the wrong places) to see the hordes of students who expect other people
> to do their work for them.  Laziness of the highest order, is what it
> is, and many long-term Usenetters see it as rudeness.

I'll try not to interpret your presumption that I'm a babe in the woods
as 
rudeness. <wink>  I know the difference between a legitimate post and a
troll.
Let's have another look at the post, shall we?

"I am doing some research and I would be grateful if anybody could give
me 
some history on the origins of Python."

Is it well thought out?  No.  Is it highly specific?  No.  Lazy?  Yeah, 
sure.  But rude?  With all due respect, I think that's a bit of a
stretch.  I
also find it a bit more of a stretch, from this single sentence, to jump
to the
conclusion that the poster is asking us to do his homework for him.  The
only
word that even hints of schoolwork is "research," and that's one that
most of
us use all the time in all sorts of contexts.

Maybe I should clarify: the reason I found it a legitimate question is
that the 
poster was asking about the _history_ of the language, not the syntax or
the 
features or whatnot.  I can imagine a newbie not knowing that this info
would 
be found on the website.

The reason I found it to be a polite question is that it was, well, not
rude.

> I'm sure you find it quite shocking to discover that rudeness is the
> frequent response to rudeness on Usenet.

Yes, this is why I was so shockingly polite in my reply to Mr. May.
<wink>

I find it shocking that somebody who asks for information about the
history
of the language is flamed, but the guy who posts "Python: what all the
hype
is about?" asking why he should choose Python over Java is given a curt
but
polite and informative response.

I'm sure I find that c.l.p is better than the rest of Usenet and I'd
like
to see it stay that way.

-n8
n8gray <at> caltech <dot> edu
-- 
                          Nathaniel Gray
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
                   California Institute of Technology                    
                     Computation and Neural Systems                      
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._


From harisri at bigpond.com  Wed Dec  6 05:05:31 2000
From: harisri at bigpond.com (Srihari Vijayaraghavan)
Date: Wed, 6 Dec 2000 21:05:31 +1100
Subject: Newbie question on 'cmpfiles':
References: <90i074$d73$1@nnrp1.deja.com> <JmeX5.39553$II2.3513820@newsread2.prod.itd.earthlink.net>
Message-ID: <sZnX5.153284$e5.99945@newsfeeds.bigpond.com>

Oldayz wrote:

> On Tue, 05 Dec 2000 05:59:02 GMT, harisri at my-deja.com
> <harisri at my-deja.com> wrote:
> >Hello,
> >
> >I am a newbie user to python 2.0
> >
> >Can someone help me by explaining 'cmpfiles' function with a simple
> >example?
> >
> >I am trying to compare 2 directories, but not sure about the syntaxof
> >this function. The "Python Reference Library" did not help me on this.
> >
> >[What does "common" mean in "cmpfiles(dir1, dir2, common)"?]
> 
> This function is used to compare *files*, not directories. Multiple files
> in these directories.
> 
> To compare directories, use dircmp class:
> 
> dircmp (a, b[, ignore[, hide]])
> Construct a new directory comparison object, to compare the
> directories a and b. ignore is a list of names to ignore, and
> defaults to ['RCS', 'CVS', 'tags']. hide is a list of names to
> hid, and defaults to [os.curdir, os.pardir].
> 
> report ()
> Print (to sys.stdout) a comparison between a and b.
> 
> etc.. It's in 2.0 lib ref under filecmp
> 
> >
> >Thanks in advance,
> >Hari.
> >
> >
> >Sent via Deja.com http://www.deja.com/
> >Before you buy.
> 
> 
Hello Andrei,

Thanks. However I am having some problem while using "dircmp" class.

I have the following 2 (identical) directories under my home directory: 
/home/hari/python1 and /home/hari/python2. 

I am getting few error messages, only while using the following attributes: 
diff_files, report(), report_partial_closure(), and report_full_closure().

I am enclosing the error messages for your reference. Am I doing something 
wrong?

Python 2.0 (#1, Oct 16 2000, 18:10:03)
[GCC 2.95.2 19991024 (release)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> from filecmp import *
>>> x=dircmp("/home/hari/python1", "/home/hari/python2")
>>> x.left_list
['about.html', 'acks.html', 'api', 'dist', 'doc', 'ext', 'icons', 
'index.html',
'inst', 'lib', 'mac', 'modindex.html', 'ref', 'tut']
>>> x.right_list
['about.html', 'acks.html', 'api', 'dist', 'doc', 'ext', 'icons', 
'index.html',
'inst', 'lib', 'mac', 'modindex.html', 'ref', 'tut']
>>> x.common
['ref', 'ext', 'acks.html', 'inst', 'api', 'mac', 'about.html', 
'modindex.html', 'tut', 'doc', 'icons', 'lib', 'dist', 'index.html']
>>> x.left_only
[]
>>> x.right_only
[]
>>> x.common_dirs
['ref', 'ext', 'inst', 'api', 'mac', 'tut', 'doc', 'icons', 'lib', 'dist']
>>> x.common_files
['acks.html', 'about.html', 'modindex.html', 'index.html']
>>> x.same_files
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/local/lib/python2.0/filecmp.py", line 147, in __getattr__
    self.phase3()
  File "/usr/local/lib/python2.0/filecmp.py", line 214, in phase3
    xx = cmpfiles(self.left, self.right, self.common_files)
  File "/usr/local/lib/python2.0/filecmp.py", line 288, in cmpfiles
    res[_cmp(ax, bx, shallow, use_statcache)].append(x)
TypeError: too many arguments; expected 2, got 4                            
>>> x.report()
diff /home/hari/python1 /home/hari/python2
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/local/lib/python2.0/filecmp.py", line 241, in report
    if self.same_files:
  File "/usr/local/lib/python2.0/filecmp.py", line 147, in __getattr__
    self.phase3()
  File "/usr/local/lib/python2.0/filecmp.py", line 214, in phase3
    xx = cmpfiles(self.left, self.right, self.common_files)
  File "/usr/local/lib/python2.0/filecmp.py", line 288, in cmpfiles
    res[_cmp(ax, bx, shallow, use_statcache)].append(x)
TypeError: too many arguments; expected 2, got 4                            
 >>> x.report_partial_closure()
diff /home/hari/python1 /home/hari/python2
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/local/lib/python2.0/filecmp.py", line 258, in 
report_partial_closure
    self.report()
  File "/usr/local/lib/python2.0/filecmp.py", line 241, in report
    if self.same_files:
  File "/usr/local/lib/python2.0/filecmp.py", line 147, in __getattr__
    self.phase3()
  File "/usr/local/lib/python2.0/filecmp.py", line 214, in phase3
    xx = cmpfiles(self.left, self.right, self.common_files)
  File "/usr/local/lib/python2.0/filecmp.py", line 288, in cmpfiles
    res[_cmp(ax, bx, shallow, use_statcache)].append(x)
TypeError: too many arguments; expected 2, got 4
>>> x.report_full_closure()
diff /home/hari/python1 /home/hari/python2
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/local/lib/python2.0/filecmp.py", line 264, in 
report_full_closure
    self.report()
  File "/usr/local/lib/python2.0/filecmp.py", line 241, in report
    if self.same_files:
  File "/usr/local/lib/python2.0/filecmp.py", line 147, in __getattr__
    self.phase3()
  File "/usr/local/lib/python2.0/filecmp.py", line 214, in phase3
    xx = cmpfiles(self.left, self.right, self.common_files)
  File "/usr/local/lib/python2.0/filecmp.py", line 288, in cmpfiles
    res[_cmp(ax, bx, shallow, use_statcache)].append(x)
TypeError: too many arguments; expected 2, got 4                            
    
Thanks in advance,

Hari.      


From fredrik at effbot.org  Thu Dec  7 15:25:49 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Thu, 07 Dec 2000 20:25:49 GMT
Subject: Simple HTTP client bug?
References: <3A2FEE2A.D1A0A645@raqia.com>
Message-ID: <hnSX5.2109$Kd1.301769@newsb.telia.net>

David Lees wrote:
> I am trying to run the following simple HTTP client code, copied
> directly from "Programming in Python with Medusa and the Async Sockets
> Library"

just curious, but what's Programming in Python with Medusa
and the Async Sockets Library"?

as for your question, the answer is found in the debugging
print-out:

> starting asyncore loop
> sending GET
> GET  HTTP/1.0

"GET  HTTP/1.0" is definitely not a valid HTTP command.
you're supposed to put an URL between the GET and the
HTTP/1.0.

either make sure you explicitly specify a path name (not
just a host name), or add a test to the URL parsing code:

    host = parts[1]
    path = parts[2]
    if not path:
        path = "/" # don't call with an empty path
    print 'host,path=',host,path

</F>




From katz at Glue.umd.edu  Tue Dec 12 22:54:42 2000
From: katz at Glue.umd.edu (Roy Katz)
Date: Tue, 12 Dec 2000 22:54:42 -0500
Subject: Dijkstra's Shortest Path algorithm
In-Reply-To: <dvrd3t40u9l7atgdro59m5ttg7dsfv7c4i@4ax.com>
References: <Pine.GSO.4.21.0012090042370.15110-100000@y.glue.umd.edu> <dvrd3t40u9l7atgdro59m5ttg7dsfv7c4i@4ax.com>
Message-ID: <Pine.GSO.4.21.0012122249480.19682-100000@y.glue.umd.edu>

On Wed, 13 Dec 2000, Andy Robinson wrote:

> Yes, there is.
> 
> Aaron Watters kjBuckets collections library include a graph type
> written in C, which has this and several other useful algorithms.
> www.chordate.com
> 
> - Andy Robinson
> 

Wonderful! That's important for my sanity. 
But that sort of takes the fun out of making something people will
use. Unless something else needs to be made? Is there anything that people
need implemented? I'm finally ending my semester next week, so it'd be
great to get back into serious Python programming (as opposed to this
semester's C++). 


Roey


btw -- we're holding a competition for the fastest program in my C++
class.  I'm probably in the top four in my section (1.77 seconds for 500
nodes with 10k edges, 1 query). The average for my class is around 20
seconds (this was the final project).  Weird! the biggest lags occur in
reading in the data! but I guess, that's what we get with
object orientation -- more layers to tunnel through. 



From cg at schlund.de  Fri Dec 15 04:08:25 2000
From: cg at schlund.de (Carsten Gaebler)
Date: Fri, 15 Dec 2000 10:08:25 +0100
Subject: Compiling Python 2.0 with Berkeley DB 3.0.55?
Message-ID: <3A39DF89.D76B6415@schlund.de>

Hi there!

In our company we have a Linux server with the Berkeley DB 3.0.55
installed and some DB-3.0.55 files. I compiled Python 2.0 out of the box
on that machine and the dbhash module worked fine for the DB files. But
on my own workstation I get the error message "Invalid argument" from
dbhash.open() when I try to open a DB-3.0.55 file. I compiled Berkeley
DB 3.0.55 and replaced the db*.h files in /usr/include/ and libdb.a in
/usr/lib/ but it just won't work. Any hints?


Regards
Carsten.


From sh at ttsoftware.co.uk  Thu Dec  7 07:20:39 2000
From: sh at ttsoftware.co.uk (Steve Horne)
Date: Thu, 07 Dec 2000 12:20:39 +0000
Subject: Schwartzian Transform (was Re: Sort)
References: <NlnX5.1444$yv1.2700363@nnrp5.proxad.net> <oa8s2tsls5foll62r8fpa6nlcpisb6v1ek@4ax.com> <90l9ir02avk@news2.newsguy.com> <3A2E5F2A.9E650D50@erols.com> <90lrhl06t@news2.newsguy.com> <Pine.NXT.4.21.0012061411450.230-100000@localhost.virginia.edu>
Message-ID: <2cuu2t8gkrjhfb5ii54h7ummhr84lpebbo@4ax.com>

On Wed, 6 Dec 2000 14:16:53 -0500, "Steven D. Majewski"
<sdm7g at Virginia.EDU> wrote:

>On Wed, 6 Dec 2000, Alex Martelli wrote:
>
>> The technique is named after its inventor Randall
>> Schwartz, who thought it up in the context of Perl
>> (but it works even better in Python:-).

>This is however, a pretty ancient Python idiom, so I wonder
>about attributing it to Randall's invention. 

Even older still, I think. I'd have to double check, but I swear
similar tricks were used decades ago in normal 3GLs, to cope with
library sort functions that didn't take parameters. Certainly I used
it in 1988, at least 8 years before I ever heard of Perl or Python.

Some people somehow seem to get their names associated with
restatements of the obvious.

I have to say I'm a little surprised by the scale of the speedup, but
even so I'd usually not bother. The lambda method is clearer, and
works just as well unless you're sorting a lot of data. If I was
worried about the difference between 0.0353952707969 and
0.220513718779 seconds, I wouldn't be using a scripting language.

-- 
Steve Horne
Home : steve at lurking.demon.co.uk
Work : sh at ttsoftware.co.uk


From boud at rempt.xs4all.nl  Sat Dec 30 03:59:36 2000
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 30 Dec 2000 08:59:36 GMT
Subject: Python for prototyping of the user interface
References: <20001228.17165633@mis.configured.host> <92ir54$ftn$1@news1.xs4all.nl> <92j6hd$k82$1@nntp6.u.washington.edu>
Message-ID: <92k85o$ke3$1@news1.xs4all.nl>

Frank Miles <fpm at u.washington.edu> wrote:

> Python and Glade seem to offer these features, though perhaps without
> the same degree of polish.  Qt appears to have a significant cost
> if you're going to run on Windows :( [at least that's my reading of
> TrollTech's FAQ].


Not if you're going to use Python and Qt: since PyQt is just an
application using Qt, the developer can make it available for free. You
don't need a developers license for Qt to develop Python apps with
it. Phil Thompson got his copy of Qt from Troll Tech for free (after I
badgered Trolltech a bit), together with a Windows compiler and another
hd, so he has produced a PyQt for windows, which can be downloaded at:
 
 http://www.thekompany.com/projects/pykde/
  
It's quite easy to install, even for novice users of your
application.  And, as I said, it's complete - designer, ui
compiler and runtime environment. And all documentation, too.

--

Boudewijn Rempt  | http://www.valdyas.org


From jeremy at alum.mit.edu  Fri Dec 15 12:39:11 2000
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Fri, 15 Dec 2000 17:39:11 GMT
Subject: what is a closure?
References: <3a39033b@news.xtml.co.uk> <S18_5.3929$Kd1.455131@newsb.telia.net> <8nck3tocsg8kp1u2g1du6nd3lt9gfeckr2@4ax.com>
Message-ID: <91dkvu$7mk$1@nnrp1.deja.com>

In article <8nck3tocsg8kp1u2g1du6nd3lt9gfeckr2 at 4ax.com>,
  Steve Horne <sh at ttsoftware.co.uk> wrote:
> On Thu, 14 Dec 2000 18:09:54 GMT, "Fredrik Lundh" <fredrik at effbot.org>
> wrote:
>
> >In the following example, the "a=a" default argument binding
> >binds add's local name "a" to the value of the "a" argument
> >from make_adder.
> >
> >    def make_adder(a):
> >        def add(b, a=a):
> >            return a + b
> >        return add
> >
> >    >>> m = make_adder(10)
> >    >>> m(20)
> >    30
>
> Are you sure that's a closure? - It looks closer to currying to me.
> And the definition seemed pretty close to the definition of an object
> - a collection of related data and methods.
>
> Currying, in functional languages, is the process of specialising a
> function by specifying one or more of its parameters to produce a
> function with fewer parameters than the original. The notation in
> Python is OK, but not very convenient...
[...]
> I don't claim to be an expert on this. I would simply appreciate it if
> someone could clear up my confusion!

A closure provides a way to resolve free variables in a function body;
it is created when the function is defined. Currying is a way to create
a binding for a formal parameter; it is used when a function is called

The use of default arguments, strictly speaking, isn't really closure
creation and isn't really currying.  It's similar to both.

In Fredrik's example, a=a is rather like a manually created closure.  If
you ignore the a=a part of the parameter list, then a is free in adder.
A closure would provide an environment in which to resolve a.

The use of default arguments in the parameter list works mostly like a
closure.  The one difference is that if a is rebound in make_adder()
after adder() is defined, adder() will still see the old value of a.

Currying occurs when a function is called.  An example using currying
would look something like this:

def make_adder(x):
    def adder(x, y):
        return x + y
    return adder(x)  # returns a function that takes one argument, y
m = make_adder(10)
m(20) == 30

In this example, make_adder() returns a curried function.  The currying
occurs in the call to adder() in the return statement, not during the
definition of adder().  It's interesting to note, though, that default
arguments and currying both cause a formal parameter to be bound to a
value.  They share the property that rebinding the name of the
right-hand side has no effect on the binding in adder().

Incidentally, I am working on a proposal to add real closure creation to
Python 2.1.  I'm working on the implementation and waiting for approval
(or rejection) from the BDFL.  The current text of the proposal is at
    http://python.sourceforge.net/peps/pep-0227.html
Comments are welcome.

-- Jeremy Hylton, <http://www.python.org/~jeremy/>


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


From aleaxit at yahoo.com  Tue Dec  5 03:17:12 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 5 Dec 2000 09:17:12 +0100
Subject: Using dictionaries
References: <3a2c9264.503856587@news-server.bigpond.net.au>
Message-ID: <90i8a001u10@news2.newsguy.com>

"sorular" <sorular at netscape.net> wrote in message
news:3a2c9264.503856587 at news-server.bigpond.net.au...
> G'day
>
> I am trying to write a python program where a "configuration.py" will
> have all the user entered values in a dictionary format and a
> "main.py" which will import these and use it as local parameters.
>
> But, there are quite a few values in the dictionary
> ie
> dict = { 'key1':'value1' , ' key2':'value2' , ......'keyN':'value2'}

So far, so good.  Some of the values could also be added
with a dict['otherkey']='othervalue' syntax, of course.  When
you "import configuration", both the dict's original definition
and the further additions/changes will be executed in order,
so we can assume that right after that configuration.dict has
all of the settings you require.


> and in the "main.py" these will be all renamed with the local names
> ie
> Val1 = dictionary.dict['key1']
> Val2 = dictionary.dict['key2']
> ....
> ValN = dictionary.dict['keyN']
> etc....

First, I assume you mean configuration.dict, unless for some
peculiar reason you've chosen to "import configuration as dictionary"
(but why would you do that?).

> But overall, I still have a large numbe of values (ValN) to list... !

What is the relationship, if any, between the names of the
variables such as Val2 and the names of the keys such as key2?

If there is no systematic relationship, then you do have to
list the whole correspondence one way or another, of course;
and the long series of assignments is one such way (or you
could have another dictionary, or list or tuple of pairs, etc,
etc, and iterate over that).

If the correspondence is systematic, such as the one that
appears to be implicit in the naming you have used above
(each key is named 'key'+suffix for some string suffix, and
the variable that must be set like that is named 'Val'+suffix),
then you have a much easier life, e.g:

    for keyname, keyvalue in configuration.dict.items():
        if keyname.startswith('key'):
            __dict__['Val'+keyname[3:]] = keyvalue

It's up to you what you want to do (if anything) with
dictionary keys that do NOT start with 'key' -- just
ignore them, print a warning, throw an exception, whatever.

> Q1- Is there a more efficient way to do this ? I thought about using a
> for loop but that didnt work....

The above for-loop will certainly work if the relationship
between keynames and variable names is as I posit it.  If
it's otherwise, the only difference is on how you compute
the variablename to be set in correspondence to a given key.


> Q2 - would it be possible only to list the parameters in first file
> (ie dictionary.py
> key1 = value1
> key2 = value2
> .....
> keyN = valueN
> )

Sure (but is it configuration.py as you said at the start,
which would seem a much better name, or dictionary.py as you
say now and would seem to imply by your other usages?) --
this way, you're setting up the configuration.__dict__
(or dictionary.__dict__, whatever) dictionary.  I.e., if
your module is this way, you can use the same loop as
above, just change the for-statement itself to:

    for keyname, keyvalue in configuration.__dict__.items():
        # body does not have to change!

> and inside main.py, do a simple import...
> import dictionary *
> so that key1 can be used inside main.py as a local parameter?

Local variables are another issue, but I think you mean
"module-global" here anyway.  Sure, if the correspondence
between keyname and variablename is simple identity, this
will also work.  But a renaming is often advisable, just
to avoid accidental clashes (where the user, editing
configuration.py, happens by mistake to set a variable
whose name equals some other identifier used in main.py),
and it's not much harder to perform (as long as it's in
some way _systematic_, not haphazard) than just sucking
everything in.

If you've not decided on a naming convention for keys in
configuration.py versus variables in main.py, I suggest
you choose a prefix (and ensure no other identifiers in
main.py start with that prefix), such as, say, conf_, and
do something like:
    for keyname, keyvalue in configuration.__dict__.items():
        if not keyname.startswith(prefix):
            keyname = prefix+keyname
        __dict__[keyname] = keyvalue

This lets the user write lines in configuration.py such
as:
    foo = 'bar'        # sets conf_foo for main.py
    conf_fee = 'fie'   # sets conf_fee ditto
and it's _almost_ equivalent to
    import configuration as conf
then using conf.foo, conf.fee, etc, instead of the
suggested conf_foo, conf_fee, etc; but it makes it
slightly easier to provide defaults in main.py for
configuration variables that the user has not set
otherwise in configuration.py, i.e. main.py could
start with something like:

conf_foo = 'a default for conf_foo'
conf_fee = 'another important default'
conf_fie = conf_fum = conf_bar = None    # common default
import configuration
# insert loop as above

then use the various conf_fie, conf_fee, etc etc, as
desired; they'd have their default values as set
before the import unless the user overrode them in
configuration.py.

If you'd rather use the conf.foo, conf.fee, etc,
syntax for this same purpose (and it would be
highly advisable in many ways most of the time),
it's also very easy to do so, for example:

class conf:
    foo = 'a default for conf.foo'
    fee = 'another important default'
    fie = fum = bar = None    # common default
    from configuration import *

assuming configuration.py only uses the plain
name and not conf_whatever.  Or, the class might
also easily use a finer degree of control with
an "import configuration", then a loop like the
above-exemplified ones.

Or you could use an instance object rather than
a class object, supposing for example you need
ALL conf.whatever variable to be None unless
explicitly set, the easiest way might be:

class Conf:
    foo = 'a default for conf.foo'
    fee = 'another important default'
    from configuration import *
    def __getattr__(self, name):
        return None
conf = Conf()

Now conf.foo and conf.fee are set by default as
above but can be overridden by lines such as
"foo = 'something else'" in configuration.py,
while ANY other conf.whatever is None unless set
in configuration.py.


As you can see, Python offers a lot of similar but not
identical approaches for the task you've set yourself.
It's not part of the party line, but, in Python, too,
there's more than one way to do it.  How to choose?  A
golden rule of programming is...:

    *Do the simplest thing that could possibly work*

Being aware of the nicest exotic possibilities is cool,
but don't use them unless the defects in SIMPLER ways
are actually giving you problems.  In other words,

    from configuration import *

might be the best solution (because it's simplest),
unless accidental name-clashes ARE actually a problem
you need to address, etc, etc.  (IMHO, the second
simplest approach is that of making conf a class
object, as above; the third-simplest, if a "default
default value" is actually useful in many cases, is
making conf an instance object, again as above; the
various loops explicitly manipulating __dict__ I see
as a little bit less simple, thus I would not tend
to use them in most cases -- unless I get convinced
that some simpler approach "can't possibly work" under
the specific conditions of a given problem).


Alex





From fredrik at effbot.org  Fri Dec 29 15:36:24 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 29 Dec 2000 20:36:24 GMT
Subject: tkinter and images - again
References: <slrn94o48o.a5p.jp@localhost.localdomain>
Message-ID: <cB636.2483$Qb7.271793@newsb.telia.net>

Jacek Pop?awski wrote:
> I played a while with Tkinter and images. Why can't I use jpeg and png
> formats? Is it any way to read and show this kind of files in Tkinter?

> I played a while with Tkinter and images. Why can't I use jpeg and png
> formats? Is it any way to read and show this kind of files in Tkinter?

PIL's ImageTk module lets you display 30+ formats
with Tkinter:

http://www.pythonware.com/products/pil/index.htm

</F>




From cfelling at iae.nl  Fri Dec 15 08:28:10 2000
From: cfelling at iae.nl (Carel Fellinger)
Date: 15 Dec 2000 14:28:10 +0100
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <915jcf0c6o@news2.newsguy.com> <9162p3$7t2$1@animus.fel.iae.nl> <MBWZ5.334$h67.37748@newsread1.prod.itd.earthlink.net> <3A397C74.6C8906E1@cosc.canterbury.ac.nz>
Message-ID: <91d69a$1ov$1@animus.fel.iae.nl>

Greg Ewing <greg at cosc.canterbury.ac.nz> wrote:
> Oldayz wrote:
>> 
>> I thought something like forever: would be even more lucid

> If there is a break in the loop somewhere, it's
> not really forever, though, is it?

> So perhaps it should *really* be

>    for a while:
>       ...

but then again, it might not end, so it really should be spelled as

   while:
      'preferably a break somewhere'

> :-)
me to, but it wouldn't break any code (I think), and to me it makes sence:

  you forgo the loop testing upfront, so you leave it out instead of
  replacing it whith some ugly 'always true' stub-test

Would such a cosmetic change stand any change of being adopted by the BDFL?
-- 
groetjes, carel


From dsh8290 at rit.edu  Sat Dec 30 23:55:31 2000
From: dsh8290 at rit.edu (D-Man)
Date: Sat, 30 Dec 2000 23:55:31 -0500
Subject: Question re: eval()
In-Reply-To: <978109654.339714327@news.silcom.com>; from clarence@netlojix.com on Fri, Dec 29, 2000 at 09:03:49AM -0800
References: <978109654.339714327@news.silcom.com>
Message-ID: <20001230235531.E19500@westsidecnc.rh.rit.edu>

On Fri, Dec 29, 2000 at 09:03:49AM -0800, Clarence Gardner wrote:
> 
<snip>
> class D:
<snip>
> 
> d = D()
> x = eval('n + 1', d)
> 
> 
> but the eval function accepts only a dictionary as the second parameter.
> Is there any amazing way to do this?

Make D a dictionary.  Have it inherit from UserDictionary (I don't
know which module it is in) and implement the necessary lookup
functions.

> 
> TIA,
> Clarence Gardner
> clarence at netlojix.com

HTH, 
-D



From etsang at my-deja.com  Wed Dec  6 15:27:36 2000
From: etsang at my-deja.com (etsang at my-deja.com)
Date: Wed, 06 Dec 2000 20:27:36 GMT
Subject: NewBie question: cannot import TKinter and os
Message-ID: <90m7fc$r6m$1@nnrp1.deja.com>

Hi,

I have build python2.0 with expect, tcl/tk. But Ifaced the folliwng
problems:
I have uncommented out the tcl/tk + resource in Setup, complilation is
okay. When I run python I run into:
1. at the very first beginning, it says import site error. I know this
module site is no longer required in Python 2.0 after dereading some
doc in pytho.org. how can I make thie error go away??
2. when I tried import Tkinter, it says No Module named TkInter, but it
is clearly compiled... I did the same thing in versio 1.5.2 and there
was no problem .....
3. when I try import expect, it says
import expect # builtin
WARNING: Python C API version mismatch for module expect:
  This Python has API version 1009, module expect has version 1007.
Where can I get the correct 1009 version for expect????

Thanks a million!!!
Attached belwo is the Setup file I used, hope this provided more info...



# -*- makefile -*-
# The file Setup is used by the makesetup script to construct the files
# Makefile and config.c, from Makefile.pre and config.c.in,
# respectively.  The file Setup itself is initially copied from
# Setup.in; once it exists it will not be overwritten, so you can edit
# Setup to your heart's content.  Note that Makefile.pre is created
# from Makefile.pre.in by the toplevel configure script.

# (VPATH notes: Setup and Makefile.pre are in the build directory, as
# are Makefile and config.c; the *.in files are in the source
# directory.)

# Each line in this file describes one or more optional modules.
# Comment out lines to suppress modules.
# Lines have the following structure:
#
# <module> ... [<sourcefile> ...] [<cpparg> ...] [<library> ...]
#
# <sourcefile> is anything ending in .c (.C, .cc, .c++ are C++ files)
# <cpparg> is anything starting with -I, -D, -U or -C
# <library> is anything ending in .a or beginning with -l or -L
# <module> is anything else but should be a valid Python
# identifier (letters, digits, underscores, beginning with non-digit)
#
# (As the makesetup script changes, it may recognize some other
# arguments as well, e.g. *.so and *.sl as libraries.  See the big
# case statement in the makesetup script.)
#
# Lines can also have the form
#
# <name> = <value>
#
# which defines a Make variable definition inserted into Makefile.in
#
# Finally, if a line contains just the word "*shared*" (without the
# quotes but with the stars), then the following modules will not be
# included in the config.c file, nor in the list of objects to be
# added to the library archive, and their linker options won't be
# added to the linker options, but rules to create their .o files and
# their shared libraries will still be added to the Makefile, and
# their names will be collected in the Make variable SHAREDMODS.  This
# is used to build modules as shared libraries.  (They can be
# installed using "make sharedinstall", which is implied by the
# toplevel "make install" target.)  (For compatibility,
# *noconfig* has the same effect as *shared*.)
#
# In addition, *static* reverses this effect (negating a previous
# *shared* line).

# NOTE: As a standard policy, as many modules as can be supported by a
# platform should be present.  The distribution comes with all modules
# enabled that are supported by most platforms and don't require you
# to ftp sources from elsewhere.


# Some special rules to define PYTHONPATH.
# Edit the definitions below to indicate which options you are using.
# Don't add any whitespace or comments!

# Directories where library files get installed.
# DESTLIB is for Python modules; MACHDESTLIB for shared libraries.
DESTLIB=$(LIBDEST)
MACHDESTLIB=$(BINLIBDEST)

# NOTE: all the paths are now relative to the prefix that is computed
# at run time!

# Standard path -- don't edit.
# No leading colon since this is the first entry.
# Empty since this is now just the runtime prefix.
DESTPATH=

# Site specific path components -- should begin with : if non-empty
SITEPATH=

# Standard path components for test modules
TESTPATH=

# Path components for machine- or system-dependent modules and shared
libraries
MACHDEPPATH=:plat-$(MACHDEP)

COREPYTHONPATH=$(DESTPATH)$(SITEPATH)$(TESTPATH)$(MACHDEPPATH)$(TKPATH)
PYTHONPATH=$(COREPYTHONPATH)


# The modules listed here can't be built as shared libraries for
# various reasons; therefore they are listed here instead of in the
# normal order.

# Some modules that are normally always on:

regex regexmodule.c regexpr.c	# Regular expressions, GNU Emacs style
pcre pcremodule.c pypcre.c	# Regular expressions, Perl style (for
re.py)
posix posixmodule.c		# posix (UNIX) system calls
signal signalmodule.c		# signal(2)
_sre _sre.c			# Fredrik Lundh's new regular
expressions

# The SGI specific GL module:

GLHACK=-Dclear=__GLclear
#gl glmodule.c cgensupport.c -I$(srcdir) $(GLHACK) -lgl -lX11

# Pure module.  Cannot be linked dynamically.
# -DWITH_QUANTIFY, -DWITH_PURIFY, or -DWITH_ALL_PURE
#WHICH_PURE_PRODUCTS=-DWITH_ALL_PURE
#PURE_INCLS=-I/usr/local/include
#PURE_STUBLIBS=-L/usr/local/lib -lpurify_stubs -lquantify_stubs
#pure puremodule.c $(WHICH_PURE_PRODUCTS) $(PURE_INCLS) $(PURE_STUBLIBS)

# Uncommenting the following line tells makesetup that all following
# modules are to be built as shared libraries (see above for more
# detail; also note that *static* reverses this effect):

#*shared*

# GNU readline.  Unlike previous Python incarnations, GNU readline is
# now incorporated in an optional module, configured in the Setup file
# instead of by a configure script switch.  You may have to insert a
# -L option pointing to the directory where libreadline.* lives,
# and you may have to change -ltermcap to -ltermlib or perhaps remove
# it, depending on your system -- see the GNU readline instructions.
# It's okay for this to be a shared library, too.

#readline readline.c -lreadline -ltermcap


# Modules that should always be present (non UNIX dependent):

array arraymodule.c	# array objects
cmath cmathmodule.c # -lm # complex math library functions
math mathmodule.c # -lm # math library functions, e.g. sin()
strop stropmodule.c	# fast string operations implemented in C
struct structmodule.c	# binary structure packing/unpacking
time timemodule.c # -lm # time operations and variables
operator operator.c	# operator.add() and similar goodies
_codecs _codecsmodule.c	# access to the builtin codecs and codec
registry

unicodedata unicodedata.c unicodedatabase.c
                        # static Unicode character database
ucnhash ucnhash.c # Unicode Character Name expansion hash table

_locale _localemodule.c  # access to ISO C locale support


# Modules with some UNIX dependencies -- on by default:
# (If you have a really backward UNIX, select and socket may not be
# supported...)

fcntl fcntlmodule.c	# fcntl(2) and ioctl(2)
pwd pwdmodule.c		# pwd(3)
grp grpmodule.c		# grp(3)
errno errnomodule.c	# posix (UNIX) errno values
select selectmodule.c	# select(2); not on ancient System V
_socket socketmodule.c	# socket(2)

# Memory-mapped files (also works on Win32).
mmap mmapmodule.c

# Socket module compiled with SSL support; you must edit the SSL
variable:
#SSL=/usr/local/ssl
#_socket socketmodule.c \
#	-DUSE_SSL -I$(SSL)/include -I$(SSL)/include/openssl \
#	-L$(SSL)/lib -lssl -lcrypto

# The crypt module is now disabled by default because it breaks builds
# on many systems (where -lcrypt is needed), e.g. Linux (I believe).
#crypt cryptmodule.c # -lcrypt	# crypt(3); needs -lcrypt on some
systems


# Some more UNIX dependent modules -- off by default, since these
# are not supported by all UNIX systems:

#nis nismodule.c -lnsl	# Sun yellow pages -- not everywhere
#termios termios.c	# Steen Lumholt's termios module
resource resource.c	# Jeremy Hylton's rlimit interface


# Multimedia modules -- off by default.
# These don't work for 64-bit platforms!!!
# These represent audio samples or images as strings:

#audioop audioop.c	# Operations on audio samples
#imageop imageop.c	# Operations on images
#rgbimg rgbimgmodule.c	# Read SGI RGB image files (but coded portably)


# The md5 module implements the RSA Data Security, Inc. MD5
# Message-Digest Algorithm, described in RFC 1321.  The necessary files
# md5c.c and md5.h are included here.

md5 md5module.c md5c.c


# The sha module implements the SHA checksum algorithm.
# (NIST's Secure Hash Algorithm.)
sha shamodule.c


# The mpz module interfaces to the GNU Multiple Precision library.
# You need to ftp the GNU MP library.
# The GMP variable must point to the GMP source directory.
# This was originally written and tested against GMP 1.2 and 1.3.2.
# It has been modified by Rob Hooft to work with 2.0.2 as well, but I
# haven't tested it recently.

# A compatible MP library unencombered by the GPL also exists.  It was
# posted to comp.sources.misc in volume 40 and is widely available from
# FTP archive sites. One URL for it is:
#
ftp://gatekeeper.dec.com/.b/usenet/comp.sources.misc/volume40/fgmp/part0
1.Z

#GMP=/ufs/guido/src/gmp
#mpz mpzmodule.c -I$(GMP) $(GMP)/libgmp.a


# SGI IRIX specific modules -- off by default.

# These module work on any SGI machine:

# *** gl must be enabled higher up in this file ***
#fm fmmodule.c $(GLHACK) -lfm -lgl		# Font Manager
#sgi sgimodule.c			# sgi.nap() and a few more

# This module requires the header file
# /usr/people/4Dgifts/iristools/include/izoom.h:
#imgfile imgfile.c -limage -lgutil -lgl -lm	# Image Processing
Utilities


# These modules require the Multimedia Development Option (I think):

#al almodule.c -laudio			# Audio Library
#cd cdmodule.c -lcdaudio -lds -lmediad	# CD Audio Library
#cl clmodule.c -lcl -lawareaudio	# Compression Library
#sv svmodule.c yuvconvert.c -lsvideo -lXext -lX11	# Starter Video


# The FORMS library, by Mark Overmars, implements user interface
# components such as dialogs and buttons using SGI's GL and FM
# libraries.  You must ftp the FORMS library separately from
# ftp://ftp.cs.ruu.nl/pub/SGI/FORMS.  It was tested with FORMS 2.2a.
# NOTE: if you want to be able to use FORMS and curses simultaneously
# (or both link them statically into the same binary), you must
# compile all of FORMS with the cc option "-Dclear=__GLclear".

# The FORMS variable must point to the FORMS subdirectory of the forms
# toplevel directory:

#FORMS=/ufs/guido/src/forms/FORMS
#fl flmodule.c -I$(FORMS) $(GLHACK) $(FORMS)/libforms.a -lfm -lgl


# SunOS specific modules -- off by default:

#sunaudiodev sunaudiodev.c


# Linux specific modules -- off by default:

#linuxaudiodev linuxaudiodev.c


# George Neville-Neil's timing module:

#timing timingmodule.c


# The _tkinter module.
#
# The TKPATH variable is always enabled, to save you the effort.
TKPATH=:lib-tk

# The command for _tkinter is long and site specific.  Please
# uncomment and/or edit those parts as indicated.  If you don't have a
# specific extension (e.g. Tix or BLT), leave the corresponding line
# commented out.  (Leave the trailing backslashes in!  If you
# experience strange errors, you may want to join all uncommented
# lines and remove the backslashes -- the backslash interpretation is
# done by the shell's "read" command and it may not be implemented on
# every system.

# *** Always uncomment this (leave the leading underscore in!):
_tkinter _tkinter.c tkappinit.c -DWITH_APPINIT \
# *** Uncomment and edit to reflect where your Tcl/Tk headers are:
	-I/tools/2.5.1/local/include \
# *** Uncomment and edit to reflect where your X11 header files are:
#	-I/usr/X11R6/include \
# *** Or uncomment this for Solaris:
	-I/usr/openwin/include \
# *** Uncomment and edit for Tix extension only:
#	-DWITH_TIX -ltix4.1.8.0 \
# *** Uncomment and edit for BLT extension only:
#	-DWITH_BLT -I/usr/local/blt/blt8.0-unoff/include -lBLT8.0 \
# *** Uncomment and edit for PIL (TkImaging) extension only:
#	-DWITH_PIL -I../Extensions/Imaging/libImaging  tkImaging.c \
# *** Uncomment and edit for TOGL extension only:
#	-DWITH_TOGL togl.c \
# *** Uncomment and edit to reflect where your Tcl/Tk libraries are:
	-L/tools/2.5.1/local/lib \
# *** Uncomment and edit to reflect your Tcl/Tk versions:
	-ltk8.0 -ltcl8.0 \
# *** Uncomment and edit to reflect where your X11 libraries are:
#	-L/usr/X11R6/lib \
# *** Or uncomment this for Solaris:
	-L/usr/openwin/lib \
# *** Uncomment these for TOGL extension only:
#	-lGL -lGLU -lXext -lXmu \
# *** Uncomment for AIX:
#	-lld \
# *** Always uncomment this; X11 libraries to link with:
	-lX11

expect expect.a

# Lance Ellinghaus's modules:

rotor rotormodule.c		# enigma-inspired encryption
syslog syslogmodule.c		# syslog daemon interface


# Curses support, requring the System V version of curses, often
# provided by the ncurses library.  e.g. on Linux, link with -lncurses
# instead of -lcurses; on SunOS 4.1.3, insert -I/usr/5include
# -L/usr/5lib before -lcurses).

#_curses _cursesmodule.c -lcurses -ltermcap



# Tommy Burnette's 'new' module (creates new empty objects of certain
kinds):

new newmodule.c


# Generic (SunOS / SVR4) dynamic loading module.
# This is not needed for dynamic loading of Python modules --
# it is a highly experimental and dangerous device for calling
# *arbitrary* C functions in *arbitrary* shared libraries:

#dl dlmodule.c


# Modules that provide persistent dictionary-like semantics.  You will
# probably want to arrange for at least one of them to be available on
# your machine, though none are defined by default because of library
# dependencies.  The Python module anydbm.py provides an
# implementation independent wrapper for these; dumbdbm.py provides
# similar functionality (but slower of course) implemented in Python.

# The standard Unix dbm module:

#dbm dbmmodule.c 	# dbm(3) may require -lndbm or similar

# Anthony Baxter's gdbm module.  GNU dbm(3) will require -lgdbm:

#gdbm gdbmmodule.c -I/usr/local/include -L/usr/local/lib -lgdbm


# Berkeley DB interface.
#
# This requires the Berkeley DB code, see
# ftp://ftp.cs.berkeley.edu/pub/4bsd/db.1.85.tar.gz
#
# Edit the variables DB and DBPORT to point to the db top directory
# and the subdirectory of PORT where you built it.
#
# (See http://electricrain.com/greg/python/bsddb3/ for an interface to
# BSD DB 3.x.)

# Note: If a db.h file is found by configure, bsddb will be enabled
# automatically via Setup.config.in.  It only needs to be enabled here
# if it is not automatically enabled there; check the generated
# Setup.config before enabling it here.

#DB=/depot/sundry/src/berkeley-db/db.1.85
#DBPORT=$(DB)/PORT/irix.5.3
#bsddb bsddbmodule.c -I$(DBPORT)/include -I$(DBPORT) $(DBPORT)/libdb.a



# David Wayne Williams' soundex module (obsolete -- this will
disappear!)
#soundex soundex.c

# Helper module for various ascii-encoders
binascii binascii.c

# Fred Drake's interface to the Python parser
parser parsermodule.c

# Digital Creations' cStringIO and cPickle
cStringIO cStringIO.c
cPickle cPickle.c


# Lee Busby's SIGFPE modules.
# The library to link fpectl with is platform specific.
# Choose *one* of the options below for fpectl:

# For SGI IRIX (tested on 5.3):
#fpectl fpectlmodule.c -lfpe

# For Solaris with SunPro compiler (tested on Solaris 2.5 with SunPro C
4.2):
# (Without the compiler you don't have -lsunmath.)
#fpectl fpectlmodule.c -R/opt/SUNWspro/lib -lsunmath -lm

# For other systems: see instructions in fpectlmodule.c.
#fpectl fpectlmodule.c ...

# Test module for fpectl.  No extra libraries needed.
#fpetest fpetestmodule.c

# Andrew Kuchling's zlib module.
# This require zlib 1.1.3 (or later).
# See http://www.cdrom.com/pub/infozip/zlib/
#zlib zlibmodule.c -I$(prefix)/include -L$(exec_prefix)/lib -lz

# Interface to the Expat XML parser
#
# Expat is written by James Clark and must be downloaded separately
# (see below).  The pyexpat module was written by Paul Prescod after a
# prototype by Jack Jansen.
#
# The Expat dist includes Windows .lib and .dll files.  Home page is at
# http://www.jclark.com/xml/expat.html, the current production release
is
# always ftp://ftp.jclark.com/pub/xml/expat.zip.
#
# EXPAT_DIR, below, should point to the expat/ directory created by
# unpacking the Expat source distribution.
#
# Note: the expat build process doesn't yet build a libexpat.a; you can
# do this manually while we try convince the author to add it.  To do
so,
# cd to EXPAT_DIR, run "make" if you have not done so, then run:
#
#    ar cr libexpat.a xmltok/*.o xmlparse/*.o
#
#EXPAT_DIR=/usr/local/src/expat
#pyexpat pyexpat.c -I$(EXPAT_DIR)/xmlparse -L$(EXPAT_DIR) -lexpat


# Example -- included for reference only:
# xx xxmodule.c


Sent via Deja.com http://www.deja.com/
Before you buy.


From cobbe at directlink.net  Sat Dec  2 11:53:45 2000
From: cobbe at directlink.net (Richard Cobbe)
Date: Sat, 2 Dec 2000 10:53:45 -0600 (CST)
Subject: [OT] Apple IIe help please
In-Reply-To: <5.0.0.25.2.20001201225938.053fcde0@mrlizard.com>
References: <5.0.0.25.2.20001201225938.053fcde0@mrlizard.com>
Message-ID: <14889.10521.92707.537876@minbar.directlink.net>

Lo, on Friday, December 1, Lizard did write:

> At 06:25 PM 12/1/2000, D-Man wrote:
> 
> >Hi all.  I apologize for the off topic-ness of the message and the 
> >cross-post.
> >I have a friend who has a program she likes (written in BASIC) on an Apple 
> >IIe.
> >I have a way to get into the code and list it on the screen.  What I am 
> >looking
> >for is someone who has had some experience using an Apple IIe who can tell me
> >how I can get the listing to be redirected to the printer port.  Also if it is
> >possible to get a copy of the program file onto a disk for an IBM compatible
> >comptuer (Windows or Linux) that would be great.
> 
> Gods, THAT brings back memories! Something like "PRINT #1" or the like? I 
> know you specify a port..."IN #1" turned on the printer? Maybe? I'm sorry, 
> I'm dredging up memories from 20+ years ago...

Close.  Try PR#1 instead (at least, I *THINK* port 1 is the printer).  That
should redirect all subsequent output to the printer.  To reset, do PR#0.
So, something like

LOAD BLAH.BAS
PR#1            (from this point on, you'll be typing blind)
LIST
PR#0            (and you'll be back to normal)

As for converting the file to a PC disk, you'll probably need to transfer
it to a PC electronically---a dial-up link or a null modem cable is
probably your best bet.

Amazed that I still remember this stuff (haven't used an Apple II in 12
years),

Richard



From ykingma at accessforall.nl  Fri Dec 22 14:14:04 2000
From: ykingma at accessforall.nl (Ype Kingma)
Date: Fri, 22 Dec 2000 20:14:04 +0100
Subject: JPython embed question
References: <3A42AA88.D9BD321A@ncsa.uiuc.edu>
Message-ID: <3A43A7F6.91A10BE0@accessforall.nl>

Randy,

you wrote:
> 
> In the JPython demo/embed example, we find how to create an instance of
> a class, e.g.:
> 
>   interp.set("foo1", new Foo());
>   interp.set("bar1", new Bar());
> 
> I'd like to be able to dynamically create an instance of a class - e.g.,
> 
>   createInstance(String instName, String className)

You might subclass the class of interp in Java:

class MyInterpreter extends PyInterpreter /* or whatever it's called */ {
  /* constructors as needed */

  void createInstance(String instName, String className) {
    set(instName, newByClassName(className));
    /* sorry, forgot the actual name of newByClassName() from java.lang */
  }
} /* class MyInterpreter */


Kind regards,
Ype

P.S. Riddle: what's my default Python indentation?

(email at xs4all.nl)


From ennisr at my-deja.com  Tue Dec 26 13:50:48 2000
From: ennisr at my-deja.com (Ennis Rosamond)
Date: Tue, 26 Dec 2000 18:50:48 GMT
Subject: python telnet script for openvms host
References: <92ainp$jab$1@nnrp1.deja.com> <kuito7rrg7.fsf@lasipalatsi.fi>
Message-ID: <92apa7$oac$1@nnrp1.deja.com>

In article <kuito7rrg7.fsf at lasipalatsi.fi>,
  Erno Kuusela <erno-news at erno.iki.fi> wrote:
> Ennis Rosamond <ennisr at my-deja.com> writes:
>
> | tn.read_until('[c')    #last char of session hearder
> | tn.write('$ dir\r')
>
> | #following this, I receive back the sequence:
> | #recv '\377\373\001'
> | #IAC WILL 251
> | #recv '\377\373\003'
> | #IAC WILL 251
> | #recv '\003\\\033Z'
> | #
> | #which I don't know how to interpret.
>
> that looks like telnet option negotiation, telnetlib should
> take care of all that automaticalyl for you. i suspect
> the above is just some debug output from telnetlib.
>
> are you sure you are seeing it after your last .write call,
> and not at the start of the telnet connection?
>
>   -- erno
>

Thanks for the quick reply Erno.

You are correct.  The sequence I listed is the debug output.  I am
seeing it after the "tn.write('$ dir\r')" call.  It may not be relevant
if the host simply isn't ready to receive the write call.

I tried using "tn.read_until('$ ')" in the previous read instead of
"tn.read_until('[c')", but that sequence is never found.  That is why I
think the host is not sending back a prompt.  But then I am pretty naive
about this whole topic.


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


From aleaxit at yahoo.com  Mon Dec 11 08:59:51 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 11 Dec 2000 14:59:51 +0100
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com>
Message-ID: <912ml00a5o@news2.newsguy.com>

"Jaroslav Gergic" <j_gergic at yahoo.com> wrote in message
news:3A34CEA7.CD9D85AE at yahoo.com...
    [snip]
> >     file = open('file.input')
> >
> >     count = 0
> >     while 1:
> >         lines = file.readlines(8192)
> >         if not lines: break
> >         count = count + len(lines)

Pretty simple, classic, and fast.


> I hate it so much I prefer to write something like this:
>
> line = fh.readline()
> while(line != ""):
>   ... do something ...
>   line = fh.readline()

_Substantially_ slower: the first form you quote reads a
bunch of up to 8k bytes at a time, which will typically
speed it up importantly, while here you only read one
line at a time, making the runtime work quite a bit more.


> OK, it is ugly to write the same line of code twice,

Yes, particularly when such extra effort results in
a slowdown of one's program.

> but 'while true' or 'while 1' is very non-programmer construction
> it smells like VisualBasic GOTO.

"non-programmer"?  Ever heard of one Don Knuth, he of "The Art
of Computer Programming", TeX, etc?  If the 'non-programmers' I
happened to work with were as 'bad' as Dr. Knuth, I think I'd
be pretty happy.  Maybe I'm content with too little, and your
programming flair puts Dr. Knuth's to shame...?

Anyway, in his landmark 1974 article on the Communications of
the ACM, "Structured Programming WITH Go-To Statements", Dr.
Knuth made the point that the generally desired form of a loop
statement is (in pseudo-language):

    loop
        statements for part 1
        exit when condition
        statements for part 2
    repeat

which at the time no language supported directly ('while' and
'until' being just special-cases of this).  I guess he did not
know C, which already existed (not quite in the modern form)
and supports such forms as:

    for(;;) {
        statements for part 1;
        if(condition) break;
        statements for part 2;
    }

or, in exact and total equivalence:

    while(1) {
        statements for part 1;
        if(condition) break;
        statements for part 2;
    }

both of which are exactly implementations of the Knuth normal
loop-form.


Among the canonical (aka trivial) transformations of this
general loop-form is one introducing a flag-variable (*NOT*
the best solution, by any means, for reasons Knuth explains
lucidly in his article):

    flag = true
    while(flag) {
        statements for part 1;
        if(condition) {
            flag=false;
        } else {
            statements for part 2;
        }
    }

I'm not sure why one *would* want to use an inferior, slower,
less-clear, worse-coupled trivially-transformed pattern, but,
hey, whatever floats your boat.  So apply the canonical
transform to the beautiful code you hate, starting from:

    count = 0
    while 1:
        lines = file.readlines(8192)
        if not lines: break
        count = count + len(lines)

and produce the semantically-equivalent:

    count = 0
    flag = 1
    while flag:
        lines = file.readlines(8192)
        if lines:
            count = count + len(lines)
        else:
            flag = 0

At least, while goofier and slower, it will still be faster
than the one-line-at-a-time variant!  (It's also easily sped
up a bit by using a number larger than 8192, on most boxes --
try it with 32768, for example).


> Do you know better construction to avoid both - duplicate lines
> and horrible 'while 1' construction?

No, I don't know any BETTER construction than the while 1
loop; that is because there is nothing horrible about having
a loop in the general (Knuth) form, and "while 1" is the way
such a loop is written in Python (in C you get the choice
between 'while(1)' and the identically equivalent 'for(;;)',
but that's just syntax-sugar-level difference).

I do know many WORSE constructions, coming from canonical or
not-so-canonical transformations, such as the one I just gave.


> Something like:
>
> while( "" != (line = fh.readline()) ):
>   ... do something ...

This is a slightly different issue, where you want to BOTH
set something ('line') AND return a value to be tested; this
is idiomatic in C (where assignment is an expression) but not
in Python (where it isn't).

If you want to write set-something-AND-return-it in Python,
you can do it in several ways, of which the most Pythonic is
probably to use a class, e.g.:

class SetAndReturn:
    def setAs(self, data):
        self.data = data
        return data

line = SetAndReturn()
while line.setAs(fh.readline()):
    ... do stuff with line.data ...

This is still a 1-line-at-a-time (slow!) idiom, but making
it faster is trivial:

lines = SetAndReturn()
while lines.setAs(fh.readlines(32768)):
    ... do stuff with lines.data ...

you get a bunch of lines at a time, just like in the while 1
idiom you detest so much, so the speed is substantially OK
(the slight overhead of working with instance-objects and
attributes is not as bad as line-at-a-time file input).


This IS concise, though many Pythonistas (typically without a
C background) would probably find it obscure when compared
to the 'while 1' approach.


Alex





From g_will at cyberus.ca  Thu Dec 21 13:09:58 2000
From: g_will at cyberus.ca (Gordon Williams)
Date: Thu, 21 Dec 2000 13:09:58 -0500
Subject: Fatal Python error: PyThreadState_Get
References: <91rcjo$2dtl$1@news2.ottawa.cyberus.ca>
Message-ID: <91tgvv$2erl$1@news2.ottawa.cyberus.ca>

Thanks Mike,

It was an old Python module that I thought I had recompiled, but evidently
had not.

"Fatal Python error: PyThreadState_Get: no current thread" seems like a
funny error to indicate that there is a problem like this.  It would have
taken me hours/days to track this one down.

Regards,

Gordon Williams


----- Original Message -----
From: "Mike Fletcher" <mfletch at tpresence.com>
To: "'Gordon Williams'" <g_will at cyberus.ca>
Sent: Wednesday, December 20, 2000 6:35 PM
Subject: RE: Fatal Python error: PyThreadState_Get


> Normally it's caused by one module using Python15.dll while the rest of
the
> system is using python20.dll .  Check your pyds using dumpbin to see which
> one is doing the incorrect thing.
>
> HTH,
> Mike

"Gordon Williams" <g_will at cyberus.ca> wrote in message
news:91rcjo$2dtl$1 at news2.ottawa.cyberus.ca...
> Using
> W95
> Python 2.0
> Pythonwin build 135
>
> I am now in the process of converting some python 1.5.2 code to python
2.0.
> I now get the message below.
>
>
> Fatal Python error: PyThreadState_Get: no current thread
>
> abnormal program termination
>
> I'm sending and receiving information in one thread via a serial port.  In
> the 1.5.2 code I was using the Marshal serial module.  As that only works
> with 1.5.2, I have now converted that to use windows native stuff using
Mark
> Hamond's work.
>
> There was a thread earlier this month about this same problem.  I think
that
> they were also using W95.  It was unclear whether it was resolved or how
it
> was resolved.
>
> Before I start tracking the problem down, is this a known problem?  What
> causes it? Any pointers on where I should look first?
>
> Regards,
>
> Gordon Williams
>
>




From jadestar at idiom.com  Sun Dec 24 18:05:17 2000
From: jadestar at idiom.com (James T. Dennis)
Date: 24 Dec 2000 23:05:17 GMT
Subject: try: ... except EOFError: ... idiom
Message-ID: <925vfd$2ad9$1@news.idiom.com>

 In the "Source code management" thread I noticed a
 code snippet that used an unusual exception handling
 idiom:

gtalvola at my-deja.com

>      import marshal
>      import win32pipe
>      stream = win32pipe.popen('p4 -G changes', 'r')
>      changes = []
>      try:
>          while 1:
>              changes.append(marshal.load(stream))
>      except EOFError:
>          for c in changes:
>              print c['change'], c['desc']

 How would we write that without the try: ... except: idiom?

 I gather that Python has no bottom tested loop
 (do: ... until... or do: ... while) construct.

 Could we do something like:

	while 1:
		if stream.eof():
			break
		changes.append(marshal.load(stream)

 Is there a <<popen object>>.eof() method?  Is that 
 try: ... except: idiom EOFError *the* generally accepted
 way to process file/streams under Python?  Why isn't there
 some sort of eof() function or object method?

 Also: how would one do non-blocking I/O on pipes and popen()
 streams?



From glenfant at equod.com.nospam  Tue Dec 12 14:39:05 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Tue, 12 Dec 2000 20:39:05 +0100
Subject: IDLE and command line arguments
Message-ID: <915uku$95c$1@reader1.imaginet.fr>

Is there some hidden feature in IDLE to run a script with command-line
arguments ?

Thanks...



From amk at mira.erols.com  Sun Dec 10 10:57:35 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 10 Dec 2000 15:57:35 GMT
Subject: Unicode coercion not the same as ASCII
References: <dde43tgd9hikfq5sd6eau5tgn91akk244b@4ax.com>
Message-ID: <slrn937ao6.h4.amk@207-172-146-162.s162.tnt3.ann.va.dialup.rcn.com>

On Sat, 09 Dec 2000 14:12:04 +0000, 
	Dale Strickland-Clark <dale at out-think.NOSPAMco.uk> wrote:
>To get a number represented as a Unicode string, do I really need to do this?
>
>>>> unicode(str(123))
>u'123'

At the moment, yes, though there has been some discussion of a
unistr() function on python-dev.  I don't remember the issues that
make it non-trivial, so I'd recommend you hunt through the python-dev
archives for the details.

--amk



From mickey at Vanille.de  Mon Dec 11 12:02:18 2000
From: mickey at Vanille.de (mickey at Vanille.de)
Date: 11 Dec 2000 19:02:18 +0200
Subject: UML and Python
References: <90o91k$eks$1@desperado.vew-telnet.net> <90stth$mea$1@newsreaderg1.core.theplanet.net> <m3aea5te8t.fsf@wilma.localdomain> <rj993tc5ig1kn6tcvovkajrgarpcgevubs@4ax.com>
Message-ID: <3a3516aa@nntp.server.uni-frankfurt.de>

Anders M Eriksson <anders.eriksson at morateknikutveckling.se> wrote:

> On 09 Dec 2000 14:35:14 -0800, Harry George <hgg9140 at seanet.com>
> wrote:

>>You could also consider Dia.  It doesn't (yet) have a model repository

> A link would be nice!

Dia can be found @ http://www.lysator.liu.se/~alla/dia/

-- 
Michael 'Mickey' Lauer . . . . . . . . . mickey at Vanille.de
How could anyone know me - when I don't even know myself ?




From jacobs at darwin.epbi.cwru.edu  Wed Dec 20 07:28:35 2000
From: jacobs at darwin.epbi.cwru.edu (Kevin Jacobs)
Date: 20 Dec 2000 12:28:35 GMT
Subject: UN/EDIFACT related packages
References: <91lg8h$2mo$1@reader1.imaginet.fr> <3A3F7049.29CCAC24@darwin.in-berlin.de>
Message-ID: <91q8lj$7gl$1@slate.INS.CWRU.Edu>

Dinu C. Gherman <gherman at darwin.in-berlin.de> wrote:
> I know that Cees de Groot did some work in that field,
> resulting in a package called "ema". Hopefully he will
> listen and speak up. 

Actually, it was Laurent Szyster (thanks to Google).  I don't know what has
come of it, but his annoucement and e-mail address are included below:

Title: EMA, an EDIFACT mapper 
Resource type: FYI 
Date: Thu, 24 Sep 1998 15:48:39 
Source: E-mail 
Author: Laurent Szyster (szyster at ibm.net)
Keywords: EMA ,EDIFACT MAPPER ,SW/OPEN SOURCE ,PROGRAM 
Original URL: 

Abstract/Summary: 

Hello, 

As part of an ongoing project, I needed a program that could map EDIFACT
interchange to HTML pages. Since it has to be integrated into a larger
network application, I decided that the best thing would be to write one
myself (I found Mercator and similar software too heavy to use and too
difficult to integrate or customize). 

EMA, an EDIFACT mapper, can parse and validate EDIFACT interchanges and
messages using definitions written in XML, then map them to various text
based format like HTML, CSV or tab-delimited using maps written in XML and
ad-hoc function libraries (using Python). 

This is alpha software but it already looks promising. Moreover it adheres
to Open Source Software requirements: it's free in every sense of the word
("free speach" as well as "free beer"), thanks to RINET s.c. good will. Any
comments, whishes and ... _code_ contributions are welcome (*). 

Some web space and mailing list are on their way, so to get EMA you'll have
to send me short mail asking for it ;-) 

To test it you'll need Python version 1.5, the language used to develop EMA.
It will run on all platform supported by Python: Windows 95/98/NT, DOS,
OS/2, all UNIXes, MacOS and even Java (using JPython). 

Laurent Szyster 

___ RINET s.c. 

(*) Don't send me money, it's FREE OF CHARGE! 



From grey at despair.rpglink.com  Sun Dec 24 06:26:14 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Sun, 24 Dec 2000 11:26:14 -0000
Subject: coding convention conversion
References: <mailman.977516469.1157.python-list@python.org> <920v4s$oqj$2@taliesin.netcom.net.uk> <slrn949c13.40l.grey@teleute.rpglink.com> <923jin$q2l$1@taliesin.netcom.net.uk>
Message-ID: <slrn94bnam.i0h.grey@teleute.rpglink.com>

On Sun, 24 Dec 2000 01:29:59 +0000, Garry Knight <garryknight at bigfoot.com>
wrote:
>In article <slrn949c13.40l.grey at teleute.rpglink.com> "Steve Lamb"
><grey at despair.rpglink.com> wrote:
>>     Anyone else find it ironic that someone is solving a problem using
>>     Perl in a Python group when the comparable Python would be an
>>     excellent exercise to perform?  :)

>IMHO some languages seem to fit some tasks better than others. Perl is
>designed for just this kind of job. A Python solution would, however,
>be far more readable: I have to look very closely at my Perl code to
>double-check it. On the other hand, the Python solution would probably
>be a fair bit longer.

    How so?  The only part I got stuck on on the Python version was the \l\1
construct in the regular expression.  Aside from that it was not significantly
longer.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From root at rainerdeyke.com  Wed Dec 20 13:08:05 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Wed, 20 Dec 2000 18:08:05 GMT
Subject: while true: !!!
References: <mailman.977127721.4398.python-list@python.org> <DnD%5.166633$U46.5342944@news1.sttls1.wa.home.com> <slrn93vcml.nh7.qrczak@qrnik.zagroda> <%8Y%5.172260$U46.5443571@news1.sttls1.wa.home.com>
Message-ID: <9A606.57777$x6.25206290@news2.rdc2.tx.home.com>

"Greg Jorgensen" <gregj at pobox.com> wrote in message
news:%8Y%5.172260$U46.5443571 at news1.sttls1.wa.home.com...
> It was an old compiler. I haven't used C++ in over a year, but I believe
> that the addition of bool caused some problems: sizeof(bool) is not
> standardized, and bool is defined as 0 is false, anything else is true,
> which means a bool expression may not be equal to "true." Please forgive
me
> if my understanding of this isn't correct: the last C++ compiler I used
> didn't have bool support built-in.

In modern C++, a 'bool' can have exactly two values: 'true' and 'false'.
When promoted to an integer type, 'true' translates to '1' and 'false'
translates to '0'.  When converting from an integer type to 'bool', '0'
becomes 'false' and everything else becomes 'true'.  In other words,
'int(bool(5))' evaluates to '1'.  The comparison operators (==, !=, <, etc.)
all yield a value of type 'bool'.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From MarkH at ActiveState.com  Sat Dec  9 19:22:44 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sun, 10 Dec 2000 00:22:44 GMT
Subject: editing python path in pythonwin
References: <90ufli$8f3$1@nnrp1.deja.com> <90ug7e$8s1$1@nnrp1.deja.com>
Message-ID: <3A32CB8F.3000806@ActiveState.com>

red5sf at my-deja.com wrote:

> just found my problem,
> it's the pythonwin path variable that i should be editing, not the
> pythonpath one!

Actually, _any_ subkey in the registry _except_ the "top" one (for the 
core) will work.  The best solution is probably to create your _own_ 
registry key, and add your own entries there.  Then if Pythonwin etc get 
uninstalled, you should not lose your new entry.

Mark.



From garryknight at bigfoot.com  Sun Dec 24 13:18:41 2000
From: garryknight at bigfoot.com (Garry Knight)
Date: Sun, 24 Dec 2000 18:18:41 +0000
Subject: coding convention conversion
References: <mailman.977516469.1157.python-list@python.org> <920v4s$oqj$2@taliesin.netcom.net.uk> <slrn949c13.40l.grey@teleute.rpglink.com> <923jin$q2l$1@taliesin.netcom.net.uk> <slrn94bnam.i0h.grey@teleute.rpglink.com>
Message-ID: <925em1$jcb$1@taliesin.netcom.net.uk>

In article <slrn94bnam.i0h.grey at teleute.rpglink.com> "Steve Lamb"
<grey at despair.rpglink.com> wrote:

>> the Python solution would probably be a fair bit longer.
> 
> How so?  The only part I got stuck on on the Python version was the
> \l\1
> construct in the regular expression.  Aside from that it was not
> significantly longer.

I meant longer as in the number of lines of code. I'd be very
interested in seeing the Python version.

-- 
Garry Knight
garryknight at bigfoot.com


From cnelson9 at gte.net  Tue Dec 19 21:09:27 2000
From: cnelson9 at gte.net (Clifford J. Nelson)
Date: Wed, 20 Dec 2000 02:09:27 GMT
Subject: Python in "math through programming" curriculum
References: <2g2t3t06r2agrrjmfnr41eu4lc2u39a8qi@4ax.com> <3A3EFF47.67DE69CF@gte.net> <513v3toslkjkggh9oumlf20qc9bi1gdtif@4ax.com> <3A3FC677.FDB19937@gte.net> <CeP%5.521$AH6.45315@newsc.telia.net>
Message-ID: <3A401C2E.1FF789A8@gte.net>


Fredrik Lundh wrote:
> 
> Clifford J. Nelson wrote:
> > I'm sure I don't attach ANY weight to any of your opinions.
> > I've seen your web sites about Synergetics and programming!
> 
> clifford, please keep your personal problems off
> comp.lang.python.
> 
> </F>

You're right. I take it personally when Bucky Fuller's
Synergetics (see 1) is made to seem like some far out
religion, by Kirby Urner or anyone else, or when computer
programming in a language (Python) that can't apparently
help itself well enough to use it to make a good interactive
programming environment for the Mac, is promoted over much
easier languages, like Clean (see 2) or Object Caml 3.0 (see
3) or even the hard to learn J programming language (see 4).

All of those languages are garbage compared to Mathematica
of course because programming (OOP) is all just really
pattern matching and replacement (see 5), but they're free
and they work on both the Mac and Windows, so they are worth
looking at.

1) http://www.rwgrayprojects.com/synergetics/synergetics.html
2) http://www.deene.ufu.br/clean/clean1.html
3) http://pauillac.inria.fr/caml/
4) http://www.jsoftware.com/
5) http://www.stephenwolfram.com/publications/talks/IMS/imstalk.html

  Cliff Nelson


From nospam at no.spam  Mon Dec 18 07:20:14 2000
From: nospam at no.spam (Carsten Geckeler)
Date: Mon, 18 Dec 2000 13:20:14 +0100 (CET)
Subject: while true: !!!
In-Reply-To: <B43D149A9AB2D411971300B0D03D7E8B0FBCBB@natasha.auslabs.avaya.com>
Message-ID: <Pine.LNX.4.21.0012181318140.22206-100000@justitia.jura.uni-tuebingen.de>

On Mon, 18 Dec 2000, Delaney, Timothy wrote:

> This is precisely the reason this thread exists ...
> 
> The idiom is actually
> 
> while 1:
> 
> 	do something
> 
> 	if condition:
> 		break
> 
> 	do something else
> 
> Basically, it's for dealing with situations where you can't tell at the
> beginning or end of a loop if the loop is completed - only somewhere in the
> middle. The canonical example is
> 
> while 1:
> 
> 	try to read a line from a file
> 
> 	if we didn't get a line:
> 		break
> 
> 	do something with the line
> 
> There are many ways of dealing with this, but they all have compromises - in
> most cases either duplication of code, or an assumption of memory
> availability.

Well, I would prefere the following:

while not f.eof():
	line = f.readline()
	# do stuff

but unfortunatly the Python file objects don't have this method.

Cheers, Carsten
-- 
Carsten Geckeler:  carsten dot geckeler at gmx dot de
To get proper email-address replace `dot' and `at' by the corresponding symbols.




From mjjacob at my-deja.com  Wed Dec 13 11:40:52 2000
From: mjjacob at my-deja.com (mjjacob at my-deja.com)
Date: Wed, 13 Dec 2000 16:40:52 GMT
Subject: Problem with += (augmented assignment)
Message-ID: <9188qi$q7d$1@nnrp1.deja.com>

Hi folks!

I tried this under Python 2.0 running on Win 2k:

x = 0
while x <= 42:
	x += 1

Python always complains a Syntax Error: Invalid Syntax on this:

    x += 1
       ^
SyntaxError: invalid syntax

Augmented assignment only seems to work when used in first indentation
level (script level). Otherwise, I always get that syntax error, it
does not matter whether I am using if, while or whatever.

Is there something terribly wrong with the += operator or is that all
my fault?

Help appreciated.


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


From grante at visi.com  Fri Dec 15 10:47:11 2000
From: grante at visi.com (Grant Edwards)
Date: Fri, 15 Dec 2000 15:47:11 GMT
Subject: Accessing serial port on windows
References: <91dds1$b0i$05$1@news.t-online.com>
Message-ID: <32r_5.6988$TU6.714464@ptah.visi.com>

In article <91dds1$b0i$05$1 at news.t-online.com>, Dennis Conrad wrote:

>I'm "ogling" to move from Perl to Python as my "main scripting language" and
>wanted to learn a bit of it by writing my new project in Python. But I need
>to read data from the serial port (on Windows) and can't find any
>information on this on the net. Is there a possibility to do it? Or do I
>have to use the WinAPI?

I've used the sio module, and it seemed to work well enough:

       ftp://ftp.python.org/pub/python/contrib/sio-151.zip

-- 
Grant Edwards                   grante             Yow!  Ha ha  Ha ha Ha ha
                                  at               Ha Ha Ha Ha -- When will I
                               visi.com            EVER stop HAVING FUN?!!


From glenfant at equod.com.nospam  Wed Dec 13 09:06:33 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Wed, 13 Dec 2000 15:06:33 +0100
Subject: Self-documenting Python??
References: <x881yvcfodl.fsf@adglinux1.hns.com>
Message-ID: <917vhe$k0v$1@reader1.imaginet.fr>

I got the Mandrake rpm burn of Python. Same mess...
Delete this documents (installed by the rpm), download the full documents
set from www.python.org.
and put the first page (index.html) in your Netscape bookmarks...
Does this seem difficult ?

For the rest...
IDLE has syntax coloring and a path browser, try it.
put __doc__, __author__ and __version__ variables in each personal module
You can make pretty print with py2html with
http://www.lemburg.com/files/python/ or in PDF with a demo package from
reportlab http://www.reportlab.com

<nbecker at fred.net> a ?crit dans le message news:
x881yvcfodl.fsf at adglinux1.hns.com...
> Has there been any proposals to integrate a documentation system with
> python?  I feel this is one feature that is sorely missing.
>
> We have a way to put some very minimal documentation into python
> sources, but there is no integrated help browser to use it.
>
> I use RedHat Linux here.  The standard Python docs are installed in a
> manner by the standard distribution that they are not accessible
> except by users sufficiently knowledgeable to know where and how to
> look for them.  Most people can use a web browser, but they don't know
> that /usr/share/doc/python.... has something of interest.
> Furthermore, documentation for user-written code is not accessible at
> all.
>
> I believe nothing would do more to further the wider acceptance of
> Python than integrated help.



From calcium at altavista.net  Sun Dec 17 05:21:21 2000
From: calcium at altavista.net (ccang (bigpond))
Date: Sun, 17 Dec 2000 21:21:21 +1100
Subject: ID3V2 : Python script to rename MP3 file using ID3 V2 tag.
Message-ID: <3A3C93A1.F7D8162E@altavista.net>

Hi there all,
I recently posted a Python script to rename an mp3 file
using it's id3 tag. However, this only did id3 v1.

I've got another script which does id3 v2. (sort of).
See http://www.id3.org for more info on id3.
However, it only reads id3v2 tags.
There is no facility to create id3 v2 tags or
handle extended frames info.
Maybe my next post.

It is far from complete. Only doing it out of interest
as there doesnt seem to be any V2 implementations
in Python.

Script available from
http://www.ozemail.com.au/~calcium/scripts.html

Regards,
Chai
http://www.ozemail.com.au/~calcium
--
| mailto:calcium at ozemail.com.au   | +61 401 688 408
|  Ob saying - Everything I say is the truth, but as I
|  get older, my view of what the truth is changes.


From DanielK at aracnet.com  Sat Dec 16 12:55:20 2000
From: DanielK at aracnet.com (Daniel Klein)
Date: Sat, 16 Dec 2000 09:55:20 -0800
Subject: Capturing stdout from a console process?
References: <91b8q3$aqu$1@nnrp1.deja.com>
Message-ID: <wXN_5.95$LU6.31235@typhoon.aracnet.com>

Have a look at the popen2 module which provides classes for managing the
stdin, stdout and stderr of sub-processes. Mr. Lundh's excellent ebook "The
Standard Python Library" has good examples of its use.

Daniel Klein

"Sam Schulenburg" <samschul at pacbell.net> wrote in message
news:91b8q3$aqu$1 at nnrp1.deja.com...
> I am using <os.system("xcopy c:*.* d: /v")> within a python program to
> copy files from one directory to another under windows NT. When the
> program executes, a new console window is created with the results of
> the xcopy command being displayed in the console window. Is their a way
> to capture this output by python so I can log error messages?
>
> Sam Schulenburg
>
>
> Sent via Deja.com
> http://www.deja.com/




From pyth at relax.dk  Mon Dec 11 05:34:59 2000
From: pyth at relax.dk (pyth at relax.dk)
Date: Mon, 11 Dec 2000 11:34:59 +0100 (CET)
Subject: variable variables
In-Reply-To: <XFMail.001211112202.mikael@isy.liu.se>
Message-ID: <Pine.LNX.4.21.0012111127460.5044-100000@relax.dk>

> I guess you want to be able to have multiple references to the same
> variable, and that you want to be able to change the value of that
> variable in such a way that all references refer to the same new value.

> /Mikael

Sorry Mikael, for using your time... I should have been more specific...
What i was thinking of, was something like this:

test1 = 'test text 1';
test2 = 'test text 2';

var1 = 1
var2 = 2

print test{var1};

This should print: 'test text 1'
This example doesn't work, but is close how you would do it in php :o)

Thanx for tryin' Mikael :o)

/Dan Larsen






From jfdeford at my-deja.com  Thu Dec  7 17:46:37 2000
From: jfdeford at my-deja.com (jfdeford at my-deja.com)
Date: Thu, 07 Dec 2000 22:46:37 GMT
Subject: returning base class pointer from function
Message-ID: <90p40c$7pq$1@nnrp1.deja.com>

Okay, this is likely an easy one for somebody who
understands even a little bit about Python (this
apparently leaves me out).  I have built an
extension dll in C++, and via SWIG have the
appropriate shadow classes and such for the
various objects.  The situation is as follows: a
member function of one of my SWIGed C++ classes
returns a base class pointer (say class A) to an
object that is derived from A (class B : public
A).  Specifically, I have the following member
function declared in one of my classes:

	virtual A*	GetObject(const char
*pcPath, bool bCopyData);


wherein an object of type B is created and
returned (this function can return any one of a
number of objects that are all derived from A).
I can convince SWIG to create the appropriate
object (B) via Py_BuildValue prior to returning
it from the GetObject function in question (using
a typemap), and the shadow class just passes out
the object it gets from the wrapped GetObject
call, i.e. in the body of the shadow class
definition I find,

    def GetObject(self,arg0,arg1):
        val = pyordatac.PyORDatabase_GetObject
(self.this,arg0,arg1)
        return val

When I make an assignment at the command line
like:

foo=blah.GetObject(....)

foo is nominally of "type" B, but it seems that I
can't use it to access functionality in B, e.g.,
foo.<func only in B>() gives me an attribute
error.  In C++ I would just cast the pointer to
the appropriate derived type - is this my only
option here?  If I do the following:

boo=swigged_B_classPtr(foo)

where swigged_B_classPtr is the shadow class
associated with my B class, then I can get at the
B class member functions.  I'd rather avoid
this "cast" step if possible, and I don't really
know what it's doing anyway.





Sent via Deja.com http://www.deja.com/
Before you buy.


From e_d_k at my-deja.com  Sat Dec 23 18:28:42 2000
From: e_d_k at my-deja.com (EdK)
Date: Sat, 23 Dec 2000 23:28:42 GMT
Subject: Problems installing PIL
Message-ID: <923cf6$s7q$1@nnrp1.deja.com>

I am unsing Python 1.6 on Windows 98. I folowed the instructions and copied
the PIL directory to c:\python16 and set PYTHONPATH=c:\python16\PIL.

I have no problem importing Image or creating an Image, but whenever I try to
manipulate the Image I get:

  ImportError: The _imaging C module is not installed

I assume this is refering to _imaging.pyd, which is located in
c:\python16\PIL. How do I install it?


    -EdK


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


From aleaxit at yahoo.com  Thu Dec 28 10:54:59 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 28 Dec 2000 16:54:59 +0100
Subject: gmpy 0.8 released (fast, high-precision arithmetic)
Message-ID: <92fnt80244v@news1.newsguy.com>

I've posted the announcement to c.l.p.announce, but its
moderation doesn't seem to be active these days.


GMP (http://www.swox.com/gmp/) is a fast free library
for arbitrary-precision arithmetic (integers, rationals,
floats), currently at release 3.1.1.

gmpy (http://gmpy.sourceforge.net/) is a C-language
module exposing GMP functionality to Python programs
(and to other C-written Python modules too, through
a simple 'C API').

Other, older modules exposed a subset of GMP's "mpz"
(integers) functionality to older Python versions;
gmpy exposes just about all of GMP 3.1.1 (including
rationals, floats, and a subset of its random number
generation functionality) to Python 2.0, using the
latter's "distutils" to ease building/installing.

The current gmpy release (0.8 pre-alpha) is (I hope!)
very close to alpha level: all relevant GMP functionality
is exposed, there are no known bugs (and workarounds for
a couple of GMP, well, 'anomalies'), documentation is
there (in docstrings _and_ in a textfile), and so are
unit-tests (almost 1,000 are routinely run, with SLOC
coverage of over 92% -- uncovered SLOCs being, almost
exclusively, tests for out-of-memory situations, and
optional debugging-output fprintf's-to-sdterr; I believe
this level of testing-coverage will be adequate for an
alpha-level release).  gmpy 0.8 includes a framework
of call-backs that's not really documented yet (mostly
meant to ease integration with PySymbolic), but that
should not disturb any 'normal' gmpy use.

Main changes currently foreseen for the planned 1.0
alpha release (which I hope to have within January):
[a] I'll rework the current approach to setting module
level options/flags, [b] I'll pretty-up the docs, both
docstrings (reformatting them as per normal Python
conventions) and external-documentation (adding examples,
choosing some spiffier format than plain text).

Any bugfixes that should prove necessary will get the
highest priority.  Performance issues will not (I plan
to address those in the later alpha->beta transition);
I consider current performance well-adequate for an
alpha release (it's faster than Python's built-in
longs for many tasks, and, according to some benchmarks
as kindly measured by Pearu Peterson, faster than many
existing interactive/interpreted systems that can
perform similar arbitrary-precision arithmetic tasks).


Gmpy sources (needed for ALL platforms, as they include
documentation, unit-tests, distutils' setup.py) are at:
http://download.sourceforge.net/gmpy/gmpy-sources-08.zip
(for non-Win32 platforms, you'll have to download and
install GMP 3.1.1 separately before you can build gmpy
from these sources, see http://www.swox.com/gmp/#DOWNLOAD;
be warned that some Linux system come with older releases
of GMP -- ensure you have GMP 3.1.1 or better installed!).


Gmpy binaries for Win32 / MSVC++6 are available at:
http://download.sourceforge.net/gmpy/gmpy-win32-library.zip
    Lundy's port of GMP.LIB 3.1.1, and GMP.H (only needed
    if you intend to rebuild GMPY.PYD from sources with
    Microsoft VC++6 on Win32, in addition to gmpy-sources).
    ftp://ftp.whiterose.net/pub/lundman/ is also OK for this
    [but you have to know exactly what to get from there...].
    Note this is NOT updated with gmpy releases; you do not
    need to download it again if you already have it!).
http://download.sourceforge.net/gmpy/gmpy-win32-binary-08.zip
    pre-built gmpy.pyd (release: pre-alpha 0.8), only needed
    on Win32 if you do NOT intend to rebuild GMPY.PYD.  Please
    note: you ALSO need to download gmpy-sources for docs &c!


Alex





From dbrueck at edgix.com  Tue Dec  5 09:48:22 2000
From: dbrueck at edgix.com (Dave Brueck)
Date: Tue, 5 Dec 2000 07:48:22 -0700
Subject: Newbie question on 'cmpfiles':
In-Reply-To: <90i074$d73$1@nnrp1.deja.com>
Message-ID: <009301c05eca$6bc5e530$760214ac@PRODUT2KDAVE>

Hi Hari,
I haven't really used this module, so this may not help too much:

common is a list of filenames that exist in both directories. Depending on
what you need, maybe you shouldn't call that function at all (instead use
filecmp.cmp to compare two files, and create a filecmp.dircmp object to
compare the contents of two directories).

HTH,
Dave

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of harisri at my-deja.com
> Sent: Monday, December 04, 2000 10:59 PM
> To: python-list at python.org
> Subject: Newbie question on 'cmpfiles':
>
>
> Hello,
>
> I am a newbie user to python 2.0
>
> Can someone help me by explaining 'cmpfiles' function with a simple
> example?
>
> I am trying to compare 2 directories, but not sure about the syntaxof
> this function. The "Python Reference Library" did not help me on this.
>
> [What does "common" mean in "cmpfiles(dir1, dir2, common)"?]
>
> Thanks in advance,
> Hari.
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
> --
> http://www.python.org/mailman/listinfo/python-list




From slhath at home.com.nospam  Mon Dec 11 13:45:44 2000
From: slhath at home.com.nospam (Scott Hathaway)
Date: Mon, 11 Dec 2000 18:45:44 GMT
Subject: winreg
Message-ID: <sh9Z5.6204$bw.481106@news.flash.net>

Hello Everyone,

I was looking at the online documentation on the winreg module that allows
editing of the windows registry.  I am too new to python to understand how
to code it.  Can someone get me started in the right direction?

I need to add a key and then add 5 values to that key.

Thanks,
Scott




From thomas at xs4all.net  Sun Dec 24 15:07:42 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Sun, 24 Dec 2000 21:07:42 +0100
Subject: Some Python 2.1 ideas
In-Reply-To: <3A464D5D.B541C04E@uswest.net>; from pbjorn@uswest.net on Sun, Dec 24, 2000 at 12:24:13PM -0700
References: <mailman.977590932.19335.python-list@python.org> <Cw616.23054$qN6.3308205@typhoon.we.rr.com> <3A452881.7686E717@uswest.net> <%vb16.187233$U46.5801214@news1.sttls1.wa.home.com> <3A464D5D.B541C04E@uswest.net>
Message-ID: <20001224210742.A18246@xs4all.nl>

On Sun, Dec 24, 2000 at 12:24:13PM -0700, Bjorn Pettersen wrote:
> Greg Jorgensen wrote:

> > "Bjorn Pettersen" <pbjorn at uswest.net> wrote in message
> > [snip]
> > How should these be handled?

> >   "foo\n\n\n"
> >   "foo\n\r\n\r"
> 
> "foo\n\n\n" -> "foo\n\n"
> "foo\n\r\n\r" -> "foo\n\r"

Note that the latter was a trick question! The proper order for DOS-EOL is
'\r\n', *not* '\n\r'. So the string "foo\n\r\n\r" is a Macintosh string, not
a DOS one :-)

> > Wouldn't it be cleaner to enhance the file object with additional
> > methods:
> >
> >   f. readline_chomp()
> >   f.readlines_chomp()

> This would be a good idea, except it wouldn't fit with the standard Python idiom
> for reading files:

And everything that poses as a file object would have to duplicate those
functions, all of a sudden. Guido is pretty conservative in adding
conveniency functions to builtin (and oft-imitated) objects because of that.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From rturpin at my-deja.com  Tue Dec 26 22:06:14 2000
From: rturpin at my-deja.com (rturpin at my-deja.com)
Date: Wed, 27 Dec 2000 03:06:14 GMT
Subject: Proposal for attribute and method caching (was: One Python 2.1 idea)
References: <mailman.977857210.10962.python-list@python.org>
Message-ID: <92bmb4$eif$1@nnrp1.deja.com>


The paper by Lowis[1] confirms that attribute and method
lookup consume a large part of a Python program's execution.
His 38% reduction in execution time is attractive, even if
its measurement is muddy.

However, I think he is taking the wrong approach. He uses
class caches based on intern'd string ID. This requires
fairly significant, additional data structures, whose
maintenance somewhat offsets the gains from caching. It
works for a small, fixed number of method names, losing
its advantage for large programs, where it is needed most.
His approach to instance attributes speeds up only the
failed lookup that is required when a method is invoked.
It does not speed attribute reference generally, which is
the more frequent case.

Python's semantics are beautiful and subtle. Method and
attribute caching need to be local, lightweight, and
tailored to Python. I've been thinking about this problem,
and propose a caching scheme below. This scheme should
eliminate dictionary lookups for the vast majority of
attribute and method references. Attributes are tackled
prior to methods. The scheme requires one or two new
primitive types, but the new data structures required are
local, and relatively lightweight. I explain the scheme in
three steps.

Two caveats. First, I'm putting this idea out, hoping
it sparks discussion for possible performance improvements
to Python. As much as I would love to take a few weeks to
experiment with this, my time is pretty heavily committed
for the next year. Second, I am no expert on how Python
now is implemented. I think I understand Python semantics
pretty well. Still .. I may be missing something, in which
case I expect those who know more than me to cover the
ideas below with well-thrown tomatoes.


STEP 1: LIGHTEN INSTANCE DICTIONARIES

In Python, every class instance has a dictionary of its
attributes that are built up at run time. This means
instances are relatively heavy-weight, compared to
statically bound OO languages such as C++. (In C++, an
instance for a complex number type typically stores
nothing more than two real numbers.) This first step
slightly modifies how instances are represented. It does
NOT improve performance generally, but is a precursor to
the next step.

Define a new Mapping type called a light dictionary.
This type has exactly the same exposed interface as the
dictionary, except for its constructor and a few additional
operations available to the Python machine. A light
dictionary is implemented by three attributes that are
NOT exposed. They are available only to the C code that
implements the type. These are:

  indexDict: A dictionary mapping n strings into range(n)
  indexedVal: A static array [0:n-1] of Python objects
  extraDict: A dictionary, initially None

At first glance, a light dictionary looks a good deal
heavier than an ordinary dictionary, since it includes two
of them as attributes. But the indexDict is a constant
dictionary, created at compile time, shared by any number
of light dictionaries. And the extraDict typically is not
used. The index dictionary holds the "expected" set of
names. For a class instance, this is all instance
attributes that can be found in the class definition,
or from any inherited class definitions. (Conditioned or
not.) Attribute values are stored in the indexedVal array:

    __dict__.indexedVal[indexDict["foo"]] = value of x.foo

The index dictionary that initialize an instance's light
dictionary is a class attribute: __attrIndexDict__. It is
created when the class is defined, it is NOT exposed, and
it never changes after it is created. It is the most
complex internal data structure whose addition I propose.

Light dictionaries support a few operations available only
to the Python machine. I will use Python code to describe
their behavior, and later, show how they are used. But
these are NOT methods. They should be implemented directly
in the Python machine, and run much faster than methods.

    def getIndex(self, key):
        if indexDict:
            return self.indexDict.get(key, None)

    def getByIndex(self, index):
        if self.indexDict:
            if type(self.indexedVal[index])!=UnassignedType:
                return self.indexedVal[index]
            else:
                raise AttributeError
        else:
            raise <some exception>

    def setByIndex(self, index, value):
        if self.indexDict:
            self.indexedVal[index] = value
        else:
            raise <some exception>

    def isIndexable(self):
        if self.indexDict:
            return 1

Getting values from and setting values in the indexedVal
array should be true, C index operations.  getByIndex()
must throw an exception if the indexed element is unassigned.
There are several ways to do this. I would suggest adding
another primitive type, Unassigned.

Using a light dictionary for __dict__, object instances
are a little more like the instances of static languages
such as C++. The initial representation of the instance's
attributes are a reference to a class dictionary, and an
array to hold the attributes' future values.

A light dictionary mimics exactly the behavior of a
dictionary. A new key that is not an intern'd string is put
into extraDict. Key lookup is done in indexDict, for
intern'd strings, and in extraDict, for other strings.
Deleting one of the keys in indexDict is done by setting
the indexed element of indexedVal to Unassigned. Deleting
a key from extraDict is done normally. A new key that is an
intern'd string copies all the index keys and values into
extraDict, sets indexDict to None, and then adds the new
key to extraDict. The light dictionary then is no longer
indexable.

The performance of a light dictionary is little different
from that of an ordinary dictionary. Except for a couple of
tests, the exposed operations are turned immediately into
the same operation on indexDict or extraDict. There is a
small gain on first assignment to the indexed keys, and a
penalty when a new intern'd string is added.

Importantly, a light dictionary has the same semantics
and exposes the same interface as an ordinary dictionary.
Anything the Python programmer does with x.__dict__ he can
still do, with the same effect. He can't reach the class's
__attrIndexDict__, because it is strictly an internal data
structure, and the internal attributes of a reference
dictionary are not exposed.


2. CACHE ATTRIBUTE REFERENCES

The big win from light dictionaries is that attribute
reference can be made a pure indexing operation, for
attributes that are determined when the class is defined.
Consider a typical attribute reference:

    y = x.foo

The byte code holds the string "foo" via a reference:

    _0: "foo"

We're going to add three other pieces of information, all
either references or integers:

    _0: "foo"
    _1: x.__class__, initially set to None
    _2: -1, reserved for methods
    _3: foo's index in x's reference dictionary

Of course, x's class might be different for different
executions of this code fragment. Local caching derives its
benefit from the fact that most times, x has the same class
as the previous pass through the same code. To retrieve the
attribute, the Python machine performs an operation that
has the following behavior:

    if x.__dict__.isIndexable:
        if x.__class__!=_1:     # New class
            _1 = x.__class__
            _3 = x.__dict__.getIndex("foo")
        if _3:
            y = x.__dict__.getByIndex(_3)
        else:
            <search other namespaces>
    else:                       # Polluted instance namespace
        if y = x.__dict__.has_key("foo"):
            y = x.__dict__.get("foo")
        else:
            <search other name spaces>

In the good case, three comparisons and an indexing operation
return the attribute value. On the first time through, the
class's dictionary must be queried. Once the instance's
dictionary is "polluted" with intern'd strings other than its
known attributes, behavior is exactly as it is now, with the
extra overhead of a comparison. If the good case is the
majority of the time, attribute references will be made MUCH
faster. A similar operation, but using setByIndex(), is used
when writing to an attribute.

The common case where this caching technique fails is when
a superclass's method is applied to instances of different
subtypes. It might be possible to optimize a bit further, by
making each class's __attrIndexDict__ an extension of its
superclass's. A little twiddling would restore caching for
"inherited" attributes, even when every instance is a
different subtype. I haven't thought about this further
optimization much, and it might not work for multiple
inheritance.

There is one last "gotcha" to handle. In Python, you can
change the class of an instance by assigning the __class__
attribute. When __class__ changes, __dict__ must be forced
non-indexable, as if an intern'd string were added to it.
After this, __dict__ behaves exactly like a normal
dictionary, because it holds a normal dictionary to which
it delegates all methods.


3. CACHE METHOD REFERENCES

There are two problems faced by any attempt to cache Python
methods. First, instance attributes (including functions)
may override defined methods. Second, class methods can be
changed on the fly.

The algorithm below caches only methods whose names (a) are
intern'd strings, and (b) are not predetermined attribute
names, i.e., don't appear in the class's __attrIndexDict__.
Methods that are so qualified are marked in a way that can
be inspected by the Python machine. Abusing notation, I
will indicate this through an attribute on the method:

    C.method.__qualified__

Given this qualification of methods, the first problem can
be solved by making sure no new intern'd names have been
added to the instance's namespace. Fortunately, this is
exactly what x.__dict__.isIndexable() does.

To solve the second problem, a new attribute is added to
class objects: __methodChanges__. It counts the number of
times that the methods of the class OR ITS SUBCLASSES have
changed. Initialized to zero, whenever a method is changed
in a class, this attribute is incremented, all the way up
the class hierarchy.

    x.meth(..)

The typical method invocation, shown above, uses a constant
name. As with attributes, this name is contained via
reference in the byte code. It is extended in a fashion
that is compatible with attribute references:

    _0: "meth"
    _1: x.__class__, initially None
    _2: the last value of x.__class__.__methodChanges__
    _3: the last value of "meth", i.e., a method pointer

To find the method, the Python machine performs the
following operation:

    if (x.__class__==_1)
    and x.__dict__.isIndexable \
    and (x.__class__.__methodChanges__==_2):
        method = _3
    else:
        <apply existing algorithm to set method>
        if method.__qualified__:
            _1 = x.__class__
            _2 = _1.__methodChanges__
            _3 = method

In the good case, the method is already cached, and it
is verified with just a few comparisons. Note that
__methodChanges__, like __class__ should be implemented
more directly than through a namespace dictionary.


[1] The Lowis paper is at:
http://www.foretec.com/python/workshops/1998-11/proceedings.html

Russell


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


From ludwig_callewaert at frontierd.com  Fri Dec  8 10:34:00 2000
From: ludwig_callewaert at frontierd.com (Ludwig Callewaert)
Date: Fri, 08 Dec 2000 16:34:00 +0100
Subject: Q for regexp experts
References: <yv2aea77xw2.fsf@lionsp093.lion-ag.de> <3A30FD3D.E3DA4639@frontierd.com>
Message-ID: <3A30FF68.169BB9C@frontierd.com>

Ludwig Callewaert wrote:
> 
> Harald Kirsch wrote:
> >
> > Even with perl-compatible regular expressions it seems to be
> > impossible to specify a pattern which matches
> >
> >    `the longest substring not containing separator' SEP
> >
> > Put another way, from a given starting point, the pattern shall match
> > everything until just before the next occurence of SEP. A first
> > attempt is something like
> >
> >    [^SEP]*
> >
> > but this disallows S', 'E' and 'P' in between the separators.
> >
> > I tried something like (.(?!SEP))+ but it somehow does not work. Is it
> > at all possible to get this to work with regular expressions?
> >
> >   Harald Kirsch
> >
> > --
> > ----------------+------------------------------------------------------
> > Harald Kirsch   | kirschh at lionbioscience.com | "How old is the epsilon?"
> > LION Bioscience | +49 6221 4038 172          |        -- Paul Erd?s
> 
> What about:
> 
> set s "the longest matching String not containing SEP"
> 
> regexp -- ".*(?=SEP)" $s match
Ooops,

forgot the non-greedy part


regexp -- "(.*?)(?=SEP)" $s match

> 
> puts $match
> 
> > the longest matching String not containing
> 
> Ludwig
> 
> --
> 
> Ludwig Callewaert       e-mail: ludwig_callewaert at frontierd.com
> Senior Software Engineer        Frontier Design, Belgium
> __________________________________________________________________

-- 

Ludwig Callewaert       e-mail: ludwig_callewaert at frontierd.com
Senior Software Engineer        Frontier Design, Belgium
__________________________________________________________________


From scarblac at pino.selwerd.nl  Tue Dec 19 05:13:42 2000
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 19 Dec 2000 10:13:42 GMT
Subject: __repr__ policy problem: should module be prefixed to output?
References: <3A3A3A6D.2C3B1A91@sbc.su.se> <91dfl0014sa@news1.newsguy.com> <Pine.LNX.4.10.10012152140170.10103-100000@kev.ioc.ee> <Pine.LNX.4.10.10012160011400.14411-100000@kev.ioc.ee> <slrn93tjv5.h7.amk@207-172-146-173.s173.tnt3.ann.va.dialup.rcn.com>
Message-ID: <slrn93ud1g.5ot.scarblac@pino.selwerd.nl>

A.M. Kuchling <amk at mira.erols.com> wrote in comp.lang.python:
> On Sat, 16 Dec 2000 00:41:59 +0200, Pearu Peterson <pearu at cens.ioc.ee> wrote:
> >Are there any better solutions to the repr string dilemma?
> 
> Yes: remove the text from the Tutorial (or wherever it is) that says
> that eval(repr(obj)) == obj.  It's true for strings, integers, and
> floats; it's not true for instances:
> 
> >>> class C: pass
> ...
> >>> c=C() ; print repr(C)
> <class __main__.C at 0x810af9c>
> >>>

But it *couldn't* be true for classes. And although it's possible to have
the instance return "C()", that wouldn't necessarily recreate an object of
the same value, and that problem can't be solved automatically.

> Practically no objects are so simple that the repr() could encapsulate
> all their state; how can repr(file_object) produce something useful?

File objects can't. The rule only states that it should produce something
useful if it can.

> If you really want to represent an object as a string for future
> recreation, pickle it.

You can't inspect and change pickles by hand.

> Frankly, the only place I ever use repr() is in debugging print
> statements and log files, where I need to be very clear about what an
> object is and avoid confusing the integer 1 with the string '1'.  I
> therefore write __repr__ on classes to be as useful for debugging
> purposes as possible; it doesn't recurse into subobjects, but does
> include the class name, the id() of the object, and any other unique
> identifier for the object (an ID number, user name, whatever).
> 
> To put my money where my mouth is, I'd like to file a documentation
> bug about this.  What bits of the Python documentation promote this
> view of repr()?  One is section 3 of the reference guide, in its
> description of the __repr__ special method name; any others?

What is the bug, precisely? You never use it?

The language reference has it when in the description of __repr__, and
the library reference in the description of repr().

This is only really an issue with instances, since you get to make your
own __repr__ for those. And in my experience, it is usually quite possible
to return a string that could be used to build the object again (for instance,
by giving the right arguments to the constructor). Of course, if you
have objects with lots of subobjects and so on, it may not be feasible
and you don't use it.

But I still like the rule. And, well, it's what repr does.

-- 
Remco Gerlich


From aahz at panix.com  Mon Dec 25 16:40:51 2000
From: aahz at panix.com (Aahz Maruch)
Date: 25 Dec 2000 13:40:51 -0800
Subject: __init__ question
References: <EvM16.270$LU6.118037@typhoon.aracnet.com> <9285oj$6rh$1@panix2.panix.com> <t0P16.273$LU6.119021@typhoon.aracnet.com>
Message-ID: <928et3$9sd$1@panix3.panix.com>

In article <t0P16.273$LU6.119021 at typhoon.aracnet.com>,
Daniel Klein <DanielK at aracnet.com> wrote:
>
>Is this part of the syntax or is this a regular expression? I can't find
>this anywhere in the online documentation or in any of the four Python books
>I have purchased. :^(

http://www.python.org/doc/current/ref/function.html
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

'"Crisp" is a good quality for crackers; less so for pot roast.'  --pnh


From aleaxit at yahoo.com  Tue Dec  5 16:11:08 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 5 Dec 2000 22:11:08 +0100
Subject: "inherited" keyword in Python?
References: <3A2B9EFA.3A64C13B@schlund.de> <90gevs02uu1@news2.newsguy.com> <m2elznipbl.fsf@mycroft.actrix.gen.nz> <90ievt025bl@news2.newsguy.com> <m266kzhua6.fsf@mycroft.actrix.gen.nz>
Message-ID: <90jlrk012u7@news1.newsguy.com>

"Paul Foley" <see at below> wrote in message
news:m266kzhua6.fsf at mycroft.actrix.gen.nz...
    [snip]
> > I think your cpl function is much more complicated than what's needed
> > to reproduce Python's very simple rule for walking the DAG of base
> > classes (it's just depth-first, left-first), even assuming it does
> > end up reproducing it perfectly.  But that's not the main point.
>
> It's probably much more complicated than it needs to be to do what it
> does, but it's not intended to reproduce Python's rule exactly -- that
> rule doesn't allow for a correctly-functioning "super" function in the
> presence of multiple inheritance.

Aha, I think I see what you mean.  But I also think you're wrong!

> My cpl() uses the CLOS rule, which is similar to Python but not
> broken.  The entire class hierarchy is sorted in such a way that
> classes appearing to the left of other classes in __bases__ appear
> before them in the sorted list, every class appears before any of its
> bases, and each appears exactly once.  Every method is reachable, not
> just the ones up the left branch, as in Python.

To clarify: Python's rule (as applied by, e.g., getattr) is anything
but broken, but it doesn't easily let you return (for a given class X
that inherits from A, B, C, ...) a single class Z that will ensure that
getattr(Z,whatever) is the same as getattr(X,whatever) for any
whatever that is not a key in X.__dict__.

Solution: generate an artificial class W with an empty __dict__
and exactly the same bases as X!

import new
def super(klass):
    return new.classobj('super_%s'%klass.__name__,
        klass.__bases__, {})

Of course, you can cache this as klass.__cached_super or whatever,
just as you did.


> > A more interesting theoretical question is how you propose to solve
> > the case in which you can't tell which class the current method
    [snip]
> If all else fails, take the easy way out and specify it in the call to
> super()

As I did above, yes.  But the 'convenience' rather takes a hit there.


> > class A:
> >     def foo(self):
> >         print super()
>
> At first reading, I didn't understand this: I thought it should be an
> error, since there are no superclasses.  Reading further, I understand
> that you expect super() to return one of the base classes.  Of course,
> that won't work.  I thought super() was just supposed to call the
> appropriate same-named method of a parent class.  Anything else can be
> done through the usual self.whatever, surely?

Not really.  To translate literally a Java class which has methods X
(which calls the base's Y) and Y (which calls the base's X method),
I'd need a Python's super() to be like Java's super (returning a class
object reference) and not like what I name 'supercall'.  self.X and
self.Y just won't do for this.


> > class B(A):
> >     def foo(self):
> >         print super()
> >         A.foo(self)
>
> > class C(A):
> >     def foo(self):
> >         print super()
> >         A.foo(self)
>
> > class D(B,C):
> >     def foo(self):
> >         print super()
> >         B.foo(self)
> >         C.foo(self)
>
> In this case, the CPL would be [D B C A], super() in D.foo() would
> call B.foo(), whose call to super() would call C.foo(), whose call to
> super() would call A.foo(), whose call to super() would signal an error.

I would have expected super() to return None for a bases-less class,
but it will now return (with my implementation above) an artificial
class with no bases and an empty dictionary.  So, my code above
would work, though not very usefully:-).


> > which reduces it to something decently-simple, like:
>
> > def supercall(_obj, _cls, _metnam, *args, **kwds):
> >     for _bas in _cls.__bases__:
> >         _metobj = getattr(_bas, _metnam, None)
> >         if not _metobj is None:
> >             return _metobj(_obj, *args, **kwds)
> >     else:
> >         raise NameError, "No superclass of %s has method %s" % (
> >             _cls, _metnam)
>
> > But, how often is it going to be satisfactory to call 'just
> > the one' (of several) inherited-method of a certain name, in
> > a multiple-inheritance context?  Just as often we may want
>
> It isn't (at least, not to me), which is why I avoided doing it that

But you did do that -- you call just one of the several 'inherited'
implementations of the method, in the code below.

> way.  Having figured out how to get the frame info, here's an
> implementation.  No need to pass in the object and method name:

I'd still want to pass them, at least optionally, to handle the
above case where I want to call a _different_ inherited method.
They can default to None to ask them to be obtained via
introspection, of course, if one wants that.

> If you just call super(__thisclass) with no args other than the class,
> it passes on the args the calling method got; or you can specify other
> args explicitly [yes, I know this prevents you calling with no args if
> super's caller has args]

As, once again, convenience gets in the way of simplicity and
generality.  Oh well:-).  "Tantum convenientia potuit suadere
malorum", as Lucretius might write if he lived today!-)


Alex





From aleaxit at yahoo.com  Sat Dec  9 07:57:54 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 9 Dec 2000 13:57:54 +0100
Subject: Python wrapper for C++ STL?
References: <3A31760E.D157EB54@erols.com> <C0gY5.126062$U46.4232859@news1.sttls1.wa.home.com> <90s3fk$j8p$1@nnrp1.deja.com> <3A31A05C.6A853969@erols.com>
Message-ID: <90taaj02qjj@news2.newsguy.com>

"Edward C. Jones" <edcjones at erols.com> wrote in message
news:3A31A05C.6A853969 at erols.com...
> shindich at my-deja.com wrote:
>
> > With all due respect, what would be wrong with a definition like
> > std::map<PyObject*, PyObject*>? Obviously there is no type safety here.
> > But Python is not type safe.
>
> This seems correct. It would give me a "sorted associative container", a
> dictionary I could loop through.

Hmmm, yes, but, take care: the 'loop-through' feature had better be
exposed through a specially-named method or attribute -- I think it
would be better if .keys() returned a normal Python list, as a caller
would expect from normal polymorphic behavior, and while it's nice
that this list is sorted rather than randomish'ly unordered, it can be
a substantial overhead to build it.

And you can't just use:
    for key in myFuncyObject:
        whatever(key)
as that would require the object to be indexable by 0, 1, ..., n with
the meaning of 'loop through it', interfering with the normal meaning
(as these integers can be perfectly normal dictionary-keys!-).

The design pattern to implement the indexability thus requires an
auxiliary object (and a little bit of care in it, since Python for-loop
protocol is based on a growing int index, and you have to mutate
that, internally, into a sequential 'current iterator'/'increment it'
pattern).  No rocket science, mind you, but it's not quite as obvious
as _just_ 'wrapping a std::map'...:-).


Alex





From rhamorim at my-deja.com  Wed Dec  6 13:15:26 2000
From: rhamorim at my-deja.com (rhamorim at my-deja.com)
Date: Wed, 06 Dec 2000 18:15:26 GMT
Subject: Finding values on sequences and .index()
Message-ID: <90lvnm$k2a$1@nnrp1.deja.com>

Hi.

I have a list that may have many ocurrences of the same data. How do I
find the indexes for all ocurrences of it? list.index(value) only works
for the first ocurrence.

I thought of this ugly hack below, but I guess there's an easier way.

a = [1, 2, 3, 4, 1, 5, 6, 7, 1, 8, 9, 1]
b = map(None, a, range(len(a)))

N = 1 #value to be searched

c = filter(lambda x, n=N: x[0]==N, b)
d = map(lambda x: x[1], c) #this is the returned index list

Thanks in advance...

Roberto Amorim


Sent via Deja.com http://www.deja.com/
Before you buy.


From aleaxit at yahoo.com  Tue Dec  5 16:19:46 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 5 Dec 2000 22:19:46 +0100
Subject: "inherited" keyword in Python?
References: <3A2B9EFA.3A64C13B@schlund.de> <90gevs02uu1@news2.newsguy.com> <1742065242.20001205223907@vega.bg>
Message-ID: <013f01c05f01$2e17eda0$518c1497@martelli>

Hi Niki,

> #class C(A,B):
> #...
> 
> class C_help(A,B):
>   pass
> 
> class C(C_help):
>   def foo( self ):
>     C_help.foo( self )
> 
> I would like to have same functionality without writing X_help class
> whenever i need it.

I posted a solution based on just this idea 1 minute ago!-)

The complete version, which should work but I haven't tested:

import new
def super(klass):
    try: return klass._cached_super
    except:
        return klass.__dict__.getdefault(
            new.classobj('super_%s'%klass.__name__,
            klass.__bases__, {})


Alex



__________________________________________________
Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.
http://im.yahoo.com



From cjensen at bioeng.ucsd.edu  Thu Dec  7 17:11:41 2000
From: cjensen at bioeng.ucsd.edu (Curtis Jensen)
Date: Thu, 07 Dec 2000 14:11:41 -0800
Subject: Parallel Computing
Message-ID: <3A300B1D.223630F3@bioeng.ucsd.edu>

I haven't had much luck on finding anything about parallel computing via
Python.  Does Python support this?  Or do we just need to write a C
module to run parallel code?  Thanks.

-- 
Curtis Jensen
cjensen at bioeng.ucsd.edu
http://www-bioeng.ucsd.edu/~cjensen/
FAX (425) 740-1451


From p.g at figu.no  Tue Dec  5 09:29:12 2000
From: p.g at figu.no (Per Gummedal)
Date: Tue, 5 Dec 2000 15:29:12 +0100
Subject: problems with installer 20_3i
In-Reply-To: <8FFED7BE9gmcmhypernetcom@199.171.54.155>
References: <_o%V5.47439$yR4.1041961@news1.rdc1.tx.home.com>
 <8FFED7BE9gmcmhypernetcom@199.171.54.155>
Message-ID: <2507146668.20001205152912@figu.no>

03.12.2000, 03:30, Gordon wrote:

> Untested: at the top level of archive_rt (after the imports), put:
> sys.path.append(imputil.BuiltinImporter())

> - Gordon

This did help in my case.

Thanks.

Per





From moshez at zadka.site.co.il  Sun Dec 31 10:13:53 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Sun, 31 Dec 2000 17:13:53 +0200 (IST)
Subject: Python vs. PHP (& Java?)
In-Reply-To: <92id18$js4$1@panix2.panix.com>
References: <92id18$js4$1@panix2.panix.com>, <69BC4C6AE83ED311BC9400805FA7B14AF84D2C@il93exp01.css.mot.com> <mailman.978059571.14515.python-list@python.org> <397C7FA100EE65A5.0FF0B81DCC646CA3.C3E3EFFAF818CC30@lp.airnews.net>
Message-ID: <20001231151353.28CCCA84F@darjeeling.zadka.site.co.il>

[Cameron Laird (about Heisenberg's Law)]
> Metaphoric application to macroscopic phenomena is, as Moshe notes,
> often overdone.

[Aahz Maruch]
> And in some cases, it's underdone as a metaphor.  I remember reading an
> article recently about how poor "focus groups" are for getting useful
> results.  One of the examples given was about how the experimenters were
> watching one focus group that was giving them bizarre results; the
> experiementers were behind one-way glass.  As the test group filed out
> of the room, one of the troublesome subjects turned to the one-way
> glass, smiled, and waved.

No, that's exactly the kind of examples I meant had nothing to do 
with Heisenberg, even as a metaphore. Heisenberg states that when you're
measuring one quantity, you're changing *another* quantity. There
are useful macroscopic metaphores. Consider a phone-conducted poll,
where the caller asks you for half an hour which brands of products
you like, and then tries to see if you're a relaxed person or not.
Naturally, you'll be exasperated. That would entitle you to say
the poll was Heisenberged -- the more exactly they'll want to know
about your consuming habits, they'll more they'll change your state
of mind.

then-again-i-see-quantum-mechanics-in-a-different-way-ly y'rs, Z.
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From pete at visionart.com  Wed Dec  6 15:23:00 2000
From: pete at visionart.com (Pete Shinners)
Date: Wed, 6 Dec 2000 12:23:00 -0800
Subject: mp3 module
References: <3A2BB89B.D4C30300@ira.uka.de>
Message-ID: <90m798$cfj$1@la-mail4.digilink.net>

"Markus von Ehr" <vonehr at ira.uka.de> wrote
> is there a module to play mp3 files with python?

it's probably a bit overkill, but if nothing else crops up...

the pygame module can playback mp3 music quite easily.
it requires a couple dependencies, but you can grab the
full windows binary for an easy install. (RPMs also avail)

import pygame.mixer, pygame.time
pygame.mixer.init(44100)
music = pygame.mixer.music
music.load('mymusic.mp3')
music.play()
while music.get_busy(): #still playing
    pygame.time.delay(1000)


anyways. pygame might be a bit overblown for your
situation, but it can do the job. other than that you
might want to look into Snack(or is it Smack?). there is
also a new wrapper for openAL. i'm not sure if either
of these play MP3, but they are more audio specific.

pygame: http://pygame.seul.org
openal: http://213.93.58.229/alpy/





From etoffi at bigfoot.com  Sat Dec 23 16:34:28 2000
From: etoffi at bigfoot.com (etoffi at bigfoot.com)
Date: Sat, 23 Dec 2000 21:34:28 GMT
Subject: python and cgi
Message-ID: <9235p5$ngo$1@nnrp1.deja.com>

hello, fellow pythoners!

i am writing an application that will communicate between two systems
using http.  note that i am not an expert in http (or python ;)

i need a way to "upload" using the HTTP POST method (ie: simulating a
browser).

basically, my app has two parts: the actual cgi script, and a backend script
that must stay alive for a short amount of time (~20 min). the problem i have
is how can these 2 scripts communicate in a cross platform manner?  i am
running the app on windows and unix, both with stock apache builds (no
mod_py, etc). is this possible, given that the cgi will always run with a new
interp. process and the backend will be using the same one until it times
out?

while i'm here i might as well mention that i am interested in soap-ly-yrs.

(snicker. hope that line is correct. snicker again.)

--
etoffi


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


From greg at cosc.canterbury.ac.nz  Thu Dec 14 21:01:34 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Fri, 15 Dec 2000 15:01:34 +1300
Subject: [Q] Are Exceptions used that much in practice?
References: <3A35D546.A4731B5B@enteract.com> <914mfd02ibt@news2.newsguy.com> <3A36F20B.7B766D0@cosc.canterbury.ac.nz> <uaea0ghlx.fsf@ctwd0143.fitlinxx.com> <jon.976778599@bezek.dstc.monash.edu.au>
Message-ID: <3A397B7E.A4D2F60D@cosc.canterbury.ac.nz>

Jonathan Giddy wrote:
> 
> Note that now that the standard exceptions are classes, IOError instances
> have an errno attribute, which is even nicer than using tuple values.

Yeah, except that the error numbers it gives you
seem to be platform-dependent.

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


From pearu at cens.ioc.ee  Fri Dec 15 09:53:48 2000
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Fri, 15 Dec 2000 16:53:48 +0200 (GMT-2)
Subject: How to determine what called __cmp__ method?
In-Reply-To: <200010231941.MAA03719@rocket.knowledgetrack.com>
Message-ID: <Pine.LNX.4.10.10012151643300.32599-100000@kev.ioc.ee>

On Mon, 23 Oct 2000, Huaiyu Zhu wrote:

> In comp.lang.python, you (David A.) wrote:
> 
> >Now, with rich comparisons.. =)  (If anyone is interested in taking on the
> >work of creating the rich comparisons patches, pushing the PEP through,
> >etc., get in touch with me).
> 
> I'm interested in this, as I need it for numerical computation as well.  I
> thought I sent you an email before, which might be lost somewhere.
> Do you already have a draft of the PEP in any form?  Otherwise I can start
> an initial draft listing the features I have in mind, which may motivate
> others to add to it or even implement it.  

How is this PEP doing?

> Right now the main feature I have in mind is to allow partial orders, so
> that a!=b does not imply a<b or a>b.  I also want the comparison operators
> to have their own names, like __lt__, __ge__, for example.

What I am looking for is that __lt__, __ge__, and other related methods
are to be allowed to return arbitrary Python objects, not just Python
integers 0,1. Or is this already obvious requirement for rich comparions?

Pearu





From Bjoern.Giesler at gmx.de  Tue Dec  5 10:44:16 2000
From: Bjoern.Giesler at gmx.de (Bjoern Giesler)
Date: 5 Dec 2000 15:44:16 GMT
Subject: Smalltalk/ROSE-like prog. environment?
Message-ID: <90j2gg$kpf$1@news.rz.uni-karlsruhe.de>

Hi,

I have a dream... :-)

I'd like to have a programming environment that allows you to do graphical
OO design (UML, ROSE-like) with access to all classes a project consists of
(Smalltalk-like). This thing would allow the programmer to graphically
design class interfaces and interactions, and edit method source code on a
per-method basis. I.e. no editing-around in large .py files but working just
on the job at hand.

This would have a couple of other nifty features: A test mode in which an
exception is presented to the user in a dialog box that allows modification
of the code _while the program is running_ (this could be done with
continuations, I believe) and stuff like that.

This is a very hazy notion that has been floating around in my head for a
while. I believe that Python would be the ideal language for this kind of
application. Does anybody know of such an environment?

If not, I'd like to give it a shot. Anybody have any ideas or want to
collaborate with me?

Regards,
		--Bj?rn


From rpgarchive at my-deja.com  Wed Dec 27 01:11:30 2000
From: rpgarchive at my-deja.com (rpgarchive at my-deja.com)
Date: Wed, 27 Dec 2000 06:11:30 GMT
Subject: zipfile.py
References: <3A37AD3E.7233E7B7@ex.ac.uk> <3A37B917.6AC49FEA@interet.com>
Message-ID: <92c16g$m35$1@nnrp1.deja.com>

I get this same problem, but only when using zip
files that have been created on Linux,
using 'zip'.

In article <3A37B917.6AC49FEA at interet.com>,
  "James C. Ahlstrom" <jim at interet.com> wrote:
> "M.J.Fisher" wrote:
> >
> > Hi,
> >
> > I have been having some trouble with using
zipfile.py, I need to unzip
> > .zip files some created in windows and
> > some in unix.
> >
> > zlib.error: Error -3 while decompressing:
incomplete dynamic bit lengths
> > tree
> >
>
> It looks like there is a version
incompatibility in your
> version of zlib on Windows or Unix.  Zipfile
depends on zlib
> to compress/uncompress files.  You can test
this by specifying
> a compression method of ZIP_STORED for all
files so they are not
> compressed.  Or you could check zlib versions.
>
> Jim Ahlstrom
>



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


From z3penguin at penguinpowered.com  Sun Dec 17 11:11:16 2000
From: z3penguin at penguinpowered.com (Z 3 Penguin)
Date: Sun, 17 Dec 2000 11:11:16 -0500
Subject: cross-platform sound?
Message-ID: <91lq6v$n5m$1@autumn.news.rcn.net>

i know windows has winsound, and i think linux has a different sound library.  is there any function that plays sounds on multiple operating systems in Python, or do i figure out which os the user is running and use the respective function?
=z



From tim.one at home.com  Sat Dec 23 13:14:46 2000
From: tim.one at home.com (Tim Peters)
Date: Sat, 23 Dec 2000 13:14:46 -0500
Subject: A Suggestion for Python Colon Syntax
In-Reply-To: <slrn949doq.ga.amk@207-172-111-138.s138.tnt1.ann.va.dialup.rcn.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEEFIFAA.tim.one@home.com>

[A.M. Kuchling]
> Two errors in a timbot posting; time to dump him to disk and restart
> from a fresh image, I think...

That's what the end of the year is for -- it's scheduled but delayed due to
license haggles.

>> ... They picked "i". [for imaginary literals]

> They picked "j", not "i".

Actually, about half did pick "i" <wink>.  Guess this proves that I couldn't
care less either.

>> [a claim that no code in the std library ends with ";"]

> [amk at mira Lib]$ grep ';' shlex.py
>         self.pushback = [];
>         self.pushback = [tok] + self.pushback;
>             nextchar = self.instream.read(1);
>                 self.token = '';        # past end of file
>
> Been meaning to fix those...  Ooh, and there's one in multifile.py and
> netrc.py, too.

I stand by my original claim -- especially since you already checked in
repairs for all the cases you found <wink>.

making-a-claim-is-a-good-way-to-make-it-happen-ly y'rs  - tim




From brianknox at home.com  Sun Dec 24 10:48:01 2000
From: brianknox at home.com (Brian Knox)
Date: Sun, 24 Dec 2000 15:48:01 GMT
Subject: Python Database Interfaces - Any Standard?
Message-ID: <RUo16.204045$RP.37382354@news1.rdc1.md.home.com>

I've been programming Perl for three years and I'm recently switching over
to Python for a project at work that is going to be JPython based. I've been
pooring over Learning Python and Programming Python (Oreilly) and I'm liking
a lot of what I'm seeing. Working with classes under Python looks to be a
pleasure and I can't wait to start digging in. I love Perl but OOP in Perl
isn't exactly clean and friendly.

One huge question I have about Python is: is there any sort of standard for
database access with Python? I've been coding Perl DBI apps as long as I've
been coding perl, and the DBI module is going to be sorely missed by me. Are
all of the Python database interfaces database specific, or is there an
abstraction layer similiar to DBI? Is there any one DB module for Python
that has been adopted as a standard by the community?

Brian Knox
Development Environment Engineer
Techtrader.com




From guido at digicool.com  Wed Dec 27 16:09:57 2000
From: guido at digicool.com (Guido van Rossum)
Date: Wed, 27 Dec 2000 16:09:57 -0500
Subject: ANNOUNCE: soundex.py
In-Reply-To: Your message of "Fri, 22 Dec 2000 15:21:19 EST."
             <LNBBLJKPBEHFEDALKOLCCECJIFAA.tim.one@home.com> 
References: <LNBBLJKPBEHFEDALKOLCCECJIFAA.tim.one@home.com> 
Message-ID: <200012272109.QAA22296@cj20424-a.reston1.va.home.com>

> [A.M. Kuchling]
> > Does this mean that Modules/soundex.c can be dropped from the Python
> > CVS tree?
> 
> Alas, I don't think so, unless we get Special Dispensation from Guido.
> Until PEP 5 is accepted, we don't have a well-defined way to get rid of a
> module.  The distributed 2.0 docs call the soundex module "obsolete", but
> don't say anything about it ever going away.  While the details of PEP 5
> aren't all agreed upon yet, I believe most people agree with the basic
> notion that modules can't go away without an explicit warning about that
> first.

Not true.  Modules have gone away at each of the recent releases:
Python code gets moved to Lib/old, and C code simply gets dropped
(e.g. stdwin), after having been disabled in the Setup file in a
previous release.

Since soundex has been labeled obsolete for a while, it's fine to drop
the file.

Note, this kind of stuff needs to be decided on a per case basis:
e.g. we're still keeping regsub and regex around (I've now added
warnings, so as of 2.1 it is strongly deprecated).

--Guido van Rossum (home page: http://www.python.org/~guido/)



From darrell at dorb.com  Fri Dec  8 22:19:27 2000
From: darrell at dorb.com (Darrell Gallion)
Date: Fri, 8 Dec 2000 22:19:27 -0500
Subject: parsing and searching big text files
References: <Pine.LNX.4.21.0012061820200.7564-100000@ns.sopron.hu>
Message-ID: <13db01c0618e$d6f64d10$6401a8c0@home>

Also untested:

def scanGaborsFile(file, scanFor='John Doe', seperator=':'):   
    buf=open(file).read()
    res=re.findall("(?:^|\012)(.*%s.*)"%scanFor,buf)
    return [ i.split(seperator) for i in res]
    
--Darrell


----- Original Message ----- 
From: "Gabor Gludovatz" 

> I have a big text file which contains 3 variable lenght columns with
> names. I have to search this file for names in either column and have to
> show to other 2 names, and I have to do this very fast!
> 
> The text file is about 3 megs long. Does someone know a method to do this
> quickly?
> 
> Which is the faster, should I read to whole text file into the memory and
> parse and search it there, or should I read it from the disk line by line
> and parse it..?
> the last seemed to be very slow.
> 
> Here is an example line from this text file:
> Foo Bar:John Doe:Bill
> 
> If I look for, for example, John Doe, the function should return Foo Bar
> and Bill and all the other records which contain John Doe.
> 
> Again: I have to do it very fast.
> 
> 
> -- 
>  Gabor Gludovatz <ggabor at sopron.hu> http://www.sopron.hu/~ggabor/
> 
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list




From jochen at pc1.uni-duesseldorf.de  Wed Dec  6 18:35:10 2000
From: jochen at pc1.uni-duesseldorf.de (Jochen =?iso-8859-1?q?K=FCpper?=)
Date: 07 Dec 2000 00:35:10 +0100
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <3A2DD2CA.14A5D7E2@engcorp.com> <90l3j9$4bn$1@kapsel.intranet>
Message-ID: <udzoi9cedt.fsf@box.home.de>

>>>>> Piotr Kuchta wrote on Wed, 6 Dec 2000 11:15:05 +0100:

Piotr> that's true that Dutchmen are very talented: common Dutchman
Piotr> speaks well two foreign languages, English and German.

Isn't that simply what Dutch is all about? 
Where's the foreign language;-))

Greetings,
Jochen

PS: Yes, most Dutch people are much better at foreign languages than I am!
-- 
Einigkeit und Recht und Freiheit                http://www.Jochen-Kuepper.de
    Libert?, ?galit?, Fraternit?                GnuPG key: 44BCCD8E
        Sex, drugs and rock-n-roll


From hgg9140 at seanet.com  Sat Dec  9 17:35:14 2000
From: hgg9140 at seanet.com (Harry George)
Date: 09 Dec 2000 14:35:14 -0800
Subject: UML and Python
References: <90o91k$eks$1@desperado.vew-telnet.net> <90stth$mea$1@newsreaderg1.core.theplanet.net>
Message-ID: <m3aea5te8t.fsf@wilma.localdomain>

You could also consider Dia.  It doesn't (yet) have a model repository
behind it, but you can draw all the diagrams, and can output them in
XML so you can generate code or whatever (and dia2code addresses C and
Java).  I've used that to generate python code.

Come to think of it, a model repository for Dia sounds like a good
python project for the holidays.  First hookup would be via the XML
files, then a direct plugin.


"Franz GEIGER" <fgeiger at datec.at> writes:

> Hi Rainer,
> 
> I'm currently lookin' for a decent priced UML IDE. All C++ UML tools are
> rather expensive. Do you have already any experiences on ObjectDomain?
> 
> Best regards
> Franz GEIGER
> 
> 
> Rainer Rasche <rrasche at dspace.de> schrieb in im Newsbeitrag:
> 90o91k$eks$1 at desperado.vew-telnet.net...
> > Hi,
> >
> > is somebody out there, who is interested in developing python applications
> > with UML-CASE Tools (e. g. ObjectDomain) and wants to exchange some
> > experiences with me?
> >
> > Thanks, Rainer
> >
> > rrasche at dspace.de
> >
> >
> 
> 

-- 
Harry George
hgg9140 at seanet.com


From hungjunglu at hotmail.com  Wed Dec 13 16:14:24 2000
From: hungjunglu at hotmail.com (Hung Jung Lu)
Date: Wed, 13 Dec 2000 21:14:24 -0000
Subject: Mapping Python objects and XML
Message-ID: <918org+7dl5@eGroups.com>

Hi,

If I need to map Python objects into XML structure and vice-versa, is 
there any Python class I can use as base class?

Mapping Python objects into XML probably is not too hard, if the 
objects don't contain things like object-bound methods. How about the 
reverse process? I think I can include the class tag in the XML, so 
during the re-creation process, I can call the appropriate 
constructors for the objects. I think a base Python class can be 
created to handle this kind of things, if the data fields are not too 
exotic (e.g: numbers, strings) This must be a very common need. So I 
am wondering whether this base class is already available.

(By the way: anything similar in Java?)

Thanks!

Hung Jung





From aleaxit at yahoo.com  Thu Dec 28 18:50:35 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 29 Dec 2000 00:50:35 +0100
Subject: One Python 2.1 idea
References: <mailman.978038049.481.python-list@python.org>
Message-ID: <92gjnu02q3r@news2.newsguy.com>

"Tim Peters" <tim.one at home.com> wrote in message
news:mailman.978038049.481.python-list at python.org...
    [snip]
> > What are the 'good Pythonic ways' to do this, as opposed to
> > 'dirty tricks'?
> > ... [assorted trickery with the "new" module] ...
>
> class UniqueOgreFoo(Ogre):
>     foo = fun
> x.__class__ = UniqueOgreFoo
>
> x.foo() # now prints 'have fun!'

Right.  So, in the general case (where the instance to 'uniquize',
the methodname to uniquize it for, and the function-object to
use for the purpose, are all runtime-determined, rather than
constants as in this toy-example), it would be:

def uniquize(instance, methodname, functionobject):
    class Uniqued(instance.__bases__):
        pass
    setattr(Uniqued, methodname, functionobject)
    instance.__class__ = Uniqued

rather than

def uniquize(instance, methodname, functionobject):
    method = new.instancemethod(functionobject,
        instance, instance.__class__)
    setattr(instance, methodname, method)

right?  OK, I'll take your word for it -- they seem about equally
'dark' to me, and the former surely has the no-refloop advantage,
after all.  But, apart from the historical accident that 'new' is,
well, new, wouldn't the latter be normally more Pythonic, i.e.,
less obscure?  The former relies on the magic and implicit
function->method conversion that classes do, the latter makes
the conversion-to-method quite explicit.


> obscurity of the code is a bigger issue in practice.  Not that mucking
> around with __class__ is recommended to begin with!  That's obscure too,
> just not as obscure as new.*.  Best would have been to make x an instance
of
> UniqueOgreFoo from the start.

Sure!  The assumption is that one _can't_ use that simplest
of approaches, because the need to uniquize x that way is
determined dynamically; i.e., that the behavior of x.foo()
must, some way or other, be now made different from what
it used to be before.

If this sort of thing happened a few times to instances of
class Ogre in a typical program run, would it be worth
refactoring Ogre to take this into account -- perhaps through
a suitable metaclass, to check an instance 'uniquization'
dictionary for possible 'uniquized' methods before going
the normal class-dictionary way, or would _that_ be too
obscure too?  My personal perception is that metaclasses
are 'obscurer' than either new, or setting __class__, but
that counts for little, since I don't see that the latter is
any clearer than the former, so I must not be seeing through
properly Pythonic eyes.  Or would a refactoring by more
elementary means, with more explicit delegation, be more
Pythonic?  E.g.:

class Ogre:
    class TypicalOgre:
        [snipped: methods as before]
    prototype = TypicalOgre()
    def __init__(self):
        self.__unique = {}
    def __getattr__(self, name):
        try:
            return self.__unique[name]
        except KeyError:
            Ogre.__temp = getattr(TypicalOgre, name)
            return self.__temp
    def uniquize(self, methodname, functionobject):
        Ogre.__temp = functionobject
        self.__unique[methodname] = self.__temp

Hmmm -- this looks cumbersome to me, and unPythonic
in relying deeply on the class's implicit function->method
magic transformation; not to mention the overhead AND
the reference-loop, plus the fact that, this way, it only
works for one class (or, at best, it can be packaged up as
a mixin), while previous approaches were more general.

So I guess that, if one does have the need for such
dynamic calisthenics, changing __class__ is still best?


> give-a-man-a-fish-and-you've-fed-him-for-a-day-but-teach-a-man-to-
>     fish-and-he'll-come-by-every-day-asking-for-bait-lures-spare-
>     lines-and-your-best-pole<wink>-ly y'rs  - tim

Right!  No good deed goes unpunished, remember... (why, I'll even
tolerate the <winks> in exchange for your precious lessons!-).


Alex





From jwmdck at monmouth.com  Sat Dec 23 13:21:38 2000
From: jwmdck at monmouth.com (Murphy/Kostick)
Date: Sat, 23 Dec 2000 13:21:38 -0500
Subject: Reality check PYZOMA = python / zope for  Mac ?
Message-ID: <3A44ED33.E203484C@monmouth.com>

 
-------------- next part --------------
An embedded message was scrubbed...
From: Bob Newhart <smegtra at goodnet.com>
Subject: Python & Zope Work!
Date: Sun, 15 Oct 2000 14:54:10 -0700
Size: 3210
URL: <http://mail.python.org/pipermail/python-list/attachments/20001223/28cf0c46/attachment.mht>

From pblanchette at pixelsystems.com  Wed Dec  6 14:17:32 2000
From: pblanchette at pixelsystems.com (Patrick Blanchette)
Date: Wed, 06 Dec 2000 19:17:32 GMT
Subject: win32api SetComputerName ?
Message-ID: <3A2E90BF.4020204@pixelsystems.com>

I,
   It seems that the function SetComputerName doesn't exist in the 
win32all-135.exe release but the function GetComputerName exist.  I'd 
like to know if the method SetComputerName will be added soon? 
   thanks



From hwanjoyu at students.uiuc.edu  Sat Dec 30 00:18:30 2000
From: hwanjoyu at students.uiuc.edu (hwan-jo yu)
Date: Fri, 29 Dec 2000 23:18:30 -0600
Subject: Q: how to sort a dictionary by the values ?
Message-ID: <Pine.GSO.4.10.10012292313350.21787-100000@ux11.cso.uiuc.edu>

Hi,
If I sort the items() of dictionary, it is sorted by first the key and
second the value.
How can I sort it by the values ?
Thanks in advance.




From shindich at my-deja.com  Fri Dec  8 22:43:32 2000
From: shindich at my-deja.com (shindich at my-deja.com)
Date: Sat, 09 Dec 2000 03:43:32 GMT
Subject: Unit Testing tool
References: <90s69s$f85$1@troll.powertech.no>
Message-ID: <90s9p4$nk8$1@nnrp1.deja.com>

In article <90s69s$f85$1 at troll.powertech.no>,
  "Syver Enstad" <syver at NOSPAMcyberwatcher.com> wrote:
> What is the Python unit testing tool preferred by readers/writers from
> comp.lang.python?
>
I have successfully used http://pyunit.sourceforge.net/.

Regards,

Alex Shindich
visit http://www.shindich.com/


Sent via Deja.com http://www.deja.com/
Before you buy.


From peter at engcorp.com  Tue Dec 12 23:26:33 2000
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 12 Dec 2000 23:26:33 -0500
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <915jcf0c6o@news2.newsguy.com> <9162p3$7t2$1@animus.fel.iae.nl>
Message-ID: <3A36FA79.C4D0E0A6@engcorp.com>

Carel Fellinger wrote:
> 
> Alex Martelli <aleaxit at yahoo.com> wrote:
> 
> > Firmitas, Utilitas, Venustas -- in *THIS* order, as Vitruvius
> 
> No no, Beauty is the one and true guiding principle, Beauty and Simplicity.
> The trouble being that what one perceives as beauty has ever so much to do
> with ones upbringing. And elas, there are very few with an eye for Beauty.

Sounds like we are rapidly converging on Zen and the Art of Motorcycle
Maintenance: see c2.com/cgi/wiki?QualityWithoutaName, which not just
coincidentally mentions Python and then links indirectly to the
c2.com/cgi/wiki?PythonPhilosophy of Tim Peters, rapidly coming full
circle back to Beauty... ;)

In other words, this ends up a question of semantics: Quality is defined
in different ways to different people and whether you call it stability,
utility, *or* beauty, chances are you are really talking about the same
thing...


From atakli at lycos.com  Thu Dec 28 23:49:34 2000
From: atakli at lycos.com (idris atakli)
Date: Thu, 28 Dec 2000 20:49:34 -0800
Subject: i wanna join the python newsgroup
Message-ID: <GEJGIAJHICHGCAAA@mailcity.com>

 



Get FREE Email/Voicemail with 15MB at Lycos Communications at http://comm.lycos.com



From Johnson  Thu Dec  7 22:01:48 2000
From: Johnson (Johnson)
Date: Thu, 07 Dec 2000 19:01:48 -0800
Subject: Stupid Newbie Question:  Python FAQ?
Message-ID: <3A304F1C.DF1224C2@springhaven.org>

Hello all,

    Chiming in with my first Stupid Newbie Question:  Is there a FAQ
available for python, so that I can avoid asking more Stupid Newbie
Questions? :)

-- 
- Johnson Earls
darkfox at springhaven.org


From aleaxit at yahoo.com  Tue Dec  5 07:37:18 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 5 Dec 2000 13:37:18 +0100
Subject: Calculating factorial
References: <3a2a9496.647995@uutiset.nic.fi> <90ghoh0lh@news2.newsguy.com> <9tkp2tsfc0hjq3dr7vc69a6hdf0at3q0rk@4ax.com>
Message-ID: <90ini102dj0@news2.newsguy.com>

"Steve Horne" <sh at ttsoftware.co.uk> wrote in message
news:9tkp2tsfc0hjq3dr7vc69a6hdf0at3q0rk at 4ax.com...
> On Mon, 4 Dec 2000 17:45:46 +0100, "Alex Martelli" <aleaxit at yahoo.com>
> wrote:
>
> >Attempts to get more sophisticated, such as
> >
> >def factorial(f):
> >    import operator
> >    return reduce(operator.mul, range(1,f+1), 1L)
> >
> >will not necessarily bring happiness.
>
> Might be faster with xrange (which doesn't build the whole list before

xrange tends to be slower than range in all normal cases.
(Lazy is beautiful, but speed still seems to be with eager evaluation,
cfr typical speed of comppiled ML applications vs Haskell ones, sigh).

> returning), but the simple for loop will probably still win for speed.
> Functional-style list manipulations (once you are used to them) are
> for speed of development, not speed of execution.

I'm pretty used to functional programming, but I still don't find

    return reduce(operator.mul, range(2,f+1), 1L)

particularly faster to think-and-write than

    result = 1L
    for x in range(2,f+1): result *= x
    return result

I guess people's mileage varies here!


Alex





From grey at despair.rpglink.com  Wed Dec 27 12:22:30 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Wed, 27 Dec 2000 17:22:30 -0000
Subject: regex, split and ()
References: <slrn94k8rq.18t.jp@localhost.localdomain>
Message-ID: <slrn94k9am.1eq.grey@teleute.rpglink.com>

On 27 Dec 2000 17:14:34 GMT, Jacek Pop?awski <jp at ulgo.koti.com.pl> wrote:
>>>> w=re.compile(r'(<[^<>]*(".*")?>)')
>>>> s="<html> one <br> two <img src=\"<blah\"> three </html>"
>>>> w.split(s)
>['', '<html>', None, ' one ', '<br>', None, ' two ', '<img src="<blah">',
>'"<blah"', ' three ', '</html>', None, '']

>works good, but why it double everything? probably becouse I used second (),
>how to fix it?

    Use ?: in the parens you don't want to retain.


-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From christian at integration-x.com  Fri Dec  8 09:50:15 2000
From: christian at integration-x.com (Christian Wermuth)
Date: Fri, 08 Dec 2000 15:50:15 +0100
Subject: Accessing "variables" in an extension module
References: <3A2E4B01.4609EE7E@integration-x.com> <7tOX5.146$JG.14399@news.shore.net>
Message-ID: <3A30F527.71138023@integration-x.com>

arcege at shore.net wrote:

> ...
> I'm assuming that you want to have predefined variables and constants.
> You might want to look at the section 'Adding "constants"' (which, in
> Python, are really variables anyway) in the "Python Extension writing"
> document.  That might help you.
>
> The URL for the document is
> <URL: http://starship.python.net/crew/arcege/extwriting/pyext.html>
>
> Remember that modules, whether a C extension or not, are mutable and
> you can always add or remove names there.
>
>   -Arcege

Thanks. It was a great help.
Now I can access variables in the "init" section of my C code, but I can't
figure out how to do it from my method. I can set and read the variable
from python, but when I try to access the variable from my C method, I just
get a value of "null". Any idea?

Christian



From arturo_perez at wanadoo.es  Tue Dec 12 14:16:32 2000
From: arturo_perez at wanadoo.es (Arturo Pérez Mulas)
Date: Tue, 12 Dec 2000 19:16:32 GMT
Subject: Fault trees structure
References: <D_IY5.49$17.2785@m2newsread.uni2.es> <9128e902uo3@news2.newsguy.com> <z8eZ5.249$YZ4.8868@m2newsread.uni2.es> <yfhZ5.23310$Ia1.4003773@typhoon.nyroc.rr.com> <91521d02r3g@news2.newsguy.com> <915pu2$6gf$1@news.kodak.com>
Message-ID: <kQuZ5.387$Ds.9939@m2newsread.uni2.es>

Hi,

I have not left the discussion, only I have not yet have the time to read
you guys.
However, on a first look, it seems to me that Darrel's model is only
including 1 or 0
values, while a fault-tree uses probabilities for the limiting nodes (Basic
Events),
and the following arithmetic

Let A and B be nodes of probabilities P(A) and P(B) respectively

then

P(A AND B) = P(A) * P(B)
P(A OR B) = P(A) + P(B) - P(A)*P(B)

And also the following boolean logic (I use '+' for OR and '*' for AND):

A + (A * B) = A
A * (A + B) = A + A*B = A

And for logic '1' and '0', the usual

A + '0' = A
A * '0' = '0'

A + '1' = '1'
A * '1' = 'A'

However, this is not the main goal of the structure I want to develop
(believe me,
probability calculation of a FT is an exponential problem -NP- which
requires much
more evolved calculations methods, and approximations)

Also, you are not considering the 'links' issue (I think, sorry I did not
look at it in detail yet!).

Arturo.

"Darrell Gallion" <darrell at dorb.com> escribi? en el mensaje
news:915pu2$6gf$1 at news.kodak.com...
> Alex Martelli wrote:
> > def connectInput(self, name):
> >     self.__dict__.setdefault('_input',{})[name] = None
> >
> > is a possible alternative (also in other cases), but
> > not very readable.  It would seem best to just set
> > empty dictionaries in __init__ for _input and _output,
> > so as to avoid having test for them at every connect,
> > and possibly having errors lurking elsewhere, too:
> >
> Sounds good.
> This wasn't required once I moved the __init__ into NodeBase by your
> suggestion.
>
> [snip]
> >
> > Of course, giving NodeBase an __init__ means having
> > to call it in lieu, or from, specific subclasses'
> > ones, but that seems OK; right now OrNode and
> > AndNode sport identical __init__ methods, so why
> > not factor them out of the subclasses and up into
> > the NodeBase common baseclass.
> >
> Done.
>
> >
> >
> > >     def logic(self, collection):
> > >         cc= reduce(lambda x,y: x and y, self._input.values())
> > >         self.propagate(collection, cc)
> >
> > If the values are constrained to "boolean" (0 and 1), then
> > using operator.and_ as the first argument to reduce might
> > well be faster, and perhaps more readable (similarly for
> > or-nodes).
> >
> Not sure that values will be constrained to boolean.
>
> When I tried this something odd happened.
> None isn't valid for and_ while it is for "and"
>
> >>> and_(1, 1)
> 1
> >>> and_(None, 1)
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> TypeError: bad operand type(s) for &
> >>> None and 1
> >>>
>
> Made some other changes as well.
> Thanks for taking the time to look this over.
> Maybe I'll have find a use for this code :)
>
> --Darrell
>
>
> class NodeBase:
>     def __init__(self, name, logicFunc):
>         self._name=name
>         self._previousOutputValue=None
>         self._logicFunc=logicFunc
>         self._output={}
>         self._input={}
>
>     def name(self):
>         return self._name
>
>     def connectInput(self, name):
>         self._input[name] = None
>
>     def connectOutput(self, otherNode, name):
>         self._output[otherNode.name()] = name
>         otherNode.connectInput(name)
>
>     def signal(self, collection, name, value=1):
>         self._input[name]=value
>         self.logic(collection)
>
>     def propagate(self, collection, value):
>         if value==self._previousOutputValue:
>             # No point in propagating the same value
>             return
>         for nodeName, inputName in self._output.items():
>             try:
>                 node=getattr(collection, nodeName)
>                 node.signal(collection, inputName, value)
>             except:
>                 print self._name
>                 print nodeName
>                 print collection
>                 raise
>
>     def logic(self, collection):
>         cc= self._logicFunc(self, collection)
>         self.propagate(collection, cc)
>
> def andFunc(self, collection):
>         return reduce(lambda x,y: x and y, self._input.values())
>
> def announceFunc(self, collection):
>         cc= reduce(lambda x,y: x or y, self._input.values())
>         if cc:
>             print self._name, self._input
>         else:
>             # Noise on the input, but nothing to report
>             print '.',
>         return cc
>
> def sourceFunc(self, collection):
>     assert(0)
>
> class SourceNode(NodeBase):
>     def __init__(self, name):
>         NodeBase.__init__(self, name, sourceFunc)
>
>     def __call__(self, collection, value):
>         self.propagate(collection, value)
>
>
> class FaultTree:#(UserDict.UserDict): Not sure about this
>     def __init__(self, *nodes):
>         for n in nodes:
>             setattr(self, n.name(), n)
>
> ft=FaultTree(NodeBase("and1", andFunc), NodeBase("and2", andFunc),
> NodeBase("and3", andFunc),
>             NodeBase("fail", announceFunc))
>
> ft.and1.connectOutput(ft.and3,"one")
> ft.and2.connectOutput(ft.and3,"two")
> ft.and3.connectOutput(ft.fail,"who cares")
>
>     # Setup some signal sources
> drivers=[]
> for x in range(3):
>     s1=SourceNode(x)
>     drivers.append(s1)
>     s1.connectOutput(ft.and1, x)
>
>     # Setup some signal sources
> for x in range(3):
>     s1=SourceNode(x)
>     drivers.append(s1)
>     s1.connectOutput(ft.and2, x)
>
>     # Make it go
> for i in drivers:
>     i(ft, 1)
>
>
>
>




From aleaxit at yahoo.com  Wed Dec 20 07:02:26 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 20 Dec 2000 13:02:26 +0100
Subject: real life examples
References: <u48q19.cth.ln@adsl.bapnt>
Message-ID: <91q75801e82@news2.newsguy.com>

"huber" <nospam at wanted.com> wrote in message
news:u48q19.cth.ln at adsl.bapnt...
> i try to teach python to my son (13 years old)!
> but instead of stupid for loops, classes nobody ever needs and so on .....
> are there some good real life examples a child could be interested in ?

Try 'Alice', http://www.alice.org/; he can start using it in
a purely interactive fashion, then, later...:

"""
Advanced uses of Alice will definitely want to use the underlying scripting
language, rather than the "drag and drop" world-building interface.  To
access this, go to the animations tab, right-mouse-click on the white area,
and select "Add Python Code Box".

Alice's scripting language is Python.  You will definiately want to visit
www.python.org which has tons of very nice documentation -- start with their
tutorial.
"""

(quote from http://www.alice.org/advancedtutorial/).


To be honest, that would have left me pretty cold at 13 --
I never was very interested in pretty pictures, animation,
and so on; the ability to use Python to compute conditional
probabilities in games (card games, but not only) would have
been just the thing to fire up my enthusiasm.  But I suspect
I wasn't a very typical 13-year-old; the 3D graphics and
animation stuff may prove interesting to many 13-y-o's.


Alex





From Bill.Scherer at VerizonWireless.com  Thu Dec 21 10:14:02 2000
From: Bill.Scherer at VerizonWireless.com (Bill Scherer)
Date: Thu, 21 Dec 2000 10:14:02 -0500
Subject: DCOracle-1.3.2 makefile problem .er. opportunity?
References: <91qtcd$dic$1@nnrp1.deja.com>
Message-ID: <3A421E3A.E88925E4@VerizonWireless.com>

I've heard one account of big problems with DCOracle on RedHat 7 
(I'm assuming that's what you mean by "Linux 7").

Anyhow, it looks like you don't have the python-devel rpm installed. 
Try installing that.  Get it from your RedHat cd or from rpmfind.net.

Then you should have the missing /usr/lib/python1.5/config/Makefile.

Alternatively, get the Python source and configure, make, make install
that.

Good Luck!

tim at 2wave.net wrote:
> 
> Hi
> 
> I'm no makefile expert, I'm afraid, and am trying to build DCOracle-
> 1.3.2 in Linux 7 (for Oracle 8.1.6), and not succeeding :-(. I've
> followed the makefile instructions provided, and get to the problem
> below. I haven't found any references to others having this problem on
> comp.lang.python, so I wonder if any of you who have been down this
> path before recognise the error - and point me in directions which
> might overcome it!
> 
> ------------------------------
> root> makefile -f Makefile.pre.in boot
> rm -f *.o *~
> rm -f *.a tags TAGS config.c Makefile.pre python sedscript
> rm -f *.so *.sl so_locations
> VERSION=`python -c "import sys; print sys.version[:3]"`;
> installdir=`python -c "import sys; print sys.prefix"`;
> exec_installdir=`python -c "import sys; print sys.exec_prefix"`;
> make -f ./Makefile.pre.in VPATH=. srcdir=.
>      VERSION=$VERSION
>      installdir=$installdir
>      exec_installdir=$exec_installdir
>      Makefile
> make[1]: Entering directory `/home/tim/rpms/DCOracle-1.3.2/src'
> make[1]: *** No rule to make target'/usr/lib/python1.5/config/Makefile'
> needed by 'sedscript'. Stop.
> make[1]: Leaving directory `/home/tim/rpms/DCOracle-1.3.2/src'
> 
> ------------------------------------------------
> 
> Thanks
> 
> Tim
> 
> Sent via Deja.com
> http://www.deja.com/
> --
> http://www.python.org/mailman/listinfo/python-list

-- 
William K. Scherer
Sr. Member of Applications Staff - Verizon Wireless
Bill.Scherer_at_VerizonWireless.com



From fgeiger at datec.at  Sat Dec  9 04:25:18 2000
From: fgeiger at datec.at (Franz GEIGER)
Date: Sat, 9 Dec 2000 10:25:18 +0100
Subject: UML and Python
References: <90o91k$eks$1@desperado.vew-telnet.net>
Message-ID: <90stth$mea$1@newsreaderg1.core.theplanet.net>

Hi Rainer,

I'm currently lookin' for a decent priced UML IDE. All C++ UML tools are
rather expensive. Do you have already any experiences on ObjectDomain?

Best regards
Franz GEIGER


Rainer Rasche <rrasche at dspace.de> schrieb in im Newsbeitrag:
90o91k$eks$1 at desperado.vew-telnet.net...
> Hi,
>
> is somebody out there, who is interested in developing python applications
> with UML-CASE Tools (e. g. ObjectDomain) and wants to exchange some
> experiences with me?
>
> Thanks, Rainer
>
> rrasche at dspace.de
>
>




From billtj at y.glue.umd.edu  Thu Dec 21 11:48:17 2000
From: billtj at y.glue.umd.edu (William Djaja Tjokroaminata)
Date: 21 Dec 2000 16:48:17 GMT
Subject: A Suggestion for Python Colon Syntax
Message-ID: <91tc8h$2pp$1@hecate.umd.edu>

Hi all,

I have been using Python for some time, and I have a suggestion for the
Python Colon (:) syntax.

I have used C/C++, Tcl/Tk, Perl, etc., and really one thing that I like
about Python is its indentation-based syntax.  Sure, we cannot write
one-liners like what Perl is famous for, but the code layout looks
beautiful, because I don't have to type { and } blocks which result
in various formatting conventions in C/C++.  Also the fact that we don't
have to write ( and ) after if, for, etc., is wonderful, because the code
has less "punctuation" and looks more clean.

However, a lot of time when I ran my Python program, I got syntax error,
because I forgot to put colon (:) after my if (or for, etc.) clause.  My
question is, is the colon really necessary, syntactically, in
Python?  Because Python is line-and-indentation-based, can we just
eliminate the colons after the if, for, etc., clause?

My suggestion is as follows:

1) If the if (or for, etc.) clause is followed by a block, then the colon
is optional (for backward compatibility):

    if condition is true
        statement_1
        statement_2
        ....

2) If the if (or for, etc.) clause is followed only by a single statement,
then the colon can be used to write a single line of code:

    if condition is true: statement_1
    next_statement
    ....

Any comments, anyone?

Regards,

Bill


From dgrisby at uk.research.att.com  Sat Dec 16 09:12:13 2000
From: dgrisby at uk.research.att.com (Duncan Grisby)
Date: 16 Dec 2000 14:12:13 GMT
Subject: no current thread warning
References: <t3ks6hg7c7rgae@corp.supernews.com> <91dtv0$g93$1@panix6.panix.com>
Message-ID: <91ft7t$5rn$1@pea.uk.research.att.com>

In article <91dtv0$g93$1 at panix6.panix.com>, Aahz Maruch <aahz at panix.com> wrote:

[...]
># Dummy thread class to represent threads not started here.
># These aren't garbage collected when they die,
># nor can they be waited for.

This behaviour is a problem for any C / C++ code which creates its own
threads, then calls into Python. It would be nice if threading.py
contained a proper API for creating and deleting threading.Thread
objects for threads created outside of Python.

In omniORBpy, I have exactly this problem. The work-around is to
create my own threading.Thread objects, and manually manipulate the
threading._active dictionary. That's ugly because it depends on the
threading module's internal structure. The Python code to do it is
below. The C++ side caches the Thread objects so they don't have to be
created on each up-call from C++ to Python.

# WorkerThread class used to make the threading module happy during
# operation dispatch.
# *** Depends on threading module internals ***

class WorkerThread(threading.Thread):

    def __init__(self):
        threading.Thread.__init__(self, name="omniORB")
        self._Thread__started = 1
        self.id = threading._get_ident()
        threading._active_limbo_lock.acquire()
        if threading._active.has_key(self.id):
            self.add = 0
        else:
            self.add = 1
            threading._active[self.id] = self
        threading._active_limbo_lock.release()

    def delete(self):
        if self.add:
            threading._active_limbo_lock.acquire()
            del threading._active[self.id]
            threading._active_limbo_lock.release()

    def _set_daemon(self): return 1


Cheers,

Duncan.

-- 
 -- Duncan Grisby  \  Research Engineer  --
  -- AT&T Laboratories Cambridge          --
   -- http://www.uk.research.att.com/~dpg1 --


From cg at schlund.de  Tue Dec 19 11:23:37 2000
From: cg at schlund.de (Carsten Gaebler)
Date: Tue, 19 Dec 2000 17:23:37 +0100
Subject: Walk through dictionary keys?
Message-ID: <3A3F8B89.1DD9E01F@schlund.de>

Hi!

Is there a way of 'walking' through the keys of a dictionary? If I have
a large dictionary I'd like to prevent 

for key in d.keys():
    do_something

because this makes a copy of all the keys. I'm looking for something
like Perl's "each".

Regards
Carsten.


From gregholmes at my-deja.com  Thu Dec  7 10:02:29 2000
From: gregholmes at my-deja.com (gregholmes at my-deja.com)
Date: Thu, 07 Dec 2000 15:02:29 GMT
Subject: os.system gives DOS mode warning in win98
Message-ID: <90o8q1$f2r$1@nnrp1.deja.com>

Using Neil Shamenaur's (sp?) standalone 1.5.2 distribution for win32, I
load the python15.dll in a windows executable.  The python script
then executed does a "os.system('start iexplore.exe')".

When this script is run from a DOS box, it works fine, (IE starts).

>From the windows executable, it gives a DOS mode warning when the
os.system happens.

Is this expected?  Is there any way to prevent it?


Sent via Deja.com http://www.deja.com/
Before you buy.


From news at dorb.com  Tue Dec 12 19:57:40 2000
From: news at dorb.com (Darrell)
Date: Wed, 13 Dec 2000 00:57:40 GMT
Subject: Problems With socket Under NT & PythonWin 2.0
References: <910jtt$nnh$1@nnrp1.deja.com> <fMQY5.12943$hp.1457332@typhoon.nyroc.rr.com> <915k4g$kcc$1@nnrp1.deja.com>
Message-ID: <8QzZ5.1711$7g.340689@typhoon.nyroc.rr.com>

I was guessing that your app was sharing an important thread with PythonWin.

PythonWin could have started your app in another thread to avoid this.
But threads bring pretty tricky problems along with them.
Although they are much simpler in Python.

--Darrell

<swanky_mode at my-deja.com> wrote:
> Just out of curiosity, is this thread blocking causing the IDE to hang
> a PythonWin issue only, or does this occur on other OSs and IDEs?
>





From just at letterror.com  Sun Dec 31 14:54:06 2000
From: just at letterror.com (Just van Rossum)
Date: Sun, 31 Dec 2000 20:54:06 +0100
Subject: Pythonic concurrency (was: Erlang (was Re: Good introduction to 
 functional programming with Python?))
References: <mailman.977984412.30211.python-list@python.org> <92f0il01a5u@news1.newsguy.com> <86wvckqw4h.fsf@localhost.my.domain> <G6AyoC.CK1@world.std.com> <92A081D417204A91.AF29627FD1BF4835.0C336CAB4734825B@lp.airnews.net>
Message-ID: <3A4F8E91.3F29F97C@letterror.com>

Cameron Laird wrote:
> 
> In article <G6AyoC.CK1 at world.std.com>, Will Ware <wware at world.std.com> wrote:
> >Brad Knotwell (knotwell at ix.netcom.com) wrote:
> >> I'm now wondering if Stackless Python's uthreads are cooperatively or
> >> pre-emptively multi-tasked.
> >
> >Unless you're digging into their inner workings, microthreads will appear
> >to you as preemptively multitasking.
>                         .
>                         .
>                         .
> Is this true:  microthreads block on system calls (I/O
> operations, most notoriously) and on invocations of
> externally-coded extensions?

They do.

Just


From plowboy_lifestyle at my-deja.com  Wed Dec 20 18:15:58 2000
From: plowboy_lifestyle at my-deja.com (plowboy_lifestyle at my-deja.com)
Date: Wed, 20 Dec 2000 23:15:58 GMT
Subject: Extracting attributes from modules
Message-ID: <91rej7$td7$1@nnrp1.deja.com>

Hi,
  I use python as an embedded scripting tool in the 3D
modeling application called Blender and I'm looking for some
clarification from a pro.
I have this difficulty when I try learn about the various features of
the embedded API from my scripts. A typical blender script looks
something like this:

import Blender
from Blender import *

print dir()

this program is supposed to give me a list of the objects in the name
space including everything from the module blender. The output looks
something like this (hopefully I can explain all this without being
application specific):

['Blender','Camera','Object','World',...etc.]

so I explore further with

print dir(Object)

and I get:

['Get','GetSelected','Update','__doc__','__name__']

if I use the get function I find out that it returns an object from
within the modelling program, like a sphere or a mesh, whatever I'm
using and I can manipulate it by setting various variable like "LocX" or
"LocY" for the various locations of the 3D objects, like so:

mySphere = Object.Get('theSphere')
mySphere.LocX = mySphere.LocX + 5

this program will have the effect of moving the object in space positive
5 units on the x axis.

so this is my actual question; Even though I can find out about various
attributes like LocX and etc. from the manuals for this application,
there is no way for me to find them from python it seems, for example:

print dir (Object.Get)
print dir (mySphere)
print dir (mySphere.LocX)
print Object.Get.__doc__
print mySphere.__doc__
print dir(type(mySphere))

all output an empty list or "none", even though

print myShere.LocX

works fine at printing out a floating point number.
I'm only trying to get a list of names of attributes of the object but
they seem hidden. Obviously I'm new to python. Can someone explain why
this works this way?

Thanks,
IB


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


From jepler at inetnebr.com  Tue Dec 12 09:02:06 2000
From: jepler at inetnebr.com (Jeff Epler)
Date: Tue, 12 Dec 2000 14:02:06 GMT
Subject: A question.
References: <9152a0$acb$1@panorama.wcss.wroc.pl>
Message-ID: <slrn93cbur.9tj.jepler@potty.housenet>

On Tue, 12 Dec 2000 12:32:09 +0100, DeHa
 <deha at poczta.onet.pl> wrote:
>Hi All!

Hello!

Are you trying to implement the algorithm known as the 'sieve of Eratosthenes'?

Here is my implementation:

def sieve(max = 1000):
	# elements of r correspond to numbers, 0 means the number has been
	# proved composite, 1 means the number is a suspected prime
	r = [1] * (max+1)

	# The list of numbers to return
	p = []

	# Look at each number in turn
	for i in range(2, max+1):

		# If it is already known to be composite, do nothing
		if not r[i]: continue

		# If it is suspected prime, and all divisors have been
		# tested, then it is known to be prime
		p.append(i)

		# Mark all multiples of the number as composite
		# eg range(2*3, 1001, 3) -> (6, 9, 12, ..., 999)
		for j in range(2*i, max+1, i):
			r[j] = 0

	# We have tested all candidates, so p contains the list of
	# all primes less than or equal to max
	return p

assert sieve(30) = [2, 3, 5, 7, 11, 13, 17, 19, 23, 29]


From swanky_mode at my-deja.com  Tue Dec 12 11:35:28 2000
From: swanky_mode at my-deja.com (swanky_mode at my-deja.com)
Date: Tue, 12 Dec 2000 16:35:28 GMT
Subject: Problems With socket Under NT & PythonWin 2.0
References: <910jtt$nnh$1@nnrp1.deja.com> <fMQY5.12943$hp.1457332@typhoon.nyroc.rr.com>
Message-ID: <915k4g$kcc$1@nnrp1.deja.com>

In article <fMQY5.12943$hp.1457332 at typhoon.nyroc.rr.com>,
  "Darrell" <news at dorb.com> wrote:
> Copy the example to a .py file and run it from the command line or
just
> double click it.
> The example blocks on conn.recv(1024), meaning that you current
thread will
> stop responding.
>
> Of interest might be select or timeout socket.
> Timeout socket is something you would need to download.
>

Hey, Darrell;

Thanks for your help.  Both the EchoClient.py and EchoServer.py worked
as expected from the command line as opposed to using the PythonWin IDE.

I'll be taking a look at Timeout Socket shortly (now downloading it
from the Valuts of Parnassus...)

Just out of curiosity, is this thread blocking causing the IDE to hang
a PythonWin issue only, or does this occur on other OSs and IDEs?

Again, thank you for your time and help...

Bill


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


From clarence at netlojix.com  Fri Dec 29 12:03:49 2000
From: clarence at netlojix.com (Clarence Gardner)
Date: Fri, 29 Dec 2000 09:03:49 -0800
Subject: Question re: eval()
Message-ID: <978109654.339714327@news.silcom.com>

I want to eval() an expression, but have some of the variables in it be
lazily evaluated.  E.g.,

class D:
    def __getitem__(self, name):
        if name == 'n':
            return 3

d = D()
x = eval('n + 1', d)


but the eval function accepts only a dictionary as the second parameter.
Is there any amazing way to do this?

TIA,
Clarence Gardner
clarence at netlojix.com


From aleaxit at yahoo.com  Wed Dec 27 04:03:52 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 27 Dec 2000 10:03:52 +0100
Subject: Python and Napster
References: <92ap0g12b6r@news1.newsguy.com>, <sx226.18970$RC1.511898@newsread1.prod.itd.earthlink.net> <20001227164900.DF7E6A84F@darjeeling.zadka.site.co.il>
Message-ID: <000c01c06fe3$efc29520$102b2bc1@cadlab.it>

"Moshe Zadka" <moshez at zadka.site.co.il> writes:


> On Tue, 26 Dec 2000 19:40:14 +0100, "Alex Martelli" <aleaxit at yahoo.com>
wrote:
>
> > Python would probably let you be more productive, I suspect.  It sure
has
> > all the network functionality you need; in your shoes, I'd look into
pyro --
> > building file exchange on top of it should be easier than on sockets or
on
> > protocols such as HTTP.
>
> I'm not so sure I agree -- SimpleHTTPServer/urllib/httplib is wonderful
> for file transfers. For anyone who wants to see a practical exmple
> (supports both upload and download), have a look at PMEP in the PMS
> source distribution (http://pythonms.sourceforge.net)

HTTP (and the Python tools around it) is great for file transfer (and
suchlike) but it's not oriented to peer-to-peer at all.  Pyro knows
nothing about files, but offers cool, simple, lightweight distributed
objects in 100% pure Python; a distributed Naming System, which can
support broadcast on a LAN, or other discovery methods if more suitable,
and is slated to become persistent at next release; pretty good
'transparency' of the remoting -- including exceptions -- and, again
in the next release, mobile objects and lifetime services (remote
activation/deactivation/persistency).

Corba is of course an alternative, but it may be harder to set up and
install properly on a great variety of platforms -- Pyro should run
wherever Python does (judging from its docs & blurbs, that is).

So the issue may boil down to pragmatics.  Is it easier to build file
transfer on top of a nice, lightweight but rich, distributed object
system; or, is it easier to build what portions you need of such an
object system on top of HTTP (given that, once machine A has found
out that machine B has file X.Y it needs, the actual transfer of that
file is a piece of cake, basically already performed by the protocol
and available supporting tools).

I think that actually transferring a file's data and metadata, once
you have a proxy for a remote object, is no big issue.  It seems to
me that naming, exceptions, resuming transfer (not supported by HTTP,
is it?), finding out what files are available where and with what
metadata (to discover what version is most up-to-date, or, possibly,
that both versions are separate branches of a version tree and some
manual reconciliation will be needed), etc, are harder problems --
and maybe using an infrastructure that basically lets you think in
term of programming Python objects, rather than inventing and
implementing special-purpose sub-protocols, may make life easier.

OTOH, if one's needs include, e.g., going through firewalls, then
HTTP may well be the best solution (in which case, SOAP on top of
it might help avoid reinventing too many wheels).  But, as I had
read the original request, it seemed to me to be more likely that
the whole setup would eventually run on a LAN or intranet, so that
such things as firewalls might not be a problem...


Alex





From MarkH at ActiveState.com  Tue Dec 19 20:09:18 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Wed, 20 Dec 2000 01:09:18 GMT
Subject: Dispatch("Excel.Application") = Access denies
References: <mailman.977220783.7381.python-list@python.org>
Message-ID: <3A400588.9080203@ActiveState.com>

andrew.j.atkinson at unilever.com wrote:

> Dear all
> 
> Have developed a python app to automate the running of an excel 
> spreadsheet. Works a treat when run on the command line, however get 
> the following 'trace' when run through Tivoli's Maestro job scheduler 
> (same user in both cases).
> 
> Any ideas (it's quite urgent).

the job scheduler is probably running your program as the wrong user  - 
quite possibly the "System" user.

Mark.



From jepler at inetnebr.com  Tue Dec 26 19:49:51 2000
From: jepler at inetnebr.com (Jeff Epler)
Date: Wed, 27 Dec 2000 00:49:51 GMT
Subject: ~Python ?
References: <3a4833f8.346848@news.t-online.de>
Message-ID: <slrn94if5e.3bu.jepler@potty.housenet>

On Tue, 26 Dec 2000 06:03:40 GMT, Gerson Kurz
 <gerson.kurz at t-online.de> wrote:
>I understand that ~ is probably handled by the shell, but if Python
>doesn't deal with it, its hard to use Python in some kinds of scripts.
>(I think Perl, Pythons ugly twin-brother, does)

I'd have to say that string-substitution-based languages screw
me up about ~2/3 of the time.  It must be something about the way my mind
works.      ^^^^
            hee hee, get it?


Seriously, the os.path module includes goodies such as os.path.expanduser
and os.path.expandvars.

If you want shorthand for this when you do an 'open', try the following:
	import __builtin__, os
	def open(file, *args):
		file = os.path.expanduser(file)
		file = os.path.expandvars(file)
		return __builtin__.open(file, *args)

>>> open("~/.bashrc").readlines()
['# .bashrc\012', '\012', '# User specific aliases and functions\012', '\012', '# Source global definitions\012', 'if [ -f /etc/bashrc ]; then\012', '\011. /etc/bashrc\012', 'fi\012']

Jeff


From aleaxit at yahoo.com  Mon Dec  4 03:37:09 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 4 Dec 2000 09:37:09 +0100
Subject: Python programs (or scripts) as drop targets
References: <hv5k2tg9cdf33j8v333qpkk7jo45ht596n@4ax.com> <3A2ACE17.6030304@ActiveState.com>
Message-ID: <90fl3d022k9@news2.newsguy.com>

"Mark Hammond" <MarkH at ActiveState.com> wrote in message
news:3A2ACE17.6030304 at ActiveState.com...
> Dale Strickland-Clark wrote:
>
> > The big question now is: is such an interface exposed anywhere by
Python?
>
> A quick look shows we need IDropTarget and IDataObject.

Why IDataObject?  I think IPersistFile is what's needed here.


Alex





From aleaxit at yahoo.com  Wed Dec 20 07:52:53 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 20 Dec 2000 13:52:53 +0100
Subject: Python in "math through programming" curriculum
References: <2g2t3t06r2agrrjmfnr41eu4lc2u39a8qi@4ax.com> <3A3EFF47.67DE69CF@gte.net> <91n9i30pr7@news2.newsguy.com> <slrn9402bt.fa6.neelk@alum.mit.edu>
Message-ID: <91qa3r01m78@news2.newsguy.com>

"Neelakantan Krishnaswami" <neelk at alum.mit.edu> wrote in message
news:slrn9402bt.fa6.neelk at alum.mit.edu...
    [snip]
> I disagree with you here about Python's appropriateness. Conceptually,
> the strongly typed FPLs are the best languages there are for learning
> about the mathematics of computer programming. The combination of

I wouldn't consider anything but Haskell if I were to teach "the
mathematics of computer programming" (that may be because my
exposure to Clean is too limited) stricto sensu (there's a lot
of math, connected to computer programming -- see all of Knuth's
'Art of Computer Programming' -- that is not the math OF computer
programming... Python, and a RISC machine language, would be good
for _that_!-).

But "math through programming" is a completely different issue,
at least as the original poster meant it (as he explained in a
followup to this post of yours) and as I took it -- middle and
high school levels, as he had said!  I would *NOT* introduce
Haskell in middle school, and even for high school, I have my
doubts -- students would have to have strong *previous* exposure
to mathematical concepts, AND to computer programming; few high
schools, I suspect, are good enough for that.


> > Caml, IMHO, has rather cluttered syntax when compared to Haskell (or
> > Python!); it only gets worse when the Object-oriented extensions are
> > also considered.  Admittedly, a minor consideration, but, _for the
> > specific issue of programming-newbies_, not a negligible one IMHO.
>
> This is the *big* weakness of the ML-family languages (including
> Haskell and Clean): they all have gnarly syntaxes. Python is proof

Haskell's syntax is *lovely*!  How can you class it in "the ML
family"?  [Again, I don't know enough Clean to have a real opinion
about it].

> that a simple syntax is extremely important; it has essentially the
> same semantics as Perl, yet I would gnaw off my arm to escape
> programming in Perl and quite enjoy programming in Python. Random

Oh, I fully agree on this one.  But, *Haskell*?!

Consider list comprehensions, for example.

In Haskell:

    [(x,y) | x <- [0,1,2], y <- [0,1], x > y]

in Python:

    [(x,y) for x in [0,1,2] for y in [0,1] if x > y]

Sure, Python is using more keywords ('for', 'in', 'if') where
Haskell uses punctuation ('|', '<-', ','), but surely that is
a _very_ surface issue -- and the end result could be argued
either way (I think somebody familiar with common mathematic
notation for sets might well find the Haskell case clearer!).

Python and Haskell also share such syntax elements as the
significance of whitespace ('offsides rule' in Haskell),
which lets one express many things unclutteredly.

What specific objections do you have to Haskell syntax...?


Alex





From aleaxit at yahoo.com  Mon Dec 25 18:54:09 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 26 Dec 2000 00:54:09 +0100
Subject: One Python 2.1 idea
References: <mailman.977771949.17911.python-list@python.org>
Message-ID: <928n2s02jko@news1.newsguy.com>

"Tim Peters" <tim.one at home.com> wrote in message
news:mailman.977771949.17911.python-list at python.org...
    [snip]
> > Can the code for list comprehensions be made faster?
>
> Sure.  "bang for the buck" is limited by how often a particular trigger is
> pulled, though, and listcomps are currently barely used.  This isn't a
> technical problem, it's a problem of finding someone motivated enough to
> bother doing something about it.  You're not -- and neither am I.

I might be, given that I think list comprehensions *deserve* wider
usage and I'd be loath to think of people who *might* use them...
and won't because of a silly 5%-performance-issue (I guess I might
'think of it as evolution in action', but, I'm a softy!-).

If I'm right that the issue is allocation, then I'd need a different
opcode at the start (a build-list variation) and one at the end to
say 'ok, listbuilding done now, please ensure it's a normal list
without too much extra fat'; the former would change the append
method pointer of the list object to {either reallocate exponentially,
or, probably better, build the list in disjoint pieces -- as nothing
BUT 'append' will be used on the list until ok-done-now time},
the latter would (if needed) recompact it up and restore the normal
append method for possible future uses.

Am I wildly off-base...?  If not, it seems to me that procuring
an acceptable diff-program is going to be the largest part of
making a patch to submit...:-).


Alex





From kkinder at tridog.com  Sun Dec  3 00:36:54 2000
From: kkinder at tridog.com (Ken Kinder)
Date: Sat, 02 Dec 2000 22:36:54 -0700
Subject: SSL with python
References: <9097pc+pspc@eGroups.com> <v04220804b64f0e47b847@[192.168.11.33]>
Message-ID: <3A29DBF6.6A074D47@tridog.com>

>From a simple search of http://www.vex.net/parnassus/ -- the premier python module
directory, I found a few. I suggest just going through the results yourself.
http://www.vex.net/parnassus/apyllo.py?find=ssl

ssl_wrapper 1.1 looks very promising but it's a broken link. Does anyone know
where a mirror of that one is?

Blair Lowe wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
>
> Hi,
>
> I am new with python, and I was wondering if anyone has got it to
> work with SSL connections.
>
> We are trying to do some socket level manipulation using urllib, and
> httplib, and we are getting some messages saying that we are using
> straight http over an SSL connection.
>
> Does anyone know of any information about SSL and python (web or printed work)?
>
> Thanks,
> Blair.
> Computer Engineering Inc. http://www.compeng.net
> Phone: 780 499 5687 (9 - 5 MST) Fax:   780 435 0693 (24 Hours)
>
> -----BEGIN PGP SIGNATURE-----
> Version: PGPfreeware 6.5.8 for non-commercial use <http://www.pgp.com>
>
> iQCVAwUBOilgDUZu3Sp72OwBAQFHRQP8CKBpVM8Wll7N2DSjbhtgueqsWohYkjmc
> 4nymzTfsTPGebMjjE+AnozMe+yzUvpSaXedABw3vGfJME/r807H/uIOrZbMLYhsw
> gp0fWL06DJpcNRuAYLHZ6QTD5VjqWxsgm8BZSNA3vzR2I40JYh5BOaVOVjy9S4mA
> alP+cyWsFA0=
> =38Vk
> -----END PGP SIGNATURE-----
>
> --
> http://www.python.org/mailman/listinfo/python-list




From brian_zhouNOSPAM at agilent.com  Fri Dec 15 13:59:01 2000
From: brian_zhouNOSPAM at agilent.com (Brian Zhou)
Date: Fri, 15 Dec 2000 10:59:01 -0800
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <3a36c85e.4409136@News.CIS.DFN.DE> <owen_nospam-1312001131590001@cs204-90.fsmodem.washington.edu> <3A37DD31.C9A5F18C@pacbell.net> <91bhlh$c64u$1@nntp6.u.washington.edu> <91d7270niu@news1.newsguy.com> <91dmda$89ei$1@nntp6.u.washington.edu>
Message-ID: <976906728.172873@emperor.labs.agilent.com>

"Russell E. Owen" <owen at astroNOJNK.washington.edu.invalid> wrote in message
news:91dmda$89ei$1 at nntp6.u.washington.edu...
> I need something that works, or will work, on Mac and unix for sure,
> Windows would also be good. The only Python GUI library that does that
> now is Tkinter, but wxPython is likely to do so the future. Are there
> any others?
>
Don't know how it compares with others but yet another alternative is
"jpython & swing"?
It might be a bit slow though.




From peter at engcorp.com  Sun Dec 31 10:05:49 2000
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 31 Dec 2000 10:05:49 -0500
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <92m9pm$d19$1@nnrp1.deja.com> <B3z36.122$9s1.18804@e420r-atl2.usenetserver.com> <3A4EC5B1.8912E571@engcorp.com> <x8G36.156$9s1.31209@e420r-atl2.usenetserver.com>
Message-ID: <3A4F4B4D.ECD6B172@engcorp.com>

Kragen Sitaker wrote:
> 
> Peter Hansen  <peter at engcorp.com> wrote:
> >Kragen Sitaker wrote:
> >>
> >> The only time I wish I wasn't subject to Python's indentation rules is
> >> when I can't outdent my "if finished(): break" statements like I do in
> >> C.
> >
> >But why would you want to outdent it?  It's *part* of the loop, not a
> >following statement.  Identation should match the block structure of the
> >code, and that statement lies _within_ the block.
> 
> I would want to outdent it because, in my view, it isn't part of the
> loop body; it's part of the loop structure itself, just like the while
> 1: at the beginning.  It is not within the block of code controlled by
> the loop.  In fact, it comes *between* two blocks of code controlled by
> the loop, much as else: comes between two blocks of code controlled by
> an if (or, in Python, a for or while).

I can see your point (remarkable, given this thread, isn't it? :-).  I
guess I consider the "block" to be everything except the code that allows
entry into the block.  Within the block I might have multiple exit
conditions, each shown as separate "if/break" blocks, which are themselves
indented appropriately (and with Python, necessarily!).

But I can't argue with another view on this.  After all, in C I've been
using "brace indentation style 3" for the longest time, in apparent
violation of 98% of coding conventions in the world.  That is, my braces
are vertically aligned but indented with respect to the "if ()" or "for
()" which gets me into the block:

while (1)
   {
   T = f();
   if (T < 25)
       break;
   }

Use of that convention is probably what has made me quite happy with
Python's indentation style, which essentially matches this pattern.


From glenfant at equod.com.nospam  Tue Dec 19 14:23:51 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Tue, 19 Dec 2000 20:23:51 +0100
Subject: MIDI extensions (WIN32, Mac)?
References: <3A3FACEF.DAF77229@cableinet.co.uk>
Message-ID: <91ocbe$jg4$1@reader1.imaginet.fr>

I don't think that Python is the best language for "real-time" stuffs like a
MIDI sequencer.
But perhaps a high level object (say a full featured sequencer) written in C
may be controled from Python script, but it's hard work.

"Richard Dobson" <rwd at cableinet.co.uk> a ?crit dans le message news:
3A3FACEF.DAF77229 at cableinet.co.uk...
> Hello,
>
>
> I have just started looking at Python seriously, with a focus on music
> applications; e.g. algorithmic composition. Python seems to me to be
> idally suited as a first programming language for Music Technology
> students who have never done programming before. So far the net trawling
> I have done has failed to turn up any already-done extensions supporting
> MIDI i/o. I could probably code this myself at least for WIN32 if need
> be, but it is the sort of thing I feel sure somebody must already have
> done. Can anyone suggest anything? Or has it indeed not been done yet?
>
>
> Richard Dobson
>
> --
> Test your DAW with my Soundcard Attrition Page!
> http://www.bath.ac.uk/~masrwd (LU: 3rd July 2000)
> CDP: http://www.bath.ac.uk/~masjpf/CDP/CDP.htm (LU: 23rd February 2000)



From joconnor at cybermesa.com  Thu Dec 21 16:30:48 2000
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Thu, 21 Dec 2000 21:30:48 +0000
Subject: Ooops....
Message-ID: <3A427688.D4B5E802@cybermesa.com>

All,

	I just accidentally sent a message to the python list about a musical
arrangement with a drummer friend of mine.  

	My apologies...I misaddressed the message

	We now return you to our regularly scheduled programming on debating
the merits of "while 1:", now in it's second season of reruns


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

"God himself plays the bass strings first when He tunes the soul"



From support at internetdiscovery.com  Wed Dec 20 11:08:03 2000
From: support at internetdiscovery.com (Mike Clarkson)
Date: Wed, 20 Dec 2000 16:08:03 GMT
Subject: Fundamental Tkinter inefficiency?
Message-ID: <3a40d46e.26612406@24.0.228.33>

I'm looking at the implementation of Tk_MainLoop in _tkinter.c:

	quitMainLoop = 0;
	while (Tk_GetNumMainWindows() > threshold &&
	       !quitMainLoop &&
	       !errorInCmd)
	{
		int result;

#ifdef WITH_THREAD
		Py_BEGIN_ALLOW_THREADS
		PyThread_acquire_lock(tcl_lock, 1);
		tcl_tstate = tstate;
		result = Tcl_DoOneEvent(TCL_DONT_WAIT);
		tcl_tstate = NULL;
		PyThread_release_lock(tcl_lock);
		if (result == 0)
			Sleep(20);
		Py_END_ALLOW_THREADS
#else
		result = Tcl_DoOneEvent(0);
#endif

		if (PyErr_CheckSignals() != 0)
			return NULL;
		if (result < 0)
			break;
	}


and was wondering if it would be better to use a Tcl Event procedure
with a MaxBlock time to check the Python signals, instead of calling
Sleep.  Does anyone know how to do this in Tcl - it shouldn't be
more than a few lines I suspect.

Many thanks in advance,

Mike.




From magnus.heino at rivermen.se  Fri Dec 15 03:33:46 2000
From: magnus.heino at rivermen.se (Magnus Heino)
Date: Fri, 15 Dec 2000 08:33:46 +0000
Subject: Scopes and exceptions
Message-ID: <91chgc$7m8$1@merlin.rembrandtstr.bocholt.de>

Could someone please explain this to me:


[magnus at daysleeper magnus]$ cat e.py
 
a = 'test'
b = 'test'
 
print 'a      = %s' % id(a)
print 'b      = %s' % id(b)
 
def func():
        print "'test' = %s" % id('test')
 
func()
 
try:
        test, '%s' % (a,)
except NameError, e:
        print 'e      = %s' % id(e)
        print 'a      = %s' % id(a)
        print 'b      = %s' % id(b)
        print "'test' = %s" % id('test')


[magnus at daysleeper magnus]$ python e.py
a      = 136227368
b      = 136227368
'test' = 136227368
e      = 135997204
a      = 136227368
b      = 136227368
'test' = 136227368
[magnus at daysleeper magnus]$    

/Magnus


From grante at visi.com  Thu Dec 28 14:05:25 2000
From: grante at visi.com (Grant Edwards)
Date: Thu, 28 Dec 2000 19:05:25 GMT
Subject: Basic MAPI
References: <nEG06.1008$Qb7.95625@newsb.telia.net> <91vfsb02cn6@news2.newsguy.com>
Message-ID: <V9M26.4016$4c.285215@ruti.visi.com>

In article <91vfsb02cn6 at news2.newsguy.com>, Alex Martelli wrote:
>"Henrik M?rtensson" <henrik.martensson at frontconsulting.se> wrote:
>
>> I'm trying to run a basic MAPI initialisation, but I get stuck
>> with the dispatcher.
>    ...
>> >>> s=Dispatch("Mapi.Session")
>
>There is no creatable coclass (that I ever heard of) with progId
>'Mapi.Session'.

That's what I use:

[...]

class MapiFetcher:
    
    def __init__(self, settings = "MS Exchange Settings"):
        self.sessionSettings = settings
        self.session = None
	self.debugLevel=0

    def Logon(self):
        self.session = win32com.client.Dispatch("MAPI.Session")
        self.session.Logon(self.sessionSettings)
	if self.debugLevel > 1:
	    sys.stderr.write("MapiFetcher: MAPI.Session Logon() with settings = '%s'\n" % (self.sessionSettings,))
    
    def CheckInbox(self):
        inbox = self.session.Inbox
        collmsg = inbox.Messages
	if self.debugLevel > 1:
	    sys.stderr.write("MapiFetcher: Found %d messages in Inbox\n" % (collmsg.Count,))
        messages = []
        
        msg = collmsg.GetFirst()

        while msg:
            [...]
            msg = collmsg.GetNext()
    
[...]
    
>> Below I have pasted the code I'm running. Is there something
>> that you have to do to make the MAPI interface available to
>> you? I'm not running this in a MS Exchange office environment.

I don't know what that means.  The machine on which my code
runs has MS-Outlook installed, and there's an Exchange Server
somewhere...

>You have to instantiate a coclass that will let you get at a
>relevant MAPI object model.  The progId you will use will
>depend on what MAPI-supplying product you want to use --
>Exchange Server, Outlook, Eudora, Netscape, whatever.  With
>Outlook, for example, the Session property of the Outlook
>application object (which is a creatable coclass, with a
>progId of 'Outlook.Application') might serve your needs.

-- 
Grant Edwards                   grante             Yow!  I like your SNOOPY
                                  at               POSTER!!
                               visi.com            


From lg at rgz.ru  Thu Dec 28 07:35:49 2000
From: lg at rgz.ru (lg at rgz.ru)
Date: 28 Dec 2000 15:25:49 +0250
Subject: Is this a BUG?
References: <3A4B0C5E.1E7976D2@wielun.dhs.org>
Message-ID: <3a4b314d@news.edunet.ru>

QdlatY <qdlaty at wielun.dhs.org> wrote:
> Hello!

> I have script:
> ----------------
> import cgi

> form = cgi.FieldStorage()

> print form
> print "!",form["index"].value,"!"
try print "!"+form["index"].value+"!"

> I launched it with parameter index=1
> And this is an output:
> ----------------
> FieldStorage(None, None, [MiniFieldStorage('index', '1')])
> ! 1 !
> ----------------
>  ^ ^
> Why the hell there are spaces? index property has value '1'!
> not ' 1 '?!

> Yours
> QdlatY

-- 
zev


From cabrera at wrc.xerox.com  Tue Dec 19 12:23:29 2000
From: cabrera at wrc.xerox.com (jicman)
Date: 19 Dec 2000 17:23:29 GMT
Subject: How to get the os.system data into a variable
Message-ID: <91o5ih$2kv$1@news.wrc.xerox.com>

Hi all!

I know that

t = os.system('ls -l')

return code and assigns it to t.  I would like to get the data, not the code, 
that is resulted from the command into a variable.  (i.e.

-rw-r--r--   1 root     other        821 Dec  2 12:20 MyLog.txt
drwxr-xr-x   2 root     root         512 Dec 21  1999 TT_DB
drwxrwxrwx   3 root     root         512 Aug 18 17:05 apps
lrwxrwxrwx   1 root     root           9 Dec 20  1999 bin -> ./usr/bin
drwxr-xr-x   3 root     nobody       512 Nov 28 12:31 cdrom
drwxr-xr-x   2 root     other        512 Nov 21 20:10 cgi-bin
-rw-------   1 root     root      975040 Oct  3 11:34 core

etc.)

I know that I could do

t = os.system('ls -l > mytemp.txt')

and then read the file, but I would like to acquire that data w/o reading the 
file.  Oh, and by the way, the command that I am sending is not really ls.  It 
does not even belong to the unix world.  It's the output of a command from an 
application that I have.

Any ideas?  It would be greatly appreciated.

thanks,

jos?



From prestonlanders at my-deja.com  Tue Dec 26 17:50:47 2000
From: prestonlanders at my-deja.com (Preston Landers)
Date: Tue, 26 Dec 2000 22:50:47 GMT
Subject: mail processing module?
References: <91rcig$rgu$1@nnrp1.deja.com> <m266kd45m0.fsf@freak.kaiserty.com> <91vumo$di3$1@nnrp1.deja.com> <92ap6g$o7n$1@nnrp1.deja.com> <92b337$h50ik$1@fido.engr.sgi.com>
Message-ID: <92b7c5$3id$1@nnrp1.deja.com>


In article <92b337$h50ik$1 at fido.engr.sgi.com>,
  offer at sgi.com (Nhi Vanye) wrote:

> You have a fetch_first_msg() API, is that all ?

No, no.  Its just what I happened to put in the example that is part
of the docs.  You will be able to at least fetch messages by 'raw
index number' which is the minimum supported on the various APIs.
Possibly other methods ... suggestions?  One will also have access to
the 'server object' (eg, poplib.POP3 object) if one needs to go around
the interface.

> What about non POP/IMAP mailboxes (does the API work for those ? ).

Well.. no, not currently, but nothing prevents other APIs from being
added to the module.  Before I release it I may rewrite it partially
to make it easier to subclass to use a new API w/o modifying the
module.

What mailbox APIs would you like to see supported besides POP3 and
IMAP4? These are the only libs which exist in the standard Python
distribution.  I'm not keen on introducing dependencies on modules not
included with standard python.

> How does the API scale (particularly for unix mailboxes where you
don't
> really want to have to keep the entire box in memory if you can avoid
it).

I'm not sure what you mean by 'for unix mailboxes.'  plain spool
files? This module doesn't handle it.  Its basically for POP or IMAP
(see above.)  spool file support could be added I guess.  The standard
POP and IMAP modules, which I am using, do not keep the entire box in
memory.  In other words, the scalability is exactly the same as for
poplib/imaplib/smtplib. ;-)

> Does this API make sense as a wrapper around the c-client code ?
(C-client
> has a syntax for specifying which 'host/port/mailbox' to use), if you
> were to pick up c-clients syntax for naming mailboxes that would help
> any future work.

I'm sorry, I don't have any idea what you're talking about here.  What
is "C-Client"? By 'naming mailboxes' are you referring to selecting
folders?  It is 'unofficially supported' for imap users and POP
doesn't have a concept of folders.  If you could direct me to a
specification of the syntax you are talking about that would help.

thanks
Preston



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


From mal at bewoner.dma.be  Sun Dec 24 06:15:23 2000
From: mal at bewoner.dma.be (Lieven Marchand)
Date: 24 Dec 2000 12:15:23 +0100
Subject: One Python 2.1 idea
References: <mailman.977590932.19335.python-list@python.org> <9237md$orm$1@nnrp1.deja.com> <924bjt0e3n@news1.newsguy.com>
Message-ID: <m38zp6nk8k.fsf@localhost.localdomain>

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

> Actually, for all-out speed in a higher-level language, Scheme
> (with the stalin compiler) or Common Lisp probably cannot be
> beat.  There ARE, after all, decades of Lisp experience and many
> groups of clever people laboring away at getting such speed.

I think Common Lisp shows that getting very good speed in Python is
quite feasable. The original CMUCL implementers weren't such a large
group. Their highly optimising compiler (which incidentally is also
called Python) has on occasion beaten FORTRAN at numerics. I don't
know why some people in the Python community think compiling Python is
such a problem. Practically all the problems have been tackled and
solved 20 years ago in the Lisp community.

-- 
Lieven Marchand <mal at bewoner.dma.be>
Lambda calculus - Call us a mad club


From darcy at vex.net  Sat Dec  9 13:04:29 2000
From: darcy at vex.net (D'Arcy J.M. Cain)
Date: 9 Dec 2000 18:04:29 GMT
Subject: Using python-PgSQL and the doublequote ("") problem...!
References: <3a31ba9e.105555070@news-server.bigpond.net.au>
Message-ID: <90ts7c$2k1$1@news.tht.net>

sorular <sorular at netscape.net> wrote:
> Is there a standard sintax to filter the data before entering it into
> PgSQL database. to prevent parsing errors ?

Using PyGreSQL or something else?  If PyGreSQL are you using the "classic"
API or the DB-SIG 2.0 API?  By double quote do you mean empty string or
are you talking about embedding quoting characters within your string
and are wondering about quote reduction as strings are evaluated.  If the
following doesn't answer your question try to clarify some of these issues.

In PyGreSQL I have a function that does the following before putting a
string value into the database.

    return "'%s'" % string.strip(re.sub("'", "''", \
                             re.sub("\\\\", "\\\\\\\\", "%s" %d)))

This return the string inside of the single quotes that SQL needs.  Of
course if you are using PyGreSQL then you don't need to worry about this
as this function will be called for you when you insert or update.

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


From tismer at tismer.com  Fri Dec 15 06:12:15 2000
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 15 Dec 2000 13:12:15 +0200
Subject: Getting a list of member functions
References: <200012142240.PAA14787@mail.cybermesa.com>
Message-ID: <3A39FC8F.6B478638@tismer.com>


Jay O'Connor wrote:
> 
> Hi,
> 
>         I'm trying to get a list of the member functions for an object.  I would think __members__ or __methods__ would work
> 
> Here's what I'm getting

[of course not working...]

> This is on Pythong 1.5.2
> 
> Did I miss something?
> 
> I know I could probably loop through the classes __dict__ and pull out the functions, but that's kinda kludgey and would miss instance specific methods

Classes do not implement __members__ and __methods__ automagically,
but you can do this yourself.

If you want to provide static members for instance, just put
a list
  __members__ = ["var1", "var2", "var3"]
into your class.
If you want to make this more dynamic or instance specific,
it is a little more complicated, since you then have to
overwrite __getattr__ and compute the __members__ attribute
on-the-fly.

>>> class demo:
... 	def __init__(self, **vars):
... 		self.__dict__.update(vars)
... 	def __getattr__(self, key):
... 		if key=="__members__":
... 			return self.__dict__.keys()
... 		raise AttributeError, key
... 	
>>> x=demo(a=5, b=7, c=42)
>>> x.__members__
['b', 'c', 'a']
>>> x.x=0
>>> x.__members__
['b', 'c', 'x', 'a']
>>> 

Of course this gets more clumsy if you also have to support
instance-specific methods. But I doubt that these are
so urgent to have at all.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From aleaxit at yahoo.com  Thu Dec 14 15:47:02 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 14 Dec 2000 21:47:02 +0100
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <r0pzoi11lhz.fsf@black.cs.washington.edu> <5jvZ5.3588$Kd1.418770@newsb.telia.net> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com> <kuu28971kp.fsf@lasipalatsi.fi> <3A380E4E.8537A0FB@san.rr.com> <91a1vr02b02@news2.newsguy.com> <Co9_5.25465$x6.13756579@news2.rdc2.tx.home.com>
Message-ID: <91bbq102mtk@news1.newsguy.com>

"Rainer Deyke" <root at rainerdeyke.com> wrote in message
news:Co9_5.25465$x6.13756579 at news2.rdc2.tx.home.com...
> "Alex Martelli" <aleaxit at yahoo.com> wrote in message
> news:91a1vr02b02 at news2.newsguy.com...
> > There is no value *IN* x.
> >
> > A variable *REFERS TO* 'a value' (aka 'an object').
>
> If somebody really wants mutable integer objects, they're not that hard to
> create.
>
> class MutableInt(UserInt.UserInt):
>   def set(self, value):
>     self.data = value

UserInt is not in the Python 2 docs, though it's nice to hear it exists:-).

Anyway, I had already provided a more generic class Mutable elsewhere
on (this or a parallel) thread 1-2 days ago -- including the explanation
of why I want a 'return value' to end this set statement...:-)


Alex





From bragi at hks.com  Wed Dec  6 16:35:46 2000
From: bragi at hks.com (Bragi Baldursson)
Date: Wed, 06 Dec 2000 16:35:46 -0500
Subject: easy question on input files in Python
References: <908jvp$dd1$1@nnrp1.deja.com>
Message-ID: <3A2EB131.31CDB5E9@hks.com>

handle = open('file.inp', 'r')
lines = handle.readlines()
for line in lines:
   print line

or in a while loop.

while 1:
    line = handle.readline()
    if line:
        print line
     if not line:
        break

Bragi


jsolack at my-deja.com wrote:

> I am just messing around with Python and I only have the tutorial off
> of python.org.  I was wondering how I would set up a while loop that
> says "while not end of file" in python?  I would like to read through
> an input file and i do not know the sytax for that!  Thanks!
>
> Jason
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.

--
___________________________________________________

Bragi Baldursson
Hibbitt, Karlsson, & Sorensen, Inc.
web:  www.hks.com             mailto: bragi at hks.com
Phone: (401) 727-4380 x5586   Fax: (401) 727-4208
___________________________________________________


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20001206/451107c1/attachment.html>

From jonathan.gilligan at vanderbilt.edu  Tue Dec  5 02:19:39 2000
From: jonathan.gilligan at vanderbilt.edu (Jonathan Gilligan)
Date: Tue, 5 Dec 2000 01:19:39 -0600
Subject: How do packages and .pth files work under Windows (Py 2.0)
Message-ID: <90i4uc$lin$1@news.vanderbilt.edu>

I am having problems understanding how .pth files are supposed to work
under windows. The Python 2.0 documentation has several inconsistent
descriptions.

In one place, it says that under Windows and Mac, .pth files are only
interpreted if they are in the python root. Elsewhere the
documentation indicates that they are interpreted if they are in the
lib\site-packages directory.

Here is the basic problem. I have packages bar, far, and gar, all of
which I want to be subpackages in the package foo, so I can write
"import foo.bar as FB". If bar is a working package, the Python
documentation suggests that II should be able to do what I want if I
put the directory bar (containing the bar package, which has a valid
__init__.py) inside a directory "site-packages\foo" and put a file
"foo.pth" in  the site-packages directory with the single line "foo".

However, this does not work for me. Somehow, Python does not seem to
understand the foo.pth file, and python complains "ImportError: No
module named foo"

Here are the details:

I am trying to get the mx tools installed on my computer and have
python 2.0 built from cvs sources in D:\Sources\python\python\dist\src
and am trying to get mxTextTools installed somewhere so that I can use
the following line (from the Martel genome parsing package):

    from mx import TextTools as TT

I have installed mxTextTools in
D:\Sources\Python\Python\dist\src\lib\site-packages\mx\TextTools. If I
set PYTHONPATH to
"D:\Sources\\Python\Python\dist\src\lib\site-packages\mx" then I can
"import TextTools" but if I set PYTHONPATH to
"D:\Sources\Python\Python\dist\src\lib\site-packages" then I cannot
"import mx.TextTools". I keep getting an error message:

    "ImportError: No module named mx"

This occurs even if I create a "mx.pth" file in "site-packages" with
the single line "mx". I have also tried putting the "mx" directory and
the "mx.pth" file in the python root
(D:\Sources\python\python\dist\src).

I have the registry correctly set up:

"HKLM\Software\Python\PythonCore\2.0" has subkeys

"InstallPath" with default value "D:\Sources\Python\Python\dist\src"
"PythonPath" with default value
"D:\Sources\Python\Python\dist\src\lib;D:\Sources\Python\Python\dist\s
rc\lib\site-packages;d:\Sources\python\python\dist\src\pcbuild;D:\Sour
ces\Python"

Can anyone help me figure out how to get packages to work in the way I
am trying to?

It would help me enormously if there were a way to get Python to print
out a list of where it looks when it tries to resolve an "import". Is
this possible?

Thanks for any help that anyone can give me.

Jonathan Gilligan




From fgeiger at datec.at  Fri Dec 29 12:18:24 2000
From: fgeiger at datec.at (Franz GEIGER)
Date: Fri, 29 Dec 2000 18:18:24 +0100
Subject: 8.3 peculiarity if run w/i console
References: <92g4dn$5ug$1@newsreaderm1.core.theplanet.net> <VjP26.36442$xW4.287502@news-server.bigpond.net.au>
Message-ID: <92ih6d$bt0$1@newsreaderm1.core.theplanet.net>

>    What is it that is saying the name is abbreviated? On Windows, all file

Well, actually that's me, I call it "abbreviated". I know there are those
two forms of file names, but I dont like the one containing "~1". How can I
get rid of that?

Thanks, Neil,
and best regards
Franz GEIGER


"Neil Hodgson" <neilh at scintilla.org> wrote in message
news:VjP26.36442$xW4.287502 at news-server.bigpond.net.au...
> Franz GEIGER:
>
> > If I run a script within the Python 2.0 console, the names of the files
> that
> > are made by that script are all "abbreviated", e.g.:
> > ALongFileName.txt becomes ALongF~1.TXT. If I run the script within
> > PythonWin, the names are o.k.
>
>    I get the same behaviour from console and PythonWin on W2K with:
> f = open("ALongFileName.txt", "w")
> f.close()
>
>    What is it that is saying the name is abbreviated? On Windows, all file
> names are stored in two forms - a long form and an abbreviated form. The
> abbreviated form is for use by old software which can't cope with the long
> names. If you do:
> dir /x
>    then you will see both forms.
>
>    Neil
>
>




From cjw at connection.com  Tue Dec  5 12:25:01 2000
From: cjw at connection.com (Colin J. Williams)
Date: Tue, 05 Dec 2000 12:25:01 -0500
Subject: PythonWin
Message-ID: <3A2D24ED.96551F5A@connection.com>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20001205/2462dab4/attachment.html>

From phd at phd.pp.ru  Tue Dec  5 08:57:06 2000
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 5 Dec 2000 16:57:06 +0300 (MSK)
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Dec  4)
In-Reply-To: <90iqs8$n6k$1@newsreaderm1.core.theplanet.net>
Message-ID: <Pine.LNX.4.30.0012051656050.25274-100000@fep132.fep.ru>

Nost links there points to nowhere. Deja complains "cannot find the
document" :(

On Tue, 5 Dec 2000, Franz GEIGER wrote:
> >     Alex Martelli shows a clever way to overload a method based on
> >     number of arguments.
> >         http://deja.com/=dnc/getdoc.xp?AN=975947338
>
> This link leads to "Invalid document request". How is the right one?
>
> Thanks in advance and
> best regards
> Franz GEIGER
>
>
> Grant Edwards <grante at visi.com> schrieb in im Newsbeitrag:
> 8C419A8040EAFDCB.1774103991763D9E.2F034C54E78FF82C at lp.airnews.net...
> >     Alex Martelli discusses the issues involved in using Python
> >     within Windows Script Control.
> >         http://deja.com/=dnc/getdoc.xp?AN=698664345
> >     Alex Martelli shows a clever way to overload a method based on
> >     number of arguments.
> >         http://deja.com/=dnc/getdoc.xp?AN=975947338
> >
> >     Dale Strickland-Clark and Alex Martelli on using Python
> >     programs as drop targets under Win32.
> >         http://deja.com/=dnc/getdoc.xp?AN=700537932
> >         http://deja.com/=dnc/getdoc.xp?AN=700557153
> >
> >     Andrew H. Chatham announces Python bindings for Ogg/Vorbis (MP3
> >     patent/licensing free replacement for MP3) -- and requests help
> >     with dlopen().
> >         http://deja.com/=dnc/getdoc.xp?AN=975946985
> >
> >     Nick Mathewson shows how to implement a general overloading
> >     module for python.
> >         http://deja.com/=dnc/getdoc.xp?AN=700076223
> >
> >     Gordon McMillan releases fixed Installer for Python 2.0
> >         http://deja.com/=dnc/getdoc.xp?AN=700179180

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




From jhg at galdon.com  Fri Dec 29 04:59:11 2000
From: jhg at galdon.com (Juan Huertas)
Date: Fri, 29 Dec 2000 10:59:11 +0100
Subject: interpretter and available memory
References: <92d01r$1vut$1@news.tht.net>
Message-ID: <92hmts$2el$1@talia.mad.ttd.net>

wxPython has a function: wxGetFreeMemory() you can aid. Try in
www.wxPython.org

Juan H. Galdon

jhg at galdon.com





From bkc at Murkworks.com  Fri Dec 22 12:45:13 2000
From: bkc at Murkworks.com (Brad Clements)
Date: Fri, 22 Dec 2000 12:45:13 -0500
Subject: Best documentation for writing wrappers for C++ classes
References: <9EB39E4E3AF64120.C259930B9E4BEFF5.1C1A4B99EAA59011@lp.airnews.net>
Message-ID: <poM06.121$Lm2.7326@news.clarkson.edu>

Certainly not a standard, but check out Wrox Press 'Professional Linux
Programming'

It has a chapter dedicated to extending and embedding with C/C++.

I'm interested in comments, if you've read it btw.. I wrote the chapter.

--
Brad Clements,
bkc at murkworks.com
"David L. Humphrey" <dhumphrey at acm.org> wrote in message
news:9EB39E4E3AF64120.C259930B9E4BEFF5.1C1A4B99EAA59011 at lp.airnews.net...
> I have several C++ classes that I would like to call from Python code.  I
> have reviewed the "Extending and Embedding the Python Interpreter" as well
> as selected portions of "Programming Python" but worry a little about the
> currency of these sources as well as the general tone of extending with C
> code rather than C++ classes.  Is there a document that is recognized as
the
> standard reference for this information?
>
> Thanks in advance,
> David L. Humphrey
> Manager, Software Development
> Bell Geospace, Inc
>
>




From loewis at informatik.hu-berlin.de  Fri Dec  8 14:50:43 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 08 Dec 2000 20:50:43 +0100
Subject: ANN.: Weak Reference Extension Module has been converted to "C"; Beta 2 is out
References: <8usorp$3i3$1@nnrp1.deja.com> <mtaeb1f4q3.fsf@astron.berkeley.edu> <8uvjmj$dt8$1@nnrp1.deja.com> <mtlmuk9hyc.fsf@astron.berkeley.edu> <B63C5A9E.172D7%graham73@telocity.com> <3A176E0F.39BF350B@felinity.com> <KaNR5.5550$QH2.654863@newsb.telia.net> <8v9dl9$si6$1@nnrp1.deja.com> <90i3e3$f8t$1@nnrp1.deja.com>
Message-ID: <p6qu28ek7zg.fsf@informatik.hu-berlin.de>

shindich at my-deja.com writes:

> I am happy to announce that Weak Reference Extension project has been
> fully converted to "C". Weak Reference extension is now completely
> CPython compatible.

That's indeed a big improvement. Congratulations!

Martin


From tino at wildenhain.de  Sun Dec 24 05:10:27 2000
From: tino at wildenhain.de (Tino Wildenhain)
Date: Sun, 24 Dec 2000 11:10:27 +0100
Subject: string split
References: <slrn94b2go.srm.jp@localhost.localdomain>
Message-ID: <3A45CB93.4FC5BC6@wildenhain.de>

Hi Jacek,
...snip...
> is it possible to split not by space (" ") but by "<" and ">" ?
> 
> I need:
> 
> ['one two','<br>','three']
> 
> How to do it in simple way?

You can use the second argument to string.split() as split-string:

split(str [,sep [,maxsplit]]) -> list of strings
splitfields(str [,sep [,maxsplit]]) -> list of strings

Return a list of the words in the string s, using sep as the
delimiter string.  If maxsplit is nonzero, splits into at most
maxsplit words If sep is not specified, any whitespace string
is a separator.  Maxsplit defaults to 0.

Anyway, using sgmllib or htmllib could be much more convenient.

Regards
Tino



From nickb at fnord.io.com  Thu Dec  7 17:42:05 2000
From: nickb at fnord.io.com (Nick Bensema)
Date: Thu, 07 Dec 2000 22:42:05 GMT
Subject: Programming Habits in Python
References: <xp0U5.76197$U46.2386878@news1.sttls1.wa.home.com> <8vsiq4$mmk$1@news.wrc.xerox.com> <cspU5.82459$U46.2617014@news1.sttls1.wa.home.com>
Message-ID: <ZmUX5.105599$DG3.2028004@news2.giganews.com>

In article <cspU5.82459$U46.2617014 at news1.sttls1.wa.home.com>,
Greg Jorgensen <gregj at pobox.com> wrote:
>I don't write them off. Part of my job description is to mentor the more
>junior programmers (who start at $50K these days if they claim to know
>Visual Basic), but it's an uphill struggle when they don't know binary math,
>can't read debugger output, are unfamiliar with common algorithms, and can't
>debug the simplest problems in their own code. 

I'll say this for DeVry.... they taught us a lot of things that made a
lot of people complain, and among them was JCL.

We didn't get to use cool things like keypunches, but we used TN3270
to do our work on a server in Chicago.  We'd submit a JCL job and a few
minutes later, stacks of extra-wide white and green paper would appear on
the FA's desk.  It was there that we learned about reading debugger output
and EBCDIC, and some of the nuances of batch processing.  The most brutal
part, of course, were the two-hour exams in which the teacher would give
us a pile of code, and a list of user input, and we had to trace through
the program and tell which subroutines were executed, and so forth.

Furthermore, even in our C++ class, we would be given the occasional
trace-the-program exercises where we'd have to keep track of global
and local variables with the same names, and recursive function calls,
and code that didn't seem to make any sense.

It went all the way back to the C course I took one summer.  I knew
already how to convert decimal to hex to binary, sometimes in my head,
but it was there that I was asked to actually add numbers in base two,
and subtract them.  

I wonder, though, if this kind of teaching is commonplace in
universities and community colleges anymore.  I wonder if DeVry
even still teaches it.  I at least suspect that a JCL course might not
be required for a computer-related degree in many universities.


-- 
Nick Bensema <nickb at io.com>      ICQ#2135445 
==== ======= ==============
CREDIT 0     THY GAME IS OVER     NO CARRIER


From jnix at medi.physik.uni-oldenburg.de  Fri Dec 22 10:35:23 2000
From: jnix at medi.physik.uni-oldenburg.de (Johannes Nix)
Date: 22 Dec 2000 16:35:23 +0100
Subject: SMP using Shared memory (was: Re: Is Stackless and uthread.py SMP friendly?)
References: <91br9v$qa4$1@nnrp1.deja.com> <ku1yva5ocq.fsf@lasipalatsi.fi> <91dp4j$5fj$1@news.gte.com> <91dt39$fbk$1@nnrp1.deja.com>
Message-ID: <91vsbu$l97@news.Informatik.Uni-Oldenburg.DE>

jbalding at hotmail.com writes:

> Okay. My next question: Is there a simple way of taking advantage of
> multiple processors using Python, or is this something that is going to
> be like pulling teeth? In other words, are there existing packages that
> allow relatively high-level, transparent access to multiple processors?
> 

I had the same problem; I needed to compute parellel using Numerical
Python on a dual processor workstation.

My solution was to get the shm module (see
http://www.vex.net/parnassus/apyllo.py/990075885.17269372 or
http://sirac.inrialpes.fr/~marangoz/python/shm/), and to write two
small wrapper classes around it to transfer arrays.

Then, I start up my process, do all the initialization, do an
os.fork(), and start passing arrays via shared memory. If you want to
use tkinter, you need to initialize it _after_ os.fork(), or strange
things will happen.


This is nearly as easy as using threads, but works fine, and seems a
lot easier than using PVM. It is also much faster, since shared memory
means that two processes share the same adress space, and no kernel or
networking code needs to run for transfers.

Some remarks about using shared memory on Linux: Note, to compile
shmmodule.c on linux, you need the patch below.  Note also, that the
keys/IDs need to have values larger than zero.

Finally note, the resource limits on Linux can be queried using the
command "ipcs -l", and you can change system parameters in
/usr/src/linux/include/asm/shmparam.h and
/usr/src/linux/include/linux/sem.h (thanks to Detlef Antoni for
pointing this out).


Personally, I would like it very much if Python would get rid of the
global lock, and introduces something like object locks. I think for a
number of people this would be a good reason to switch from Matlab to
Python, or at least give it a try.

and believe me, parallel crunching with Python is FUN!

Johannes


=======================================================================
# file: shmbuf.py
#!/usr/bin/python

""" this module provides a shared memory buffer for arbitrary strings.
It can be used to exchange data between different Python processes,
to overcome the problem with the Global Interpreter Lock
on SMP machines."""

import shm
import sys
from struct import calcsize, pack, unpack

WLOCK='wlock'
RLOCK='rlock'
MUTEX='mutex'

class SHMBuf:
    def __init__(self, shm_key, sem_mutex_key, sem_wlock_key, sem_rlock_key, msize):
        if not (shm_key and sem_mutex_key and sem_wlock_key and sem_rlock_key):
            raise "zero keys do not work."
        self.sems={}
        for tag, semkey, initval in [ (MUTEX,sem_mutex_key,1),
                                      (WLOCK, sem_wlock_key,1),
                                      (RLOCK, sem_rlock_key,0)]:
            try:
                if shm.semaphore_haskey(semkey):
                    s = shm.semaphore(shm.getsemid(semkey))
                    s.setval(initval)
                else:
                    s = shm.create_semaphore(semkey,initval)
            except:
                print "can't acces / allocate semaphore with key %s" % semkey
                raise
            self.sems[tag]=s
        try:
            if shm.memory_haskey(shm_key):
                self.sharedmem = shm.memory(shm.getshmid(shm_key))
            else:
                self.sharedmem = shm.create_memory(shm_key,msize)
        except:
            print "can't access / allocate shared memory with key %s" % shm_key
            raise
        self.sharedmem.attach()

        self.size_long = calcsize('l')

    def write(self,string):
        self.sems[WLOCK].P()
        self.sems[MUTEX].P()
        self.sharedmem.write(pack('l',len(string))+string)
        self.sems[MUTEX].V()
        self.sems[RLOCK].V()

    def read(self):
        self.sems[RLOCK].P()
        self.sems[MUTEX].P()
        size = unpack('l',self.sharedmem.read(self.size_long))[0]
        string = self.sharedmem.read(size,self.size_long)
        self.sems[MUTEX].V()
        self.sems[WLOCK].V()
        return string
===========================================================================
# File: shmarray.py
#!/usr/bin/python


"""This module provides exchange of Numerical array instances
between independent Python processes.

Array shape has to be known, because all arrays come out flattened
(no problem to fix this, but we want it very efficient and it isn't
needed)"""


from shmbuf import *
from Numeric import fromstring

class SHMArray(SHMBuf):
    def write_array(self,array):
        SHMBuf.write(self,array.typecode() + array.tostring())


    def read_array(self):
        s = SHMBuf.read(self)
        return fromstring(s[1:],s[0])



##################################################################

if __name__ == "__main__":
    print "testing..."
    smar = SHMArray(1,1,2,3,1024)

    import os
    import sys
    import time
    from Numeric import arange

    if len(sys.argv) > 1:
        print sys.argv[1]
        tty = open(sys.argv[1],'w')
        print "hello"
        tty.write('Hi there\n')
    else:
        tty = sys.stdout


    child_pid=os.fork()


    if not child_pid:
        # child path
        
        a = arange(20)
        try:
            for i in range(1000):
                smar.write_array(a)
		a  = a * 1.1

        finally:
            print 'child: ready, terminating'
            sys.exit(0)       



    else:
        # this is the parent process
        print "child is:",child_pid
        for i in range(1000):
            tty.write(repr(smar.read_array())+'\n')
        tty.write('parent: ready, terminating\n\n')
        time.sleep(0.05)          # this is for safety, in case
        os.kill(child_pid,9)   # ... child is still waiting.
        sys.exit(0)
            

            
========================================================================
/* patch for shmmodule.py */


--- shm.orig/shmmodule.c	Wed Nov 22 20:12:25 2000
+++ shm/shmmodule.c	Fri Dec 22 16:21:22 2000
@@ -173,6 +173,11 @@
     struct semid_ds ds;		/* data structure	*/
 } PyShmSemaphoreObject;
 
+#ifdef _SEM_SEMUN_UNDEFINED
+#undef HAVE_UNION_SEMUN
+#endif
+
+
 #ifndef HAVE_UNION_SEMUN
 union semun {
     int val;                    /* used for SETVAL only */
@@ -429,7 +434,7 @@
     PyObject *args;
 {
     char *data;
-    unsigned long n, offset = 0;
+    int n, offset = 0;
     char buf[128];
     char *addr;
 


From mail at codex.pe.kr  Thu Dec 21 11:44:24 2000
From: mail at codex.pe.kr (Ahn, Jae-wook)
Date: Fri, 22 Dec 2000 01:44:24 +0900
Subject: Python Image Resize Problem
References: <mailman.977412309.21408.python-list@python.org>
Message-ID: <B6686277.143F%mail@codex.pe.kr>

No, it's Korean language.
Maybe someone made a mistake.




From bobhicks at adelphia.net  Fri Dec 29 20:43:27 2000
From: bobhicks at adelphia.net (Robert L Hicks)
Date: Sat, 30 Dec 2000 01:43:27 GMT
Subject: Komodo
Message-ID: <B672A66F.72CC%bobhicks@adelphia.net>

Has anyone tried Komodo from ActiveState yet? What are you thoughts on it? I
see that a Linux beta is out now as well.

Will Komodo for Linux have to have the ActiveState versions of Perl and
Python to work?

Bob



From spahievi at vega.bg  Tue Dec 12 17:43:39 2000
From: spahievi at vega.bg (Niki Spahiev)
Date: Wed, 13 Dec 2000 00:43:39 +0200
Subject: DangerousExceptions (was Re[2]: [Q] Are Exceptions used that much in practice?)
In-Reply-To: <Q4vZ5.19457$x6.10649942@news2.rdc2.tx.home.com>
References: <3A35D546.A4731B5B@enteract.com> <914mfd02ibt@news2.newsguy.com>
 <Q4vZ5.19457$x6.10649942@news2.rdc2.tx.home.com>
Message-ID: <196647430.20001213004339@vega.bg>

12.12.2000, 21:34:08, Rainer Deyke wrote:

RD> "Alex Martelli" <aleaxit at yahoo.com> wrote in message
RD> news:914mfd02ibt at news2.newsguy.com...
>> (Some might prefer a wider 'except: return 0' clause, but
>> most often it's better to be specific in catching exceptions).

RD> I would go further and say that 'except: return 0' should never be used,
RD> since it would catch and ignore MemoryError, KeyboardInterrupt, SystemExit,
RD> ComputerExplodingError, or any number of other exceptions that are totally
RD> unrelated to the issue at hand and require careful handling and/or program
RD> termination.

While at this i always wished that Exception has two subclasses:
"NormalException" and "DangerousException" so one can use

except NormalException: return 0

and not worry about  MemoryError, KeyboardInterrupt, SystemExit,
 ComputerExplodingError :) being improperly handled.


-- 
Best regards,
 Niki Spahiev





From wware at world.std.com  Thu Dec  7 21:09:43 2000
From: wware at world.std.com (Will Ware)
Date: Fri, 8 Dec 2000 02:09:43 GMT
Subject: Music synthesizer in Python/C
References: <G588B3.6A5@world.std.com>
Message-ID: <G588o8.DC7@world.std.com>

Will Ware (wware at world.std.com) wrote:
> http://www.deja.com/getdoc.xp?AN=702341030&fmt=text

I posted the resulting tune as a uuencoded MP3 on
alt.binaries.sounds.music.amateur, it hasn't appeared on deja.com
yet but here's the posting headers if anybody is curious:
From: wware at world.std.com (Will Ware)
Subject: Synthesizer results
Message-ID: <G587D6.BDu at world.std.com>
Date: Fri, 8 Dec 2000 01:41:30 GMT
Lines: 17966
-- 
# - - - - - - - - - - - - - - - - - - - - - - - -
# Resistance is futile. Capacitance is efficacious.
# Will Ware	email:    wware @ world.std.com


From dalke at acm.org  Wed Dec 13 22:26:02 2000
From: dalke at acm.org (Andrew Dalke)
Date: Wed, 13 Dec 2000 20:26:02 -0700
Subject: Programming Habits in Python
References: <cspU5.82459$U46.2617014@news1.sttls1.wa.home.com> <8vvb6p$ct2$1@news.wrc.xerox.com> <XiMU5.87123$U46.2803855@news1.sttls1.wa.home.com> <900cpl02ss1@news2.newsguy.com> <Ue3V5.91231$U46.2956112@news1.sttls1.wa.home.com> <slrn92tv0u.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com> <910m5d$t6v$1@slb3.atl.mindspring.net> <k_j_r_a_y-958C59.19032913122000@news>
Message-ID: <919ejl$8l3$1@slb2.atl.mindspring.net>

Keith Ray wrote:
>Could you have put the code for each of those 12 states into separate
>classes, and put instances of those classes into a dictionary where the
>key is the state-id?

You are refering to the mxTextTools example, I presume, and not
the C++/MIX one?

No, I could not.  I think you don't know how mxTextTools works.
It does all the work in C, which makes it fast but precludes
easy calls to Python.  Now, it does allows Python calls, but the
peformance would be even worse, and I needed that performance.
It ended up, btw, that mxTextTools was faster than Python's
readlines solution *and* let me parse \n, \r\n and \r line
endings.  (I believe the speed comes from using Boyer-Moore
because I'm actually looking for a line which starts with a
given text, so there's more text it can know to skip.)

                    Andrew
                    dalke at acm.org





From quinn at dinar.ugcs.caltech.edu  Thu Dec 14 15:29:58 2000
From: quinn at dinar.ugcs.caltech.edu (Quinn Dunkan)
Date: 14 Dec 2000 20:29:58 GMT
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <tcmay-FCA22F.14073713122000@news.got.net> <918v7a02scn@news1.newsguy.com> <tcmay-04E80C.16181313122000@news.got.net> <91a1ed02a4j@news2.newsguy.com> <3A389B1A.4F33FFD4@cfmu.eurocontrol.be> <91ae8s02llf@news2.newsguy.com>
Message-ID: <slrn93ibe6.cg1.quinn@dinar.ugcs.caltech.edu>

On Thu, 14 Dec 2000 13:25:51 +0100, Alex Martelli <aleaxit at yahoo.com> wrote:
>"Ian Wild" <ian at cfmu.eurocontrol.be> wrote in message
>news:3A389B1A.4F33FFD4 at cfmu.eurocontrol.be...
>> Alex Martelli wrote:
>> >
>> > I can't find a source any more, but one of the most interesting
>> > arguments I've heard for "only inherit from abstract classes"
>> [...otherwise...]
>> > it would break the idea that your concrete
>> > classes partition the universe of discourse, if it were.
>>
>> Where does this partitioning requirement come from?
>
>From the fact that an object 'is an instance' of one,
>and only one, class -- 'THE class' (singular) 'of the
>object'.  If an object 'belongs' to n>1 classes, in
>this framing, this is modeled by having the (inevitably
>singular) class 'of the object' _inherit_ from the
>said n>1 classes (i.e., 'belongs' is a weaker relationship
>than 'being an instance of').
>
>I am not aware of any OOA approaches in which is-instance-of
>(as here defined) isn't an exhaustive and many-to-one
>relationship between objects and (concrete) classes;
>in other words, OOA approaches which let some objects
>(in the universe-of-discourse that is being modeled)
>be an instance of 'no class', or of several classes
>at the same time.

Well, I'm not quite sure if it's what you're talking about (all this high
falutin' "OOA" stuff tends to zip by over my head), but perhaps haskell's
typeclass system is like this.  In haskell, a class describes an interface,
and any type that supplies that interface can function as an instance of that
class.  So:

data Rectangle = Rect Int Int
data Sailor = Sailor String
data Poodle = Poodle String

class Drawable t where
    draw :: t -> String
class Pirate t where 
    pirate_talk :: t -> String

instance Drawable Rectangle where
    draw (Rect x y) =
        "a rectangle with sides " ++ (show x) ++ " and " ++ (show y)

instance Drawable Sailor where
    draw (Sailor name) = "a sea-dog named " ++ name
instance Pirate Sailor where
    pirate_talk (Sailor name) = name ++ " growls, 'arr, matey!'"

So, a Rectangle is an instance of Drawable (so it will be accepted by any
function that wants a Drawable type), a sailor is an instance of both Drawable
and Pirate, and a Poodle not an instance of any class (so it will only be
accepted by functions that want a Poodle type).  No inheritance is involved.


Or maybe you're saying there is no developed method of analysis for this sort
of thing.  It would surprise me if there weren't, since they're a pretty
academic bunch over there in FP-land :)


From garryknight at bigfoot.com  Sat Dec 23 08:38:54 2000
From: garryknight at bigfoot.com (Garry Knight)
Date: Sat, 23 Dec 2000 13:38:54 +0000
Subject: coding convention conversion
References: <mailman.977516469.1157.python-list@python.org> <920v4s$oqj$2@taliesin.netcom.net.uk> <mailman.977547906.22999.python-list@python.org>
Message-ID: <9229te$8nl$2@taliesin.netcom.net.uk>

In article <mailman.977547906.22999.python-list at python.org> "David
Porter" <jcm at bigskytel.com> wrote:

> * Garry Knight <garryknight at bigfoot.com>:
>> In article <mailman.977516469.1157.python-list at python.org> "Issac
>> Trotts" <trotts at llnl.gov> wrote:
>> 
>> > Does anyone know of a utility to change between the coding
>> > conventions
>> >  of
>> > 
>> > def fooBarBaz(): pass
>> > 
>> > and 
>> > 
>> > def foo_bar_baz(): pass
>> > 
>> > It is a pain to do this by hand.

[snip program]
 
> It only changes the def lines. It should also alter the lines that
> call the function (but not imported functions):

Sorry, this was one of those 1:30 am postings. I'll amend it when I get
back later tonight.  :o(

-- 
Garry Knight
garryknight at bigfoot.com


From clgonsal at keeshah.penguinpowered.com  Thu Dec  7 12:04:17 2000
From: clgonsal at keeshah.penguinpowered.com (C.Laurence Gonsalves)
Date: Thu, 07 Dec 2000 17:04:17 GMT
Subject: DCOracle Memory Leak?
References: <m3g0k1qih5.fsf@h0050047c4794.ne.mediaone.net> <t2v9fq7v9ns565@corp.supernews.com>
Message-ID: <slrn92vgq7.130.clgonsal@keeshah.penguinpowered.com>

On Thu, 7 Dec 2000 10:47:55 -0400, Jose Martin <joseamartin at venmex.com> wrote:
> Hi, sorry i cant help you, but , can you help me ?
> I am trying to use DCOracle on Python 2.0 under Windows or Linux
> but when i try to make my scritp :
> 
> The first line is
> from DCOracle import *
> 
> when i try to run:
> Python say me:
> 
> File "d:\python20\DCOracle\__init__.py", line 52, in ?
>     import sys, Buffer
> ImportError: No module named Buffer
> 
> I was searched my python path for a module called Buffer, but i dont have it
> Whats happennig ?

The Buffer module is actually a shared library. There are two shared
libraries that are part of DCOracle: Buffer and oci_. Look for
"Buffer.so" and "oci_.so". If you're missing both of these, then it
sounds like you might have just installed the *.py files without
building the *.so files. If that's the case, see the Readme file that
comes with DCOracle for instructions on building it.

-- 
  C. Laurence Gonsalves                "Any sufficiently advanced
  clgonsal at kami.com                     technology is indistinguishable
  http://cryogen.com/clgonsal/          from magic." -- Arthur C. Clarke


From MarkH at ActiveState.com  Thu Dec 28 07:23:14 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Thu, 28 Dec 2000 12:23:14 GMT
Subject: COM / Variants / VT_BOOL
References: <mailman.977467384.30398.python-list@python.org> <91v2sv01rpp@news2.newsguy.com>
Message-ID: <3A4B2F8C.4000909@ActiveState.com>

Alex Martelli wrote:


> *Almost* right... but not *quite*, and we hit this specific problem
> quite recently (we weren't driving with Python, but I think that
> would have made no difference).

Quite possibly it would.  If win32com had type information, it would 
have dont the right thing.

> Setting the property .ValidateOnParse of the Xerces XML parser
> (originally from IBM, now, I believe, from Apache), in its COM
> setting, to 0 *or* 1, still keeps the parser non-validating; you

Without type info, the parser should call VariantChangeType() to get a 
valid VT_BOOL back.  So really the parser is at fault (or the client for 
ignoring it)

> *Most* COM servers will happily accept 'server.BoolProp = 1'
> as setting the property to 'true', but, I suggest, don't count
> on that; use 'server.BoolProp = -1' and you might be safer.

Also, a feature of the pythoncom type conversion means that you can say:

true = 1==1
false = 1==0

And passing these variables (or the expression as a literal) _will_ 
force a VT_BOOL of the correct type.

Mark.



From sholden at holdenweb.com  Thu Dec 21 08:36:13 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 21 Dec 2000 08:36:13 -0500
Subject: while true: !!!
References: <3A4024E5.98843CFC@cosc.canterbury.ac.nz> <91qadb$1di$1@plutonium.compulink.co.uk> <3A41539E.DF6977D6@cosc.canterbury.ac.nz>
Message-ID: <UIn06.2878$PP2.31080@e420r-atl2.usenetserver.com>

Greg Ewing <greg at cosc.canterbury.ac.nz> wrote in message
news:3A41539E.DF6977D6 at cosc.canterbury.ac.nz...
> gbreed at cix.compulink.co.uk wrote:
> >
> > Take any type, and if
> > there's a way of it being nothing, it evaluates false.
>
> Hmmm, let's try that out. Is [[]] true or false?
> It's a list containing a list containing nothing.
> Since [] is false, a list containing nothing must be
> nothing. So [[]] is a list containing nothing,
> which is nothing. Ah, it must be false! :-)
>
> Nothing-is-weird-stuff,
>
if [[]]:
    print "Ewing is wrong"
else:
    print "Ewing is right"

Ewing is wrong

> --
> Greg Ewing, Computer Science Dept, University of Canterbury,
> Christchurch, New Zealand
> To get my email address, please visit my web page:
> http://www.cosc.canterbury.ac.nz/~greg
>
Clearly your sophistry does not suffice to fool the interpreter, which
obviously "knows" that [[]] is not nothing, but a list containing a list
containing nothing.

if-you-wrap-nothing-up-you've-got-something-ly y'rs  - steve






From aahz at panix.com  Sun Dec 17 13:58:45 2000
From: aahz at panix.com (Aahz Maruch)
Date: 17 Dec 2000 10:58:45 -0800
Subject: Error evaluating numbers starting in zero
References: <3A3D07BD.85EFEB3D@wag.caltech.edu>
Message-ID: <91j2d5$9ai$1@panix3.panix.com>

In article <3A3D07BD.85EFEB3D at wag.caltech.edu>,
Richard P. Muller <rpm at wag.caltech.edu> wrote:
>
>I'm sure people already know about this, but I just got hit with a
>really screwy error.
>
>I have a python checkbook program that evaluates date strings, such as
>"01/31/99". Most dates are fine, but if I have a date such as
>"01/08/99", python runs into an error when it tries to eval("08").

Use mxDateTime
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

"Using LOCs as a measure of programmer productivity is like estimating the 
productivity of a welder by the amount of acetylene used."  


From tim.one at home.com  Wed Dec 20 02:18:35 2000
From: tim.one at home.com (Tim Peters)
Date: Wed, 20 Dec 2000 02:18:35 -0500
Subject: Walk through dictionary keys?
In-Reply-To: <apl04ts8h4vl52cdv9tcbva2n1gcabtuqd@4ax.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEMAIEAA.tim.one@home.com>

[Emile van Sebille]
> I'm not sure when or how long this has been in, but the cvs
> version of dict's has a popitem method that looks like it does
> what you want.
>
> d = {1:11,2:22,3:33,4:44}
> for i in xrange(len(d)):
>      print d.popitem()

[Tim Roberts]
> But is popitem destructive?  That is, when that loop is complete, is
> d empty?

Yes, d is empty.  It's not like Perl's "each" that way (which is what the
original poster asked about); OTOH, .popitem() is well-defined if d is
mutated inside the loop.

The original motivation for .popitem() was in the use of dicts to implement
sets.  Many set algorithms are of the form:

   while myset:   # would "not myset.empty()" really be clearer?!
       e = myset.pick_some_element()
       myset.remove(e)
       # deal with e, possibly adding more stuff to myset

There is no efficient way to implement the "pick an element and remove it"
operation in Python 2.0 if the set is represented as a dict.  .popitem()
provides one.

I expect it will also be handy for iterating over huge dicts not used to
represent sets; it's common enough for the last operation on a dict to be

    stuff = dict.items()
    for k, v in stuff:
        yadda

.popitem() can be used instead if dict is dead at this point, and so avoid
the memory burden of materializing a full list of (key, value) pairs.

.popitem()-is-the-right-answer-to-questions-that-weren't-
    asked-ly y'rs  - tim




From wolfson at midway.uchicago.edu  Thu Dec  7 01:40:01 2000
From: wolfson at midway.uchicago.edu (Ben Wolfson)
Date: Thu, 07 Dec 2000 06:40:01 GMT
Subject: OT: _Fifth Business_ (was Re: Python Origins)
References: <90k3pq$50p$1@nnrp1.deja.com> <3A2EE2AF.AD8A51E2@caltech.edu> <slrn92tu7t.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com> <90n1b0$4jr$1@panix3.panix.com>
Message-ID: <5hGX5.68$E3.1197@uchinews>

In article <90n1b0$4jr$1 at panix3.panix.com>, Aahz Maruch <aahz at panix.com> wrote:

>What's _Fifth Business_?  (I did a quick search on Google and Amazon and
>didn't find anything related to you.)

_Fifth Business_ is the first book in the Deptford trilogy by Robertson
Davies, a Canadian and a man with a very impressive beard (the other
books being _Manticore_ and _World of Wonders_).

-- 
BTR  |  "I see you're wearing a white undershirt, a blue shirt, and a dark tie.
         From this I deduce that you are twenty."  
        "French.  Close." -- I and Andy Hagan


From richard at NOstarfighterSPAM.freeuk.com  Sat Dec 16 13:12:20 2000
From: richard at NOstarfighterSPAM.freeuk.com (Richard)
Date: Sat, 16 Dec 2000 18:12:20 -0000
Subject: __del__() in a Tkinter app
References: <MlI_5.53768$eT4.4433606@nnrp3.clara.net> <ZKJ_5.110$u17.27823@newsb.telia.net>
Message-ID: <ojO_5.60536$R77.4608980@nnrp4.clara.net>

"Fredrik Lundh" <fredrik at effbot.org> wrote in message
news:ZKJ_5.110$u17.27823 at newsb.telia.net...
> "Richard" wrote:
> > I have tried adding a __del__() method to a Tkinter application so it
can
> > remove temporary files when the program exits.
>
> Don't ever do things like that in a __del__ method.   Python isn't
> C++, and doesn't make any guarantees that __del__ methods will
> ever be called.
>
> To properly shut down a Tkinter application, bind the "<Destroy>"
> event on the root/toplevel window.
>
> You can also overload the destroy method:
>
>     class myGUI(Tk):
>
>         def __init__(self, ...):
>             Tk.__init__(self)
>             # not really needed in 2.0 and later, but it doesn't hurt
>             self.protocol("WM_DELETE_WINDOW", self.destroy)
>
>         def destroy(self):
>             Tk.destroy(self)
>             ... handle shutdown
>
> </F>
>
>

Thanks Fredrik, I've used your suggestion to handle shutdown in my app,
without using a __del__() method.

regards,
Richard





From reichowa at tcfreenet.org  Wed Dec 13 12:15:50 2000
From: reichowa at tcfreenet.org (Aaron Jon Reichow)
Date: Wed, 13 Dec 2000 11:15:50 -0600
Subject: Smalltalk and Python
In-Reply-To: <917r2b$4n7$1@news.rz.uni-karlsruhe.de>
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de>
Message-ID: <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org>

On 13 Dec 2000, Bjoern Giesler wrote:

> : Would it perhaps not be better to use Ruby ?
> 
> Not for me, for the following reasons:
> a) I don't know Ruby well enough
> b) Ruby is "even more" unknown in The Real World(tm) than Python
> c) I think Ruby code looks ugly (what with @, #, $, no 'return' etc.)
> d) I don't see any of the features Ruby has over Python as 'killer'
>    features.

Proper blocks rather than awkward lambda functions, for starters.

Aaron Reichow                   :: "In essence, Smalltalk is a programming
Twin Ports ACM Vice President   ::  language focused on human beings rather
http://www.d.umn.edu/~reic0024/ ::  than the computer." - Alan Knight





From qneill at my-deja.com  Wed Dec 13 16:56:55 2000
From: qneill at my-deja.com (qneill at my-deja.com)
Date: Wed, 13 Dec 2000 21:56:55 GMT
Subject: smtplib
References: <918lo8$6pg$1@nnrp1.deja.com>
Message-ID: <918rb3$buq$1@nnrp1.deja.com>

In article <918lo8$6pg$1 at nnrp1.deja.com>,
  Brad <blosh at tsbrass.com> wrote:
> anyone know what this means...I am running Python 1.6 on HPUX10.20
>
> smtplib.SMTPHeloError:
>   (504, 'DNS verification of sending machine failed: no
>    mail will be accepted')


Most mail transfer agents (e.g. sendmail) require DNS verification
of hosts requesting services.  The remote MTA is doing a reverse DNS
lookup on your IP address as a way of finding out if you are a
legitimate client.

If you have a dynamic IP or are behind a firewall then you might have
problems with a remote system doing a reverse lookup on your machine.
Try another mailserver, or get reverse DNS working.

For more background info, try:
    http://linux.com/howto/DNS-HOWTO-4.html#ss4.3
    http://linux.com/howto/Mail-Administrator-HOWTO-3.html



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


From thomas at xs4all.net  Sat Dec 23 19:06:38 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Sun, 24 Dec 2000 01:06:38 +0100
Subject: A Suggestion for Python Colon Syntax
In-Reply-To: <922olo$e43$1@nnrp1.deja.com>; from ullrich@math.okstate.edu on Sat, Dec 23, 2000 at 05:50:49PM +0000
References: <mailman.977446446.16022.python-list@python.org> <922olo$e43$1@nnrp1.deja.com>
Message-ID: <20001224010638.J5515@xs4all.nl>

On Sat, Dec 23, 2000 at 05:50:49PM +0000, David C.Ullrich wrote:
> In article <mailman.977446446.16022.python-list at python.org>,
>   "Tim Peters" <tim.one at home.com> wrote:
> 
> > [...] (ditto the whole "while 1" thread
> > that's been going on virtually non-stop for a decade).

> and we'll never be certain whether it was an intentional
> joke or not. Keen.

No, I'm certain Tim did not make that horrible, horrible pun by accident. No
way could he inflict such amounts of nausea and pain by accident. Just not
possible :)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From owen_nospam at astro.omitme.washington.edu  Wed Dec 13 14:31:59 2000
From: owen_nospam at astro.omitme.washington.edu (Russell E. Owen)
Date: Wed, 13 Dec 2000 11:31:59 -0800
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <3a36c85e.4409136@News.CIS.DFN.DE>
Message-ID: <owen_nospam-1312001131590001@cs204-90.fsmodem.washington.edu>

In article <3a36c85e.4409136 at News.CIS.DFN.DE>, costas at springmail.com (Costas Menico) wrote:

><snip>..........
>>
>>-- Russell
>>
>>Having said that Python is not really a great substitute for Smalltalk, 
>>I will admit that I am working on a reasonable scale GUI project in 
>>Python because even though I adore ST I have doubts about ST's long-term 
>>viability (cincom seems to be back to ParcPlace's old tricks on insane 
>>pricing) and it'd be an incredibly hard sell to my management. I may be 
>>setting up for failure, due to Python's mediocre GUI, but I'm having a 
>>go at it.
>
>Cincom is not the only Smalltalk out there. What is wrong with the
>other lower cost ones?

I need something that is:
-- reasonably well supported
- likely to be around for awhile
- runs on Mac, unix and possibly PC
- can write GUI code with reasonably native look and feel

I would also prefer a language that is fairly clean and not evolving very quickly (which I dearly hope Squeak eventually becomes).

I don't think there is a Smalltalk that fits these requirements other than Cincom, is there?
- Squeak runs on everything but its GUI stuff is slow and uses a rooted main window, making it pretty hopeless for native feel, at least on Mac
- IBM and Dolphin are mature but only runs on PC and (IBM) unix
- I tried QKS years ago but got tired of it always being in beta, always promises for a real release some year.

Also, as I said, it'd be a tough fight. The conservatism in astronomy is astonishing. If it's not C or Perl then it's going to be a struggle to put it through. We're already using Python and selling it as a "better Perl", but for me to push Smalltalk there has to be an obvious first-class implementation that we can afford and will obviously do what we need. Cincom is too expensive, and given their pricing policies I expect it to continue to be a language with limited public knowledge and acceptance.

-- Russell


From shindich at my-deja.com  Tue Dec  5 01:53:56 2000
From: shindich at my-deja.com (shindich at my-deja.com)
Date: Tue, 05 Dec 2000 06:53:56 GMT
Subject: ANN.: Weak Reference Extension Module has been converted to "C"; Beta 2 is out
References: <8usorp$3i3$1@nnrp1.deja.com> <mtaeb1f4q3.fsf@astron.berkeley.edu> <8uvjmj$dt8$1@nnrp1.deja.com> <mtlmuk9hyc.fsf@astron.berkeley.edu> <B63C5A9E.172D7%graham73@telocity.com> <3A176E0F.39BF350B@felinity.com> <KaNR5.5550$QH2.654863@newsb.telia.net> <8v9dl9$si6$1@nnrp1.deja.com>
Message-ID: <90i3e3$f8t$1@nnrp1.deja.com>

I am happy to announce that Weak Reference Extension project has been
fully converted to "C". Weak Reference extension is now completely
CPython compatible.
Check out Beta 2 at Source Forge
(http://pyweakreference.sourceforge.net/ ;
http://sourceforge.net/projects/pyweakreference/)

Enjoy!

Alex Shindich
Visit http://www.shindich.com/


Sent via Deja.com http://www.deja.com/
Before you buy.


From dsavitsk at my-deja.com  Thu Dec 14 12:38:22 2000
From: dsavitsk at my-deja.com (dsavitsk at my-deja.com)
Date: Thu, 14 Dec 2000 17:38:22 GMT
Subject: Rummaging around inside an Access DB
References: <c5es2ts2l66bhg8qb7ne28cj0qa3gi9595@4ax.com>
Message-ID: <91b0id$3dh$1@nnrp1.deja.com>

> (I can't find a help file on the object model
> anywhere!!)

If you are going to use DAO, then there is a book from O'Reilly
called "DAO Object Model, The Definitive Reference" that is probably
worth having (although at first glance i don't see anything about
Reports in it.)

you might also try ADO. Something like ...

>>> import win32com.client
>>> conn = win32com.client.Dispatch(r'ADODB.Connection')
>>> DSN = 'PROVIDER=Microsoft.Jet.OLEDB.4.0;DATA SOURCE=C:/MyDB.mdb;'
>>> conn.Open(DSN)
>>> oCat = win32com.client.Dispatch(r'ADOX.Catalog')
>>> oCat.ActiveConnection = conn
>>> oRep = oCat.Reports
>>> for x in oRep:
...     if x.Type == 'REPORT':
...         print x.Name

... might work (by the way, this little bit of code was actually for
finding tables so i just changed 'TABLES' to 'REPORTS' etc. I don't
have any idea if it really works for reports, but the point is that
ADOX.Catalog might be a good place to poke around. (There is a slight
bit more info on access & ADO @ www.e-coli.net/pyado.html))


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


From grey at despair.rpglink.com  Sun Dec 24 13:53:43 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Sun, 24 Dec 2000 18:53:43 -0000
Subject: coding convention conversion
References: <mailman.977516469.1157.python-list@python.org> <920v4s$oqj$2@taliesin.netcom.net.uk> <slrn949c13.40l.grey@teleute.rpglink.com> <923jin$q2l$1@taliesin.netcom.net.uk> <slrn94bnam.i0h.grey@teleute.rpglink.com> <925em1$jcb$1@taliesin.netcom.net.uk>
Message-ID: <slrn94chhm.kje.grey@teleute.rpglink.com>

On Sun, 24 Dec 2000 18:18:41 +0000, Garry Knight <garryknight at bigfoot.com>
wrote:
>I meant longer as in the number of lines of code. I'd be very
>interested in seeing the Python version.

    Well, aside from that one construct I got stuck on I think it would have
been less.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From db3l at fitlinxx.com  Wed Dec 27 22:20:02 2000
From: db3l at fitlinxx.com (David Bolen)
Date: 27 Dec 2000 22:20:02 -0500
Subject: python telnet script for openvms host
References: <92ainp$jab$1@nnrp1.deja.com> <kuito7rrg7.fsf@lasipalatsi.fi> <92apa7$oac$1@nnrp1.deja.com> <D8D70FCFA8CC11C0.4BEE4AB5F2A3BCF7.EAD0DA106483FAE8@lp.airnews.net> <kuelyvrmqw.fsf@lasipalatsi.fi> <92d6oh$ghf$1@nnrp1.deja.com>
Message-ID: <ud7edz0yl.fsf@ctwd0143.fitlinxx.com>

Ennis Rosamond <ennisr at my-deja.com> writes:

> Looking at the actual client/host dialog with the aid of the telnetlib
> debug option, it appeared that perhaps the host was proposing 'suppress
> go ahead', but that acceptance of the option was not being returned by
> the client.  (The actual host message was '\377\373\003'.  I haven't yet
> found the interpretation of either 377 or 373.)

These values are octal, so \377 is the same as 0xFF (255 decimal),
which is the IAC (interpret as command) octet in telnet (which starts
all special sequences, such as options).

\373 is the same as 0xFB (251) which is WILL, and yes, option 3 is
suppress go ahead.

> To fix the script, I added a write call to the host with the message
> '\255\253\003', meaning 'do suppress go ahead', following the login

Hmm, except that's not what that means - if they were decimal values
you'd be ok, but they're octal - you should have to use '\377\375\003'
to send a DO SGA message.

However, that seems to tickle another problem - the telnetlib write()
method will automatically quote (double) any IAC character.  So it
doesn't appear that you can use tn.write() in your code to answer any
options.  You'd have to go through directly to tn.sock.send() (or use
the nicer get_socket() method to get the socket).

The fact that it seems to work may indicate that the underlying
problem is more subtle and it's just the fact that you're generating
data at that point in the process that is cleaning something up.

Of course, at some level it may not matter if it now does what you
want :-)

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From jerry_spicklemire at my-deja.com  Fri Dec 22 13:05:13 2000
From: jerry_spicklemire at my-deja.com (jerry_spicklemire at my-deja.com)
Date: Fri, 22 Dec 2000 18:05:13 GMT
Subject: Wanted: Python White Papers
References: <91oj8v$i0r$1@nnrp1.deja.com> <91tmkq$lru$1@nnrp1.deja.com> <91uj5j$dg4$1@nnrp1.deja.com>
Message-ID: <92054k$j6c$1@nnrp1.deja.com>

  Noah Spurrier replied:
> Thanks! This is great. It makes good ammo
> for my defense of Python.

If you liked those, check out:

 http://www.cgl.ucsf.edu/home/sparky/

 http://sandbox.xerox.com/webdav/

 http://www.uk.research.att.com/omniORB/omniORBpy/

 http://amber.intel.com/scripts-jobs/uRequisition.asp?Posting=23615

 http://amber.intel.com/scripts-jobs/uRequisition.asp?Posting=24124

Later,
Jerry S.


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


From dsh8290 at rit.edu  Sat Dec  9 01:21:39 2000
From: dsh8290 at rit.edu (D-Man)
Date: Sat, 9 Dec 2000 01:21:39 -0500
Subject: Python wrapper for C++ STL?
In-Reply-To: <3A31A05C.6A853969@erols.com>; from edcjones@erols.com on Fri, Dec 08, 2000 at 22:00:44 -0500
References: <3A31760E.D157EB54@erols.com> <C0gY5.126062$U46.4232859@news1.sttls1.wa.home.com> <90s3fk$j8p$1@nnrp1.deja.com> <3A31A05C.6A853969@erols.com>
Message-ID: <20001209012139.E31063@westsidecnc.rh.rit.edu>

Python already has it :

dict = { } # make up a dictionary
keys = dict.keys()  # get the keys
keys.sort() # sort the keys , too bad this returns None, it would be useful as
an expression

for key in keys :
	print dict[ key ]


:-)
-D

On Fri, 08 Dec 2000 22:00:44 Edward C. Jones wrote:
 | shindich at my-deja.com wrote:
 | 
 | > With all due respect, what would be wrong with a definition like
 | > std::map<PyObject*, PyObject*>? Obviously there is no type safety here.
 | > But Python is not type safe.
 | 
 | This seems correct. It would give me a "sorted associative container", a
 | dictionary I could loop through.
 | 
 | Ed Jones
 | 
 | 
 | -- 
 | http://www.python.org/mailman/listinfo/python-list
 | 




From Daniel.Kinnaer at AdValvas.be  Mon Dec 18 14:18:26 2000
From: Daniel.Kinnaer at AdValvas.be (Daniel)
Date: Mon, 18 Dec 2000 19:18:26 GMT
Subject: Timers in Python?
References: <3a3e42ef.16144323@news.skynet.be> <91li8l$33h$1@reader1.imaginet.fr>
Message-ID: <3a3e61cf.24049354@news.skynet.be>

On Mon, 18 Dec 2000 18:46:13 +0100, "Gilles Lenfant"
<glenfant at equod.com.nospam> wrote:

>Have a look to "threading" package and time.sleep(...) function.
>Seems that's all you need!
>Important, use threading.Lock() objects if threads may use not sharable
>resources.
>Warning, binary distributions of Python/Unix are not always compiles with
>threads options.
>You may need to compile Python from sources if required.
>

Dear Gilles,
Thanks for the reply.  It seems I ran into trouble when trying the
above code.  Python replied me the following :

timers.py :

import threading, time

Mylock = threading.Lock()

class TimeScheduled(threading.Thread):
    def __init__(self, function, delay):
        self.function = function
        self.delay = delay

    def run(self):
        self.alive = 1
        while self.alive:
            Mylock.acquire()
            # optional (depends whether unsharable resources used or
not)
            self.function()
            Mylock.release() # optional
            time.sleep(self.delay)

    def kill(self):
        self.alive = 0

def sayhello():
    # Very complex stuff
    print "Hello world"

# Every 2 minutes
scheduledHello = TimeScheduled(sayhello, 120)
scheduledHello.start() #this is line28

scheduledHello.kill()



Reply by Python : 

Traceback (innermost last):
  File "C:\Python20\AProject\timers\timers.py", line 28, in ?
    scheduledHello.start()
  File "c:\python20\lib\threading.py", line 351, in start
    assert self.__initialized, "Thread.__init__() not called"
AssertionError: Thread.__init__() not called

Is there something that can be done about it?  Thanks

Daniel


From jochen at pc1.uni-duesseldorf.de  Sun Dec 10 05:44:11 2000
From: jochen at pc1.uni-duesseldorf.de (Jochen =?iso-8859-1?q?K=FCpper?=)
Date: 10 Dec 2000 11:44:11 +0100
Subject: SA (was: Python Origins)
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu> <slrn92tu7t.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com> <tcmay-8BCFAB.23025408122000@news.got.net> <90uddk$31f@gap.cco.caltech.edu> <tcmay-4AACBE.16294509122000@news.got.net>
Message-ID: <udvgsszhc4.fsf_-_@box.home.de>

>>>>> Tim May wrote on Sat, 09 Dec 2000 16:29:45 -0800:

Tim> In article <90uddk$31f at gap.cco.caltech.edu>, Nathaniel Gray 
Tim> <nospam at caltech.edu> wrote:

>> He didn't ask you to do a search.

Tim> He has expressed happiness that he could find 1200 references, many of 
Tim> them applicable, to Python and simulated annelaing by using a better 
Tim> search engine.

Hmm, probably my English really is that bad???

I wasn't exactly happy. After following your hint I found a few basic,
simple, nonetheless working implementations of SA in python.
I posted URLs of the code I found and asked again whether anybody on
this list had a more elaborate and sophisticated version...

As I don't get any replies, I am wondering whether people are not
doing SA in python or "simply" not willing to share their code?

Number one could mean people are using C-modules from python for that
kind of stuff, and I thought about that from the beginning. Number two
would be very strange (and let me sit here uncomfortable) regarding
pythons origins and scientific spirit.

As this is my first python project, I am going to implement it in
python completely (for the beginning). As such I will write my own
python-implementation of SA (when I get time to work on it again later
this month).

Tim> I didn't see _you_ providing any help, either in that thread or
Tim> in this "Python origins" thread.
[...]  
Tim> You remind me of Alan Dershowitz.

Getting personal shouldn't be done in the group but by private mail.

Greetings,
Jochen
-- 
Einigkeit und Recht und Freiheit                http://www.Jochen-Kuepper.de
    Libert?, ?galit?, Fraternit?                GnuPG key: 44BCCD8E
        Sex, drugs and rock-n-roll


From grey at despair.rpglink.com  Sat Dec 30 20:12:19 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Sun, 31 Dec 2000 01:12:19 -0000
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <slrn94shbe.150.grey@teleute.rpglink.com> <X8s36.2731$Qb7.305339@newsb.telia.net> <slrn94smds.150.grey@teleute.rpglink.com> <FOt36.17$vd1.2224@e420r-atl2.usenetserver.com>
Message-ID: <slrn94t1vj.150.grey@teleute.rpglink.com>

On Sat, 30 Dec 2000 23:00:53 GMT, Kragen Sitaker <kragen at dnaco.net> wrote:
>Hahaha!  You just plonked Fredrik Lundh in comp.lang.python.  He was
>being helpful and polite, and you plonked him because you disagreed
>with him.  I can guess how much help you'll get next time you have a
>Python question.  :)

    I can tell you I'll get far better without him than with him when he is
giving advice that runs contrary to about 10+ years of programming that I've
been taught from various different sources.  I don't care if he is God
himself, advocating not initing a variable is stupid, plain and simple.  He
wanted to argue the point for nothing more than argument's sake, so be it.  I
don't need his advice any further.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From aleaxit at yahoo.com  Mon Dec 18 12:26:35 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 18 Dec 2000 18:26:35 +0100
Subject: Singleton vs Proxies DP (was Re: Solution: Direct access to  Printer   I/O lines)
References: <917k9b$9i9$1@nnrp1.deja.com> <slrn93hrjn.2vp.ssthapa@localhost.localdomain> <91cnp3$g59$1@nnrp1.deja.com> <91d23f0jk7@news1.newsguy.com> <91gfnp$7r4$1@animus.fel.iae.nl> <91gkh801e49@news2.newsguy.com> <3A3CB83D.45C8775E@darwin.in-berlin.de> <91iu95013ta@news2.newsguy.com> <3A3D1D49.1A453B51@darwin.in-berlin.de> <91kmfd020hp@news1.newsguy.com> <3A3DF9AA.74552346@darwin.in-berlin.de>
Message-ID: <91lhdb0mb@news1.newsguy.com>

"Dinu C. Gherman" <gherman at darwin.in-berlin.de> wrote in message
news:3A3DF9AA.74552346 at darwin.in-berlin.de...
    [snip]
> Actually, I'm pretty sure there is already something like a
> singleton-like pattern with more of the properties you like.

Of course there is -- I've been saying all along that I
first met this design pattern (whose name I forget!) in
C++ terms in the C++ Report (and in COM it's rather old
hat...:-).

> Maybe you should check Linda Rising's "Pattern Almanac" or
> similar sources?
>
>   http://www.amazon.com/exec/obidos/ASIN/0201615673

That might help me remember the name, yes... or else, any
cumulative table of contents of C++ Report (or, if I
could just put some order in my old C++ Report issues!-).


Alex





From fredrik at pythonware.com  Thu Dec 21 06:44:27 2000
From: fredrik at pythonware.com (fredrik at pythonware.com)
Date: Thu, 21 Dec 2000 11:44:27 GMT
Subject: mail processing module?
References: <91rcig$rgu$1@nnrp1.deja.com>
Message-ID: <91sqer$tsn$1@nnrp1.deja.com>

Preston Landers improves the signal/noise ratio:
> I'm trying to find out if there's any interest in a
> kitchen-sink Python module for mail handling. /.../
> It uses the standard Python smtp/pop/imap/MIME modules,
> just basically ties them all together.

I sure wouldn't mind seeing a "maillib.py" module in the
next Python release (similar to "urllib", which also pro-
vides an easy-to-use layer on top of httplib/ftplib etc).

so if you decide to release it, don't forget to post a
copy to the patch manager.

</F>


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


From sbrunning at bigfoot.com  Mon Dec  4 12:11:10 2000
From: sbrunning at bigfoot.com (Simon B.)
Date: Mon, 04 Dec 2000 17:11:10 GMT
Subject: Programming with Python
References: <90fq39$jgq$1@nnrp1.deja.com> <9iOW5.55847$L26.16633720@typhoon.southeast.rr.com>
Message-ID: <90gj77$7a0$1@nnrp1.deja.com>

In article <9iOW5.55847$L26.16633720 at typhoon.southeast.rr.com>,
  "Don Tuttle" <tuttledon at hotmail.com> wrote:
> >
> > Any good?
>
> IMO, this is the only bad Python book on the market.  Don't waste your
> money.

Thanks - you may have just saved me thirty quid!

--
Simon Brunning
Are you going to come quietly, or do I have to use earplugs? - from The
Goon Show


Sent via Deja.com http://www.deja.com/
Before you buy.


From pyp at gmx.net  Tue Dec 12 12:16:00 2000
From: pyp at gmx.net (Miike Mueller)
Date: Tue, 12 Dec 2000 18:16:00 +0100
Subject: passing list from C to Python
References: <9154s2$2sb@micky.ibh-dd.de> <9158la030jp@news2.newsguy.com>
Message-ID: <915n2p$asi$03$1@news.t-online.com>

"Alex Martelli" <aleaxit at yahoo.com> schrieb im Newsbeitrag
news:9158la030jp at news2.newsguy.com...

>
> PyListObject *newList = PyObject_CallMethod(module_object,
>     "pythonModul", "(O)", list);
>
> Note that PyObject_CallMethod, much like Py_BuildValue, takes
> a format-string describing the following arguments, and builds
> the tuple according to it.  I think you do need the parenthesis
> in the case of a format-string describing just one object, to
> make sure it's still wrapped into a (1-element) tuple, which
> is how Python callables expect their arguments (but maybe the
> PyObject_CallMethod is smart enough to do it implicitly; the
> docs don't make it clear and I haven't checked the sources).
>

O works wihtout (O)

I managed to pass the list in. The calculations in Python works as expected.
I can get a list back from Python (checked with PyListCheck()).
The only problem remaining to convert the list back to the C-args.
I tried PyArgs_ParseTuple() but I could not get to work, the C-args just
stay the same as before.
PyListGetItem works but I have 35 arguments and writing one
PyArgs_ParseTuple() seems to be more elegant than 35 PyListGetGetItem and 35
PyFloat_AsDouble(). (Don't know how to loop over them.)


> I think that it would be even simpler if the Python code you
> call just modified its list-argument, but it's a minor issue.
I think this would the the simplest way.
I tried this by passing the list in as an argument and changing the list in
place but this has no effect an the *args in C. I am not sure if understand
the concept of pointers right here.

Mike





From SBrunning at trisystems.co.uk  Fri Dec  8 11:06:03 2000
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Fri, 8 Dec 2000 16:06:03 -0000 
Subject: Refactoring browser
Message-ID: <31575A892FF6D1118F5800600846864D5B159A@intrepid>

> From:	Stefan Boiadjiev [SMTP:stefanb at mail.vega.bg]
> Do in python world exist something similar to Smalltalk Refactoring
> browser ?
 
<http://bicyclerepair.sourceforge.net/>.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk





-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From nospam at no.spam  Mon Dec 18 17:32:23 2000
From: nospam at no.spam (Carsten Geckeler)
Date: Mon, 18 Dec 2000 23:32:23 +0100 (MET)
Subject: while true: !!!
In-Reply-To: <slrn93snid.tvg.grey@teleute.rpglink.com>
Message-ID: <Pine.LNX.4.10.10012182326490.1345-100000@erde.kosmos.all>

On Mon, 18 Dec 2000, Steve Lamb wrote:

> On Mon, 18 Dec 2000 13:36:27 +0100 (CET), Carsten Geckeler <nospam at no.spam>
> wrote:
> >Anyway it's a very common idiom.  But the point is that using true and false
> >would make the code more readable in cases where it's not clear, that the
> >value is used as a boolean value later.  E.g:
> 
> [Example of if conditions snipped]
> 
>     You're mixing examples.  I doubt that anyone has a problem with storing a
> value in an appropriately named variable.  However, how does that address
> having true equate to false?  That would be the requirement of having "while
> true:" and then terminating the loop with assigning "true = 0" thus making
> that consition false.  Meanwhile the very valid break would go unused for the
> exact purpose that it was designed for, breaking out of a loop.

Sorry, but that was not my point.  Of course, defining true = 0 would be a
bad way of ending the loop.  In case you use true and false variable,
those should be defined to 1 and 0 and never be changed.  The break
statement is definetly the best way to terminate a loop.  I just said that
`while true' would be more readable and clear than `while 1'.  By the way,
explain a newbie that -1 is true. ;)

Cheers, Carsten
-- 
Carsten Geckeler:  carsten dot geckeler at gmx dot de
To get proper email-address replace `dot' and `at' by the corresponding symbols.




From ljz at asfast.com  Sat Dec 16 11:02:40 2000
From: ljz at asfast.com (Lloyd Zusman)
Date: 16 Dec 2000 11:02:40 -0500
Subject: Alarms in threads - how to get/mimic?
References: <Pine.LNX.4.21.0012160248520.7483-100000@traal.eecs.berkeley.edu>
Message-ID: <ltbsucbbhb.fsf@asfast.com>

Luca de Alfaro <dealfaro at eecs.berkeley.edu> writes:

> I am trying to write a multi-threaded program to read web pages.  The
> program spawns (say) 10 threads, assigning to each thread a file to
> read with the ftp or http protocol; each thread reads it, and returns
> the info.
> 
> The problem is that the threads sometimes hang for a very long time
> waiting for a connection (httplib seems to have very long timeouts??),
> whereas instead I don't care for the answer if it comes later than
> (say) a few seconds.
> 
> [ ... ]

Perhaps Timothy O'Malley's `timeoutsocket' class will help you.  It
allows you to set timeouts for selected socket operations, such as the
httplib connect calls you're referring to.  You can get it here ...

  http://www.timo-tasi.org/python/

Here's an excerpt from Timothy O'Malley's documentation:

  """Timeout Socket

  This module enables a timeout mechanism on all TCP connections.  It
  does this by inserting a shim on top of the socket module.  After
  this module has been imported, all socket creation goes through this
  shim.  As a result, every TCP connection will support a timeout.

  [ ... ]

  The timeout applies to the socket functions that normally block on
  execution:  read, write, connect, and accept.  If any of these 
  operations exceeds the specified timeout, the exception Timeout
  will be raised.

  [ ... ]

  """

Using this module, you can cause each httplib connection statement
which times out to throw an exception, and each thread can then
intelligently respond to its own exception.

I've been using this module for a similar purpose, and I'm quite
happy with it.


And by the way, I want to take this opportunity to express my public
thanks and congratulations to Timothy O'Malley for having created this
much-needed module.


Question to the Python community: the lack of a timeout mechanism
seems to be a glaring hole within the standard `socket' module's API.
Are there any plans to add this functionality to this module?  If so,
it should then be quite straightforward to encapsulate timeout
capabilities within the all the standard socket-based libraries such
as `httplib', `pop3lib', `imaplib', `smtplib', etc.


> [ ... ]
>
> Luca de Alfaro


-- 
 Lloyd Zusman
 ljz at asfast.com


From chris at araidesign.com  Thu Dec  7 20:59:53 2000
From: chris at araidesign.com (Chris Arai)
Date: Thu, 07 Dec 2000 17:59:53 -0800
Subject: How to use BLT2.4 and Tcl/tk8.3 under Windows9X/NT ?
References: <39F6BBCD.C25D4FEA@mailux.berl.tek.com> <3a0351b4.7467539@news-server.socal.rr.com>
Message-ID: <3A304098.40157C82@araidesign.com>

Hello,

I am having the same problem trying to get Pmw.blt to work.  The all.py
demo works for the Pmw part, but gives me the "BLT Package not installed"
message when I tried to use Blt.Graph.

My application that I wrote in Python 1.6 on linux gives me the
following:

Traceback (most recent call last):
  File "polar_ed.py", line 332, in ?
    tl.append(editor(root))
  File "polar_ed.py", line 79, in __init__
    self.grf= scoop.polgrf(mainpane.pane('plot'))
  File "scoop.py", line 394, in polgrf
    graph=Pmw.Blt.Graph(parent)
  File "d:\program files\python20\Pmw\Pmw_0_8_4\lib\PmwBlt.py", line 260,
in __init__
    Tkinter.Widget.__init__(self, master, _graphCommand, cnf, kw)
  File "d:\program files\python20\lib\lib-tk\Tkinter.py", line 1758, in
__init__

    self.tk.call(
TclError: invalid command name "::blt::graph"


Has anyone provided the answer to these problems yet?

Thanks in advance,

Chris


howard at eegsoftware.com wrote:

> Following up on myself as it were.....
>
> I was unable to get the blt2.4u for 8.3 bltwish shell to run until I
> copied the library directories to be under Python20\Libs.  Bltwish
> then was able to run its own demos (graph1.tcl).
>
> The following program :
>    from Tkinter import *
>    import Pmw
>
>    root=Tk()
>    graph=Pmw.Blt.Graph(root)
>
> works in 1.5.2 but fails as follows on 2.0:
>
> Traceback (most recent call last):
>   File "test.py", line 5, in ?
>     graph=Pmw.Blt.Graph(root)
>   File "c:\python20\Pmw\Pmw_0_8_4\lib\PmwBlt.py", line 260, in
> __init__
>     Tkinter.Widget.__init__(self, master, _graphCommand, cnf, kw)
>   File "c:\python20\lib\lib-tk\Tkinter.py", line 1758, in __init__
>     self.tk.call(
> TclError: invalid command name "::blt::graph"
>
> Significantly, the Dos window hangs after this and must be terminated
> forcefully (forceably ?).  It seems to me that this is a symptom of
> mismatched DLL's somewhere.
>
> Still hoping for a good idea from someone......
> Howard Lightstone
> EEGSoftware
> howard at eegsoftware.com

--
Chris Arai
Arai Design
chris at araidesign.com
http://www.araidesign.com
707 431 1299
FAX 431 1599

-------------- next part --------------
A non-text attachment was scrubbed...
Name: chris.vcf
Type: text/x-vcard
Size: 224 bytes
Desc: Card for Chris Arai
URL: <http://mail.python.org/pipermail/python-list/attachments/20001207/9d0464a2/attachment.vcf>

From aleaxit at yahoo.com  Tue Dec 19 07:59:17 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 19 Dec 2000 13:59:17 +0100
Subject: How to rewrite this as list comprehension?
References: <91nj2q$s3p$1@buty.wanadoo.nl>
Message-ID: <91nm3v0150s@news2.newsguy.com>

"Serge Beaumont" <serge at sergebeaumont.com> wrote in message
news:91nj2q$s3p$1 at buty.wanadoo.nl...
> Hello,
>
> I'm trying to get my head around comprehensions. How would I rewrite this
> method?
>
>   def deadThings(self, type = None):
>     deadThings = self.destroyed[:]
>     for thing in self.things:
>       if thing.destroyed:
>         if type:
>           if isinstance(thing, type):
>             deadThings.append(thing)
>         else:
>           deadThings.append(thing)
>     return deadThings

Writing this as a _single_ list comprehension will likely
put 'too much fat on the fire':

def deadThings1(self,type=None):
  return [ thing for thing in self.destroyed+self.things
      if thing in self.destroyed or thing.destroyed and ... ]

this would be a seriously unnatural approach -- if you
want to return the concatenation of an existing list with
some subset of another one, expressing it as a concatenation
(of the existing list with a comprehension) is a serious
readability win over expressing it as one huge comprehension:

def deadThings2(self,type=None):
  return self.destroyed + [ thing for thing in self.things
      if thing.destroyed and ... ]

(besides, the form used in deadThings1 would have
different semantics if repetitions are allowed in
the lists, AND would be _seriously_ slower too!).

So the only remaining problem is how to express the
'...' part.  My preference would be for a local
named function -- names sometimes clarify things
a lot...:

def deadThings3(self,type=None):
    if type:
        def isDestroyed(thing, type=type):
            return thing.destroyed and isinstance(thing,type)
    else:
        def isDestroyed(thing):
            return thing.destroyed
    return self.destroyed + [ thing for thing in self.things
        if isDestroyed(thing) ]

I'd rather test 'if type' once, choosing the appropriate
selection function, than test it every time through the
loop, given it's not changing inside the loop anyway; not
so much an issue of performance, but of clarity and
simplicity.  But if you disagree, it's easy to put the
test for type in a single function rather than use it
to choose the right function:

def deadThings4(self,type=None):
    def isDestroyed(thing, type=type):
        return thing.destroyed and (
            not type or isinstance(thing,type))
    return self.destroyed + [ thing for thing in self.things
        if isDestroyed(thing) ]

If you're allergic to local named function, you _can_
expand this form if 'isDestroyed' right in the list
comprehension, though I'd say this makes things a
bit more obscured (an epsilon of clarity may be gained
through the idiom 'if X if Y or Z' rather than
'if X and (Y or Z)', though that's debatable...):

def deadThings5(self,type=None):
    return self.destroyed + [ thing for thing in self.things
        if thing.destroyed if not type or isinstance(type,thing)]

Fanatics of conciseness at all costs may prefer this
last form (which CAN be considered a single-liner, if
your terminal has long-enough lines), but my personal
tastes favour form number 3, above, or some variation
emphasizing slightly different factoring:

def deadThings6(self,type=None):
    if type:
        def inType(thing, type=type):
            return isinstance(thing,type)
    else:
        def inType(thing):
            return 1
    return self.destroyed + [ thing for thing in self.things
        if thing.destroyed and inType(thing) ]

(which I find clearer, but that's just me), or:

def deadThings7(self,type=None):
    def inType(thing, type=type):
        return not type or isinstance(thing,type)
    return self.destroyed + [ thing for thing in self.things
        if thing.destroyed and inType(thing) ]

which, I suspect, would be the first form to 'flow off
the keyboard' for most of us:-).


Alex





From see at below  Fri Dec 22 22:43:37 2000
From: see at below (Paul Foley)
Date: 23 Dec 2000 16:43:37 +1300
Subject: Python interpreter in python?
References: <mailman.977335880.30954.python-list@python.org> <91vuan$l97@news.Informatik.Uni-Oldenburg.DE>
Message-ID: <m2y9x79506.fsf@mycroft.actrix.gen.nz>

On 22 Dec 2000 17:08:53 +0100, jnix  wrote:

> in what language the world might be interpreted ?

http://artists.mp3s.com/artist_song/234/234762.html

-- 
C/C++/Java/Perl/etc are for people who want to make things that work.
Common Lisp is for people who want to make things that don't break.
                                                           -- Erik Naggum
(setq reply-to
  (concatenate 'string "Paul Foley " "<mycroft" '(#\@) "actrix.gen.nz>"))


From boud at rempt.xs4all.nl  Sat Dec 16 12:40:17 2000
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 16 Dec 2000 17:40:17 GMT
Subject: Unicode and default encodings
Message-ID: <91g9e1$4t2$1@news1.xs4all.nl>

I'm currently trying to work with Unicode strings, and
I was kind of wondering where sys.setdefaultencoding has
gone to - it's still mentioned in Misc/unicode.txt, but
it isn't there anymore ;-).

Is it still possible to define a default encoding for all
the strings I use?

-- 

Boudewijn Rempt  | http://www.valdyas.org


From nospam at no.spam  Fri Dec 15 13:51:03 2000
From: nospam at no.spam (Carsten Geckeler)
Date: Fri, 15 Dec 2000 19:51:03 +0100 (MET)
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
In-Reply-To: <rOr_5.31120$x6.15241856@news2.rdc2.tx.home.com>
Message-ID: <Pine.LNX.4.10.10012151949450.1208-100000@erde.kosmos.all>

On Fri, 15 Dec 2000, Rainer Deyke wrote:

[snip]
> Yes, in-place modification should be supported.  I had something like this
> in mind:
> 
> class UserInt:
>   def __init__(self, value):
>     self.data = value
>   def __add__(self, other):
>     if isinstance(other, UserInt):
>       return self.__class__(self.data + other.data)
>     else:
>       return self.__class__(self.data + other)
>   __radd__ = __add__
>   def __iadd__(self, other):
>     if isinstance(other, UserInt):
>       self.data += other.data
>     else:
>       self.data += other

There is a `return self' missing at the end of the __iadd__ method.

Cheers, Carsten
-- 
Carsten Geckeler:  carsten dot geckeler at gmx dot de
To get proper email-address replace `dot' and `at' by the corresponding symbols.




From cerutti at together.net  Fri Dec 15 16:00:29 2000
From: cerutti at together.net (Neil Cerutti)
Date: 15 Dec 2000 21:00:29 GMT
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <tcmay-FCA22F.14073713122000@news.got.net> <918v7a02scn@news1.newsguy.com> <tcmay-04E80C.16181313122000@news.got.net> <91a1ed02a4j@news2.newsguy.com>
Message-ID: <slrn93l1tf.5l.cerutti@fiad06.norwich.edu>

Alex Martelli posted:
>What classes are 'abstract' (not actually instantiated as
>themselves, but only in terms of their subclasses) or 'concrete'
>(directly instantiable without subclassing) is part of what your
>O-O analysis is determining, and will depend on what exactly
>you're trying to do.  The root of the 'only abstract classes
>ever get subclassed' idea is that there is an inconsistency of
>levels-of-detail in your analysis if it allows subclassing of
>instantiable classes.  If your application-needs (or more
>generally, your domain-modeling needs) are such that you need to
>distinguish between, e.g., different kinds of 'vehicles' (cars
>vs trucks vs planes vs ...), then 'Vehicle' itself, the common
>superclass, cannot be instantiable in a logically consistent way
>-- it would break the idea that your concrete classes partition
>the universe of discourse, if it were.
>
>This is based on the idea of subclassing as asserting IS-A.
>
>As, say, a Plane IS-A Vehicle, it cannot be the case (if your
>concrete classes partition the universe of discourse) that a
>(generic) Vehicle is instantiable; else, you would have
>ambiguity in what concrete class to use to model an actual plane
>-- either a concrete Plane or a concrete generic Vehicle would
>satisfy all expressed constraints.  In practice, object-models
>which allow instantiation of a superclass (such as Vehicle) do
>so by adding a NON-expressed constraint about using the most
>'leaf' class an actual object can possibly be modeled as; but,
>that analysis-constraint is left non-explicitly-expressed in the
>object model... it CANNOT be expressed in it.  Whatever semantic
>constraints we attach to Vehicle will get inherited by Plane, so
>we cannot say *of Vehicle instances* 'and this is not
>specifically a Plane', else, bye-bye IS-A.

Luckily, the objects we model in software are generally simpler
and easier to define than a Vehicle class. In the real world,
things have "family relationships", (Wittgenstein?)--you can't
always make an easy determination about what class a thing
belongs to. In an ocean simulation, you might have a Fish class
and it would be useful for the Dolphin class to inherit, but it
IS-NOT a fish.

Perhaps it makes sense to instantiate a Vehicle if you believe in
Plato's idea of an ideal vehicle. All other vehicles will be
different or imperfect in some way, but the ideal could exist and
be instatiated--by God, er... by the programmer. ;-)

-- 
Neil Cerutti <cerutti at together.net>


From wolfson at midway.uchicago.edu  Sat Dec 23 15:51:44 2000
From: wolfson at midway.uchicago.edu (Ben Wolfson)
Date: Sat, 23 Dec 2000 20:51:44 GMT
Subject: coding convention conversion
References: <mailman.977516469.1157.python-list@python.org> <920v4s$oqj$2@taliesin.netcom.net.uk> <slrn949c13.40l.grey@teleute.rpglink.com>
Message-ID: <Af816.222$x3.2694@uchinews>

In article <slrn949c13.40l.grey at teleute.rpglink.com>,
Steve Lamb <morpheus at here.not.there> wrote:
>On Sat, 23 Dec 2000 01:29:00 +0000, Garry Knight <garryknight at bigfoot.com>
>wrote:
>>--------8X-cut here-----------------------------------------------
>>#!/usr/bin/perl
>>
>>while (<>) {
>>  if (/^\s*def\s+\w+/) {
>>    s/([A-Z])/_\l\1/g;
>>  }
>>  print;
>>}
>>--------8X-cut here-----------------------------------------------
>
>    Anyone else find it ironic that someone is solving a problem using Perl in
>a Python group when the comparable Python would be an excellent exercise to
>perform?  :)

Well, no:  now you have to change every invocation of the function.  
Also, wouldn't that give, if fed BarBaz, _bar_baz?  I thought the goal was
bar_baz.
-- 
Barnabas T. Rumjuggler
I say, an object can move whether or not an "external (male, masculine) 
force" chooses to act upon (rape) it!
 -- David Zacuto


From rjroy at takingcontrol.com  Sat Dec 16 14:32:48 2000
From: rjroy at takingcontrol.com (Robert Roy)
Date: Sat, 16 Dec 2000 19:32:48 GMT
Subject: Compiling PyQT-2.2
References: <TSN_5.8418$wX2.112805@typhoon.austin.rr.com>
Message-ID: <3a3bc07f.10013671@news1.on.sympatico.ca>

On Sat, 16 Dec 2000 17:45:23 GMT, "Marcus Urban"
<MURBAN at houston.rr.com> wrote:

>I've been trying to build PyQt for several days now and it continues to
>fail during the "make install" process. SIP compiles just fine. No
>complaints during ./configure or make/install. PyQt configures fine, too,
>with no complaints. The make process succeeds. But "make install" aborts.
>And yes, QTDIR is set correctly.
>
>
>/bin/sh ../mkinstalldirs /usr/local/lib/python2.0/site-packages
> /usr/bin/ginstall -c -m 644 ./qt.py /usr/local/lib/python2.0/site-packages/qt.py
>make  install-data-hook
>make[3]: Entering directory `/home/urban/py/PyQt-2.2/qt'
>(cd /tmp; PYTHONPATH=/usr/local/lib/python2.0/site-packages:/usr/local/lib/python2.0/site-packages python -c "import qt")
>Traceback (innermost last):
>  File "<string>", line 1, in ?
>  File "/usr/local/lib/python2.0/site-packages/qt.py", line 43, in ?
>    import libqtc
>ImportError: /usr/local/lib/libsip.so.3: undefined symbol: PyObject_Init
>make[3]: *** [install-data-hook] Error 1
>make[3]: Leaving directory `/home/urban/py/PyQt-2.2/qt'
>make[2]: *** [install-data-am] Error 2
>make[2]: Leaving directory `/home/urban/py/PyQt-2.2/qt'
>make[1]: *** [install-am] Error 2
>make[1]: Leaving directory `/home/urban/py/PyQt-2.2/qt'
>make: *** [install-recursive] Error 1


I had some installation problems too. My system is FreeBSD 4.2 with
KDE 1.4x. libqtcmodule.so was finding the wrong version of the
library. The following couple of paragraphs are extracted from an
email I sent to the maintainer of PyQt and describe what I had to do
to get things to work. Your problem sounds different than mine in that
the unresolved symbol you are trying to load is different. Yous might
be a sip problem. Re-run configure for sip and ensure that the output
displays the path to the QT libraries. You might also want to search
the mailing list archives: 

http://mats.gmd.de/mailman/listinfo/pykde 

for "undefined symbol", there is a problem identical to yours listed
there though the solution is not clear.

 http://www.mail-archive.com/pykde at mats.gmd.de/msg00541.html


This is how I fixed my problem:

At the root of this is that libqtcmodule loads libqt.so rather than
the version of the library against which it was built, libqt.so.2.2
(probably no advantage to being more specific). Strangely enough it
does specifically load libSM.so.6. I observed this first hand when I
played with the load paths. This is also consistent with what is in
the output of  'strings libqtcmodules.so'. 

My workaround was to create a symlink from /usr/X11R6/lib/libqt2.2.so
to my libqt.so.2.2.2 library and to change the libqtcmodule_la_LIBADD
string to use '-lqt2.2' instead of -lqt and remake. This is less than
ideal but it works. Ideally the linker would recognize which version
it is linking against and ask for the appropriate file like it does
for libSM. I don't know what prevents this from happening. Perhaps
there is some version value that has to be set in qt.  


Good Luck

Bob


From roy at panix.com  Tue Dec 26 08:55:51 2000
From: roy at panix.com (Roy Smith)
Date: Tue, 26 Dec 2000 08:55:51 -0500
Subject: One Python 2.1 idea
References: <mailman.977771949.17911.python-list@python.org> <9292tn$jka$1@nnrp1.deja.com> <929n9c013m3@news1.newsguy.com> <mailman.977835669.28993.python-list@python.org>
Message-ID: <roy-9332FF.08555126122000@news.panix.com>

Thomas Wouters <thomas at xs4all.net> wrote:
> You *can* speed it up, by doing 'x.foo = x.foo' somewhere early in 
> your call chain :) I experimented with this sometime during 2.0 
> development, but doing this for every call (by adjusting 
> instance_getattr2 to store bound methods in the instance dict after 
> creating them) proved only a marginal (and disputable) improvement, 
> and broke the pickle modules (because bound methods can't be 
> pickled.) And I'm not sure if the benifit is large enough to actually 
> build a framework of caching (which Python lacks entirely, currently) 
> and use it in the few places that might use it.

I do this on occasion when I have deeply nested data structures (not 
classes) and I make frequent use of a value or method burried deep in 
the chain.  Depending on how deep the chain gets, "frequent" might mean 
"more than once" :-)  For example, I might do:

x.port = x.page.frame.server.connection.input.socket.port

but I do it because I think it makes the code easier to read (and 
certainly easier to type!), not because I'm particularly worried about 
how efficient it is.


From egbert at bork.demon.nl  Thu Dec 28 09:06:37 2000
From: egbert at bork.demon.nl (Egbert Bouwman)
Date: Thu, 28 Dec 2000 15:06:37 +0100
Subject: MySQL and __init__
Message-ID: <20001228150636.A242@bork.demon.nl>

Creating an instance in a database-context means either
filling an instance with data from an existing table-row, 
or collecting data and adding that as a new row to a table, or both.
Is there any received wisdom or other advice how to do this ?
egbert 
-- 
Egbert Bouwman - Keizersgracht 197 II - 1016 DS  Amsterdam - 020 6257991
========================================================================



From franzja at anubis.network.com  Tue Dec  5 10:14:46 2000
From: franzja at anubis.network.com (Joseph A. Franzmeier)
Date: Tue, 05 Dec 2000 09:14:46 -0600
Subject: Programming with Python
References: <90fq39$jgq$1@nnrp1.deja.com> <9iOW5.55847$L26.16633720@typhoon.southeast.rr.com> <90gj77$7a0$1@nnrp1.deja.com> <3A2BFD12.7AB3A628@anubis.network.com> <3A2CC304.8C451CD6@eq.uc.pt>
Message-ID: <3A2D0666.22D6F4DF@anubis.network.com>

Yes Pedro,

You are correct I meant  "Python Essential Reference", by author Beazley.

---------

Pedro Vale Lima wrote:

> "Joseph A. Franzmeier" wrote:
>
> > The other to recommend is  "The Python Reference"  by  Guido Van
> > Rousson
> > the creater of Python.
>
> I never heard of this one. Are you talking of the "Library Reference"
> coming with the
> python docs, or are you talking about "Python Essential Reference" by
> Beazley the author
> of SWIG (by the way, this is the book I always carry around).
>
> The best book for learning is Programming Python, it's a bit dated but
> still you learn all from
> the basic to the expert. ORA is probably going to do a 2nd edition.
>
> regards
> pedro

--

|    StorageTek    |
| cout << C_ya << later << Joe;


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20001205/067f5e34/attachment.html>

From mhuening at zedat.fu-berlin.de  Wed Dec  6 07:58:44 2000
From: mhuening at zedat.fu-berlin.de (Matthias Huening)
Date: Wed, 6 Dec 2000 13:58:44 +0100
Subject: Changing frames with a python script
References: <976097884.20994.0.pluto.c30bfd5c@news.demon.nl>
Message-ID: <90ld5l$118ov$1@fu-berlin.de>

Paul van Erk <paul at vanerk.demon.nl> wrote:

> My first solution was that the script would make a new html document,
which
> was also divided in frames. But then I get a document with three frames
> (logical), because it would publish the html document only in the main
> frame.
>

Well, it's not a Python-problem, but your first solution should be just
fine. You only have to insert a 'target' in the HTML-page which calls your
script:
<FORM ACTION="..." METHOD="..." TARGET="_top">

Matthias




From aleaxit at yahoo.com  Tue Dec  5 04:04:22 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 5 Dec 2000 10:04:22 +0100
Subject: Calculating factorial
References: <3a2a9496.647995@uutiset.nic.fi> <3a2c8c2a_3@corp.newsfeeds.com>
Message-ID: <90ib2e0216k@news2.newsguy.com>

"Joshua Muskovitz" <josh at open.com> wrote in message
news:3a2c8c2a_3 at corp.newsfeeds.com...
> What about using xrange() instead of range() to support *really* large
> numbers?

Of marginal usefulness in the context of computing factorials.

The memory space taken by a number X grows, roughly, with the
logarithm of X.  From Stirling's formula, we know that
    log(N!) about= N log(N) - N
for large N; for N large enough, this means that N! takes up
memory space growing as O(N logN).

Since range(N) takes up memory space growing as O(N), for a
large enough N this memory consumption becomes irrelevant
compared to the memory which N! itself takes up.

There _may_ be a value or two of N such that you have just
enough memory to compute N! if you carefully husband every
little bit, which is why the usefulness of xrange is
"marginal" rather than "non-existent".  But I doubt you'd
want to wait for the computation times needed for such
calculations on a typical machine of today (with a few
hundreds megabytes and several hundreds Meaningless
Instructions Per Second), particularly with the multiplication
algorithm that is used by Python's builtin 'long integer'
numbers.

For example, 10000! is a 35,600-digits number (starting
with 284) and (on my old box) is computer in just over
2 seconds with the simple loop using Python longs, half
a second with the loop using gmpy.mpz integers, and just
under 1/4 of a second using gmpy.fac (the gmpy module is
at http://gmpy.sourceforge.net/).

Here, the memory used for range(10000) is a piddling
40k bytes or so, and the memory used for the multiplicand
and result at the final step (about 70,000 digits between
them) is of similar magnitude (perhaps 20+ kbytes).

20000! is a 77,338 digits number (starting with 181) and
its computation takes (again, on my specific box) a bit
more than 11 seconds with ordinary Python long integers,
more than 3 seconds usign gmpy.mpz, and 0.6 seconds using
gmpy.fac.  The memory is still a tiny amount, computation
times are beginning to become noticeable (and growing
far faster than memory consumption, which, all in all, is
only marginally worse than linear).

40000! is a 167,714 digits number (starting with 209) and
its computation takes over 54 seconds with Python long ints,
over 9 with gmpy.mpz's, 1.6 with gmpy.fac.


We can see that a doubling of N means (roughly) a doubling of
the needed memory, with a computation about 5 times slower
for Python longs, 3 to 4 times for gmpy.mpz, 2 to 3 times
for gmpy.fac.  And we're still below 1/4 of a megabyte of
space taken -- with 8 more doublings left to get to a still
reasonably modest 64 MBytes of memory space, we could well
extrapolate that computing the factorial of 10,000,000 would
take over 270 days of time with Python longs, and be a pretty
big problem even for gmpy.fac.  (gmpy.fac actually takes
150 seconds to compute the factorial of 1,000,000, which is
a 5,565,709-digits number).


Alex





From scott at csihq.com  Wed Dec 13 13:27:04 2000
From: scott at csihq.com (Scott Bortman)
Date: Wed, 13 Dec 2000 13:27:04 -0500
Subject: Moving from Perl to Python
References: <3a3a9552.76673185@news.tcn.net> <slrn93fcdt.a8m.scarblac@pino.selwerd.nl> <3a38b798.85448901@news.tcn.net>
Message-ID: <t3fffuhm038120@corp.supernews.com>

Richard,

I too own Learning Python by O'Reilly and it is a great
intro to Python.  It covers the basics well.  I though about
getting Programming Python but like the other guy said,
it seems a bit dated now.  You can check out the TOC at
the O'Reilly site as well as a few sample chapters.  And
there is also a cheap (yet handy) pocket reference.

Hope this helps,
scottie

"Richard Zilavec" <rzilavec at tcn.net> wrote in message
news:3a38b798.85448901 at news.tcn.net...
> On 13 Dec 2000 17:30:56 GMT, scarblac at pino.selwerd.nl (Remco Gerlich)
> wrote:
>
> >Personally I don't like Programming Python much. It's becoming old (there
> >are rumors of a new version soon) and not very well structured. But then,
> >it's a lot of book, there are many examples, and the important things are
> >probably all in there somewhere.
>
> I hope I enjoy it.  I don't think I will ever stop using Perl.. but I
> really need something new.
>
> >I started with Learning Python, and I like that book. It explains the
>
> Wish I could see this book first, I don't want anything too basic
> maybe just a quick start, then more advance features.
>
> >Python Essential Reference is a nice reference book. It has a short
> >tutorial, then a good reference on the language and of the standard
> >library. One of the best books out there at the moment.
> >
> >More books at http://www.python.org/psa/bookstore/
> >
> >There is good online documentation at http://www.python.org/doc/current/
.
>
> Thanks!
>
> >There are things out there on the web that explain Python to Perl
programmers,
> >but I can't find them at the moment.
>
> I just found a man pages on another server, should be a good start..
>
> >
> >There are some posts by Tom Christiansen to this group that you can find
> >on Deja about Perl v Python, those may be useful.
>
> I found this article at perl.com, it was quite good.
>
> --
>  Richard Zilavec
>  rzilavec at tcn.net




From cjensen at bioeng.ucsd.edu  Fri Dec 29 12:43:32 2000
From: cjensen at bioeng.ucsd.edu (Curtis Jensen)
Date: Fri, 29 Dec 2000 09:43:32 -0800
Subject: Python core dumping
References: <92deoj$b2q$1@bob.news.rcn.net>
Message-ID: <3A4CCD44.FDCF2082@bioeng.ucsd.edu>

David Allen wrote:
> 
> I have a medium sized python application which is causing
> python to segfault under Linux at very odd times.  I'd
> like to get a core dump, so I can try to figure out
> what on earth is going on with it, and possibly post
> a bug report or something similar.  But as odd as this
> sounds, the application will NOT dump core.  The
> program I'm running does not use any C extensions,
> it's written in pure python.
> 
> I've tried using:
> Python 1.5.2 (#0, Apr  3 2000, 14:46:48)  [GCC 2.95.2 20000313 (Debian GNU/Linux)] on linux2
> and
> 
> Python 2.0 (#1, Dec 17 2000, 18:36:14)
> [GCC 2.95.2 20000220 (Debian GNU/Linux)] on linux2
> 
> Both after issuing "ulimit -c unlimited" which gives
> this:
> 
> [x at test pscr]$ ulimit -a
> core file size (blocks)     unlimited
> data seg size (kbytes)      unlimited
> file size (blocks)          unlimited
> max locked memory (kbytes)  unlimited
> max memory size (kbytes)    unlimited
> open files                  1024
> pipe size (512 bytes)       8
> stack size (kbytes)         8192
> cpu time (seconds)          unlimited
> max user processes          256
> virtual memory (kbytes)     unlimited
> 
> So the lack of core dumping is not due to my
> environment as far as I can tell.
> 
> I did not compile 1.5.2, that's the default version
> that's in Debian GNU/Linux (Woody) right now.  As
> for 2.0, I only "./configure"'d it with a --prefix
> option.
> 
> How can I get python to dump core so I can figure out
> what's wrong with my program, or alternatively, what's
> wrong with python?  Any help would be appreciated.
> 
> --
> David Allen
> http://opop.nols.com/

You can also run with the -v option and you'll get all kinds of debuging
output.  It should tell you at which point it dumps.

-- 
Curtis Jensen
cjensen at bioeng.ucsd.edu
http://www-bioeng.ucsd.edu/~cjensen/
FAX (425) 740-1451


From dieter at handshake.de  Sun Dec  3 18:00:31 2000
From: dieter at handshake.de (Dieter Maurer)
Date: 04 Dec 2000 00:00:31 +0100
Subject: [Bug?] Python 2.0: memory leak in "pdb"?
Message-ID: <x7r93p6rg0.fsf@lindm.dm>

When I tried to analyse a bad time complexity in 4XSLT,
I noticed that the program required at least twice as much main memory
if run under pdb than in a free run.

This suggests there might be a severe memory leak in pdb.
The debugger had to look for two breakpoints.


Dieter


From phd at phd.pp.ru  Thu Dec 28 09:15:05 2000
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 28 Dec 2000 17:15:05 +0300 (MSK)
Subject: Interesting behaviour of the assignment
In-Reply-To: <92f4gk$qk6$2@news.nuri.net>
Message-ID: <Pine.LNX.4.30.0012281713450.2265-100000@fep132.fep.ru>

On Thu, 28 Dec 2000, June Kim wrote:
> >>> a=100
> >>> b=100
> >>> a is b
> 0
> >>> a=99
> >>> b=99
> >>> a is b
> 1
>
> Is this an intended/expected result? What is the lying structure beneath
> this?

   First 100 integers (0-99) are preallocated by interpreter. So every time
you use it - the same object is used. All other integers are allocated anew
every time.

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




From trotts at llnl.gov  Wed Dec 27 14:30:56 2000
From: trotts at llnl.gov (Issac Trotts)
Date: Wed, 27 Dec 2000 11:30:56 -0800
Subject: coding convention conversion
In-Reply-To: <3A45196E.26B5BA6A@recombinant.demon.co.uk>; from stevee@recombinant.demon.co.uk on Sat, Dec 23, 2000 at 09:30:22PM +0000
References: <20001222122015.A7987@llnl.gov> <mailman.977522226.5104.python-list@python.org> <oSV06.176$g03.67749@sapphire.mtt.net> <mailman.977550847.24965.python-list@python.org> <3A45196E.26B5BA6A@recombinant.demon.co.uk>
Message-ID: <20001227113056.A20002@llnl.gov>

Stephen,

That seems to be what I had in mind.  I'll have a look.

Thanks,
Issac

On Sat, Dec 23, 2000 at 09:30:22PM +0000, Stephen D Evans wrote:
> Issac
> 
> There is a utility by the name of fixcid.py with the python distribution in the
> Tools\scripts directory that does 'massive identifier substitution on C source
> files'. It looks like it will work on C++ too. You'd probably have to get the
> (unmangled) function names from the linker.
> 
> Stephen D Evans
> 
> 
> Issac Trotts wrote:
> 
> > OK, that should do the trick for python.  Any ideas about how to
> > do it for C++ (using python, so that this isn't off-topic ;).
> >
> > Issac
> >
> > On Sat, Dec 23, 2000 at 04:29:08AM +0000, Bruce Dodson wrote:
> > > Better yet, import the file as a module, and scan through its public symbols
> > > (via dir()) to find out which names need to be changed, and then use that
> > > information to build a sed script which can be applied globally.  You can
> > > easily take advantage of other meta-information such as the type, e.g. limit
> > > it to just functions if you want.  This approach is good because the parser
> > > will recognize more variations than a regexp.
> > >
> > >
> > > Something like this is what I had in mind:
> > >
> > > from types import ModuleType, ClassType
> > >
> > > def print_symbols(symbols, prefix=''):
> > >     for sym in dir(symbols):
> > >         if sym[:2] == '__': continue
> > >
> > >         v = getattr(symbols, sym)
> > >         if type(v) is ModuleType: continue
> > >
> > >         print prefix + sym
> > >
> > >         if type(v) is ClassType:
> > >             print_symbols(v, prefix + sym + '.')
> > >
> > >
> > > See also the pyclbr module for a way to do something similar without
> > > executing the module.
> > >
> > > Bruce
> > >
> > >
> > >
> > > "D-Man" <dsh8290 at rit.edu> wrote in message
> > > news:mailman.977522226.5104.python-list at python.org...
> > > >
> > > > This shouldn't be too hard with a python script and some regexes with
> > > > some more string processing in between (or maybe multiple pass).
> > > >
> > > > Everything you want to change is bounded by "def" on one side and "("
> > > > on the other.  Then replace the capitals with _ followed by the lower
> > > > case version.
> > > >
> > > > A quick (partial) solution:
> > > >
> > > > re = re.compile( "def ([a-z]+)([A-Z])([a-z]*)\(" )
> > > >
> > > > for line in source.readlines() :
> > > > match = re.match( line )
> > > > if ( match ) :
> > > > res = "def " + re.group( 0 ) + "_" + \
> > > > string.lower( re.group( 1 ) + re.group( 2 ) + "("
> > > >
> > > >
> > > > This isn't complete and won't have the arguments in the result (nor
> > > > preserve indentation before the "def").
> > > >
> > > > I think the re.sub() function will do more of what you want it to.
> > > >
> > > >
> > > > HTH,
> > > > -D
> > > >
> > > >
> > > > On Fri, Dec 22, 2000 at 12:20:15PM -0800, Issac Trotts wrote:
> > > > > Does anyone know of a utility to change between the coding conventions
> > > > > of
> > > > >
> > > > > def fooBarBaz(): pass
> > > > >
> > > > > and
> > > > >
> > > > > def foo_bar_baz(): pass
> > > > >
> > > > > It is a pain to do this by hand.
> > > > >
> > > > > Thanks,
> > > > > Issac
> > > >
> > >
> > > --
> > > http://www.python.org/mailman/listinfo/python-list
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list



From kare at speech.kth.se  Sat Dec  9 11:34:47 2000
From: kare at speech.kth.se (Kare Sjolander)
Date: Sat, 09 Dec 2000 17:34:47 +0100
Subject: ANNOUNCE: The Snack Sound Toolkit v2.0.4
Message-ID: <3A325F27.880D44F9@speech.kth.se>

This release contains:

* Bug-fixes

* Updated documentation

* Updated demos

The Snack Sound Toolkit is designed to be used with a scripting
language.
Currently it supports Tcl/Tk and Python.
Snack adds commands to play, record, and process sound and supports
in-memory sound objects, file based audio, and streaming audio.
It handles fileformats such as WAV, MP3, AU, AIFF, and NIST/Sphere.
Snack is extensible, new commands, filters, and sound file formats can
be added using the Snack C-library. An easy to build example extension
is contained in the source distribution.
Snack also does sound visualization, e.g. waveforms and spectrograms.
The visualization objects update in real-time and can output postscript.

Snack works with Tcl8.0 - Tcl8.4 and Python 1.5.2-2.0
Platforms: Linux, Solaris, HP-UX, IRIX, NetBSD, Macintosh, and
Windows 95/98/NT/2K.

Source and binaries can be downloaded from
http://www.speech.kth.se/snack/

Regards,
Kare Sjolander
kare at speech.kth.se




From robin at jessikat.fsnet.co.uk  Mon Dec 11 05:48:47 2000
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Mon, 11 Dec 2000 10:48:47 +0000
Subject: ReportLab 1.02
Message-ID: <UPzUkAAPELN6EwLN@jessikat.demon.co.uk>

ReportLab 1.02 has been released.

See http://www.reportlab.com for download instructions.

This is mainly a bug fix release, but there ar many improvements and
fixes especially to layout and paragraphing code.
Paragraph now allows <onDraw...> callbacks, background colors.
Pythonpoint now has a DTD.
-- 
Robin Becker


From loewis at informatik.hu-berlin.de  Thu Dec 14 16:19:15 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 14 Dec 2000 22:19:15 +0100
Subject: Deleting a stored file in a ZipFile
References: <91aoe1$1cno$1@news.rchland.ibm.com>
Message-ID: <p6qelzavgz0.fsf@informatik.hu-berlin.de>

"Larry Whitley" <ldw at us.ibm.com> writes:

> Looking over ZipFile, it looks like it's not possible to delete a file from
> the archive.  True?

True. I guess code contributions are welcome.

Regards,
Martin


From root at rainerdeyke.com  Sun Dec 10 13:30:47 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Sun, 10 Dec 2000 18:30:47 GMT
Subject: imputils problem
References: <aGzY5.6237$x6.4746529@news2.rdc2.tx.home.com> <90viai$slg$1@merlin.rembrandtstr.bocholt.de>
Message-ID: <rZPY5.11237$x6.6306116@news2.rdc2.tx.home.com>

"Magnus Heino" <magnus.heino at rivermen.se> wrote in message
news:90viai$slg$1 at merlin.rembrandtstr.bocholt.de...
>
> > I've been having a problem with imputils: a module imported through a
> > custom
> > importer seems unable in import .pyd modules.  In particular, if I
import
> > the zipfile module though a custom importer, it is unable to find the
zlib
> > module.  Does anybody here know how to fix this?
>
> I have it working, but I had to tweek imputils.py a bit.

Strangely enough, the problem went away when I moved zlib.pyd from the DLLs
subdirectory to the main directory of my program (which is a C++ program
with embedded Python 2.0).  The question: why did this change anything?
Does the default __import__ search directories that are not in sys.path?

> How are you doing things?

I took inspiration from the "small" Python distribution and provided a
custom site.py:

--- start.py ---
import marshal
import imputil
import cStringIO
import sys
import struct


class ArchiveImporter(imputil.Importer):

  def __init__(self, name):
    self.file = open(name, 'rb')
    (pos,) = struct.unpack('<l', self.file.read(4))
    self.file.seek(pos)
    self.toc = marshal.load(self.file)

  def get_code(self, parent, modname, fqname):
    name = fqname.replace('.', '/')
    if self.toc.has_key(name):
      return (0, self.__load(name), {})
    elif self.toc.has_key(name + '/__init__'):
      return (1, self.__load(name + '/__init__'), {})
    else:
      return None

  def __load(self, name):
    self.file.seek(self.toc[name])
    return marshal.load(self.file)


if __name__ != '__main__':
  imputil.ImportManager().install()
  sys.path.append(imputil.BuiltinImporter())
  sys.path.append(ArchiveImporter('code'))

--- end site.py ---



--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From gregj at pobox.com  Wed Dec 13 02:55:02 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Wed, 13 Dec 2000 07:55:02 GMT
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <29lZ5.138780$U46.4638547@news1.sttls1.wa.home.com> <uhf49kf8r.fsf@roundpoint.com>
Message-ID: <qXFZ5.142740$U46.4719533@news1.sttls1.wa.home.com>

"Ben Hutchings" <ben.hutchings at roundpoint.com> wrote in message
news:uhf49kf8r.fsf at roundpoint.com...
> "Greg Jorgensen" <gregj at pobox.com> writes:
> <snip>
> > A flame war broke out a few years ago in comp.lang.c++ over which
> > was better: for ( i=0; i < N; i++ ) or for (i=0; i < N; ++i ). The
> > result is exactly the same, and K&R use the first (postincrement)
> > idiom, but it turns out that ++i is usually a tiny bit more
> > efficient than i++.
> <snip>
>
> For iterators of built-in types (like int) it's unlikely to matter,
> but for iterators of user-defined types the latter may be less
> efficient as it's harder for the compiler to optimise the code for the
> overloaded post-increment operator.

The problem is the temporary variable. The value of i++ is i before the
increment, so the compiler has to save the current value before
incrementing:

    int postincrement(int &i)
    {
        int temp = i;
        i = i + 1
        return temp;
    }

The value of ++i is i after the increment, so no temporary is necessary:

    int preincrement(int &i)
    {
        i = i + 1
        return i
    }

The difference is negligible in most cases, and modern compilers easily
optimize the temporary out anyway. I was using the flame war over this as an
example of programmers defending a familiar idiom.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From bkc at Murkworks.com  Fri Dec 22 12:40:34 2000
From: bkc at Murkworks.com (Brad Clements)
Date: Fri, 22 Dec 2000 12:40:34 -0500
Subject: <> vs != was Re: A Suggestion for Python Colon Syntax
References: <mailman.977446446.16022.python-list@python.org>
Message-ID: <2kM06.120$Lm2.7518@news.clarkson.edu>

"Tim Peters" <tim.one at home.com> wrote in message
news:mailman.977446446.16022.python-list at python.org...

> Guido abhors gratuitous variations in coding sytle; that's largely why
> indentation is used instead of curly braces or begin/end to begin with.  A
> choice about whether or not to use a colon is simply not something people
> *need*.  The few places Python allows gratuitous choices today are places
> Guido regrets (although not enough to lose sleep over <wink>).  For
example,
> Guido is unhappy that both "<>" and "!=" are accepted to mean "not equal";
> the older "<>" survives only for compatibility with a very early release
of
> the language.

Horrors, I just checked some of my code and find that I use both != and <>
in the same module!

I don't know why.. Python is so easy to write for, I just use whatever is
easier to type, so if my hands are low on the keyboard, out comes <> ..
hands higher up then !=

I think I'll stick with <>, but then I'll have to *think* about the syntax,
rather than the algorithm.. something I normally don't have to do in
Python.. What a lovely language.






From nbecker at fred.net  Wed Dec 13 08:21:26 2000
From: nbecker at fred.net (nbecker at fred.net)
Date: 13 Dec 2000 08:21:26 -0500
Subject: Self-documenting Python??
Message-ID: <x881yvcfodl.fsf@adglinux1.hns.com>

Has there been any proposals to integrate a documentation system with
python?  I feel this is one feature that is sorely missing.

We have a way to put some very minimal documentation into python
sources, but there is no integrated help browser to use it.

I use RedHat Linux here.  The standard Python docs are installed in a
manner by the standard distribution that they are not accessible
except by users sufficiently knowledgeable to know where and how to
look for them.  Most people can use a web browser, but they don't know
that /usr/share/doc/python.... has something of interest.
Furthermore, documentation for user-written code is not accessible at
all.

I believe nothing would do more to further the wider acceptance of
Python than integrated help.


From tim.one at home.com  Thu Dec 21 12:52:37 2000
From: tim.one at home.com (Tim Peters)
Date: Thu, 21 Dec 2000 17:52:37 GMT
Subject: A Suggestion for Python Colon Syntax
References: <91tc8h$2pp$1@hecate.umd.edu>
Message-ID: <Frr06.194843$RP.35747437@news1.rdc1.md.home.com>

[William Djaja Tjokroaminata]
> ...
> However, a lot of time when I ran my Python program, I got syntax error,
> because I forgot to put colon (:) after my if (or for, etc.) clause.  My
> question is, is the colon really necessary, syntactically, in
> Python?  Because Python is line-and-indentation-based, can we just
> eliminate the colons after the if, for, etc., clause?

Python's indentation rules are inherited from the ABC language (Guido was a
member of the ABC implementation team).  The ABC designers tried several
styles of denoting block structure, and ran usability studies with newbies
to determine which were easiest to learn.

Guido reported that indentation worked best in these studies, but--
curiously enough! --only if the line immediately preceding the new block
ended with a colon.  And that's why Python requires the colon too:  it's not
because the parser needs it, but because a study once concluded *people* do
better with it.

> My suggestion is as follows:
                             ^
> 1) If the if (or for, etc.) clause is followed by a block, then the
> colon is optional (for backward compatibility):
                                                ^
> ...

Also curious:  you never forget the colon at the ends of your introductory
sentences <wink>.

without-a-colon-life-itself-is-impossible-ly y'rs  - tim





From slhath at home.com  Sun Dec 24 21:07:35 2000
From: slhath at home.com (Scott Hathaway)
Date: Mon, 25 Dec 2000 02:07:35 GMT
Subject: xml-rpc server in python returning python lists...
References: <hih16.85504$yR4.2492277@news1.rdc1.tx.home.com> <925o9s$pqk$1@bunyip.cc.uq.edu.au>
Message-ID: <HZx16.87713$yR4.2521840@news1.rdc1.tx.home.com>

I am using the one from Steve Ivy and Joe Massey.  The other one for COM, I
cannot seem to get working.  Which one did you use with VB?

Scott

BTW, here is the vb code I cannot get working:

Dim rpc, theArray, x
Set rpc = CreateObject("XMLRPC.XMLRPCClient")
x = rpc.vbcall(theArray, "http://localhost:8000", "GetParamDefaultValues",
"c:\projects\core\test.crf", "bgColor")

Thanks,
Scott


"Chui Tey" <teyc at bigfoot.com> wrote in message
news:925o9s$pqk$1 at bunyip.cc.uq.edu.au...
> As far as I know, VB takes it as an array of Variants. Worked for me
before.
> Which XMLRPC Client are you using?
>
> Scott Hathaway <slhath at home.com> wrote in message
> news:hih16.85504$yR4.2492277 at news1.rdc1.tx.home.com...
> > Hello All,
> >
> > I am writing an xml-rpc server in python that has a function which
returns
> a
> > python list.  When I use the python client to connect to the python
> server,
> > it works great.  When I try from VB, ASP, PHP, and Delphi, they all have
> > trouble with the return type of the variable.  By the way, all the
clients
> > mentioned work fine if the server's method returns a string or a number.
> >
> > Any ideas?
> >
> > Scott
> >
> >
>
>




From duraip at extendsys.com  Wed Dec 13 07:21:48 2000
From: duraip at extendsys.com (pjd)
Date: Wed, 13 Dec 2000 12:21:48 GMT
Subject: unpacking Python tuple in C
References: <917e64$3bp@micky.ibh-dd.de>
Message-ID: <917pkp$d7a$1@nnrp1.deja.com>

>
> PyArg_ParseTuple(result, "dd",  &double1, &double2)
>

> PyObject* floatFromPy0 = PyTuple_GetItem(result, 0);
>   *double1= PyFloat_AsDouble(floatFromPy0)

I suspect you are passing a pointer to pointer (to a double) in
PyArg_ParseTuple. It should just be a pointer.

cheers
pj


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


From aleaxit at yahoo.com  Wed Dec 13 05:01:46 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 13 Dec 2000 11:01:46 +0100
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <kuzoi15d60.fsf@lasipalatsi.fi>
Message-ID: <917heq021si@news2.newsguy.com>

"Erno Kuusela" <erno-news at erno.iki.fi> wrote in message
news:kuzoi15d60.fsf at lasipalatsi.fi...
    [snip]
> | - garbage collection is reference counted, though now enhanced in Python
> | 2.0 so it can handle most cases of circular references
>
> i don't think there are any situations in which the 2.0 system
> would not handle them.

I think that, if the objects in a circular-references loop have
user-defined __del__ methods, the 2.0 GC (quite reasonably)
will not remove them (as any order of removal might cause some
trouble or other in the general case).

> in fact the refcounting + "real" gc combination is much nicer than
> pure gc, because of the predictable finalisation behaviour.

The counterpoint to that is performance -- a good modern 'pure
GC' approach can squeeze substantially more performance out of
a system, than a 'pure RC' or 'mixed GC/RC' approach.  It's a
moot point whether somewhat-predictable-finalization is worth
that much (judging from the squeals about it on discussion
forums related to Microsoft .NET -- which moves to "pure GC",
while Microsoft's previous tradition was for reference counting --
one might think so -- but, IMHO, that's partly an artefact of
what people developing for MS platforms were previously used to;
I have not seen anything like that amount of heat concerning
the same "pure GC" when a language/environment uses it right
from the start, cfr. Lisp, Scheme, Smalltalk, Java, Eiffel...).


> | It's an absolutely stunning language for free, and although not as fast
> | as Perl
>
> (it's as fast in my experience)

Mine too.  But, more significantly, Lutz Prechelt's important
empirical study also showed no statistically significant
difference between equivalent programs for a given task as
developed in Perl or Python (there _was_, if I recall this
correctly, statistical significance to other performance
comparisons -- C/C++ programs a bit faster, Java, Tcl,
and Rexx slower, than the Perl/Python 'cluster').


Alex





From ennisr at my-deja.com  Wed Dec 27 11:52:36 2000
From: ennisr at my-deja.com (Ennis Rosamond)
Date: Wed, 27 Dec 2000 16:52:36 GMT
Subject: python telnet script for openvms host
References: <92ainp$jab$1@nnrp1.deja.com> <kuito7rrg7.fsf@lasipalatsi.fi> <92apa7$oac$1@nnrp1.deja.com> <D8D70FCFA8CC11C0.4BEE4AB5F2A3BCF7.EAD0DA106483FAE8@lp.airnews.net> <kuelyvrmqw.fsf@lasipalatsi.fi>
Message-ID: <92d6oh$ghf$1@nnrp1.deja.com>

The script is now working.  Thanks to everyone who responded to my
request for help.  As several people pointed out, the problem was in the
telnet protocol negotiation.  Since I knew virtually nothing about the
nitty gritty of telnet communication, I quickly read a short article on
telnet protocol that I found on the web
(http://www.scit.wlv.ac.uk/~jphb/comms/telnet.htm).  That gave me just
enough information get the script working, though not very elegantly.

Looking at the actual client/host dialog with the aid of the telnetlib
debug option, it appeared that perhaps the host was proposing 'suppress
go ahead', but that acceptance of the option was not being returned by
the client.  (The actual host message was '\377\373\003'.  I haven't yet
found the interpretation of either 377 or 373.)

To fix the script, I added a write call to the host with the message
'\255\253\003', meaning 'do suppress go ahead', following the login
dialog but before scanning for the '$ ' prompt.  I doubt that this a
good general fix-- seems to work here but probably not in other
situations.  The new script is shown below.

#python test script:
import sys, telnetlib
host = 'hostname'
user = 'username\r'
password = 'mypasswd\r'
tn = telnetlib.Telnet(host)
tn.read_until('Username: ')
tn.write(user)
tn.read_until('Password: ')
tn.write(password)
tn.read_until('[c')             #last char of session header message
tn.write('\255\253\003')        #this is the 'fix' command added
tn.read_until('$ ')             #scan for prompt
tn.write('dir\r')               #issue directory command
d = tn.read_until('$ ')         #save results in d
tn.close()
print d

The underlying problem is obviously the lack of option negotiation logic
in the telnetlib module.  (It is on the 'To do' list in the telnetlib
source code.)  Someone with telnet protocol expertise would do the
Python community a valuable service by adding some sophisticated
negotiation code to the module.

(Posted to both Python and OpenVMS discussion groups.)


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


From tim at degree.ath.cx  Mon Dec 11 20:49:47 2000
From: tim at degree.ath.cx (Tim Hammerquist)
Date: Tue, 12 Dec 2000 01:49:47 GMT
Subject: Python Problem - Could not find platform independent
References: <9105de$20l$1@venus.telepac.pt> <912vql$llc$03$1@news.t-online.com>
Message-ID: <slrn93b15n.63r.tim@degree.ath.cx>

Richard Sharp <rbsharp at gmx.de> wrote:
> It reminds me of a
> problem with Mandrake which is just a packaging of Redhat.

I was unaware that Red Hat and Mandrake were distributed by the same
source.  I know Mandrake claims to be (and so far has been) Red Hat/RPM
compatible.  So MacMillan distributes Red Hat in addition to Mandrake?

-- 
-Tim Hammerquist <timmy at cpan.org>
Democracy encourages the majority to decide
things about which the majority is ignorant.
	-- John Simon


From gregj at pobox.com  Mon Dec 11 01:15:07 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Mon, 11 Dec 2000 06:15:07 GMT
Subject: Programming Habits in Python
References: <cspU5.82459$U46.2617014@news1.sttls1.wa.home.com> <slrn92tv0u.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com> <udn1e8k0qi.fsf@box.home.de> <slrn937caf.h4.amk@207-172-146-162.s162.tnt3.ann.va.dialup.rcn.com> <682DDF8D24ABBCEB.7AA1428334F7B3A8.A771CF7758416C27@lp.airnews.net>
Message-ID: <Lh_Y5.133482$U46.4493934@news1.sttls1.wa.home.com>

The quote you are probably thinking of is from Samuel Johnson: "Sir, a
woman's preaching is like a dog's walking on his hind legs. It is not done
well; but you are surprised to find it done at all." Anything similar is a
derivative work. From Boswell's "Life of Johnson."


"Tres Seaver" <tseaver at starbase.neosoft.com> wrote in message
news:682DDF8D24ABBCEB.7AA1428334F7B3A8.A771CF7758416C27 at lp.airnews.net...
> In article
<slrn937caf.h4.amk at 207-172-146-162.s162.tnt3.ann.va.dialup.rcn.com>,
> A.M. Kuchling <akuchlin at mems-exchange.org> wrote:
>
> >In the back of my mind I had a quotation, which I cannot track down on
> >the Web.  The Principia uses geometrical methods throughout, even
> >where analytical methods would have been much simpler, and someone
> >(Halley, perhaps) said that reading the Principia was like walking
> >through an arsenal of giant weapons of a past age, and that one is
> >amazed that anyone could wield them at all.
>
> I don't think it could be Halley;  he was actively involved (as the
> then secretary of the Royal Society) in getting the Principia published.
> It sounds a bit like an analogy my memory wants to attribute to
> Bertrand Russell:  "the amazing thing about a waltzing bear is not
> how well he waltzes, but that he waltzes at all."

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From sholden at holdenweb.com  Fri Dec 15 11:49:09 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 15 Dec 2000 11:49:09 -0500
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <3a36c85e.4409136@News.CIS.DFN.DE> <owen_nospam-1312001131590001@cs204-90.fsmodem.washington.edu> <3A37DD31.C9A5F18C@pacbell.net> <91bhlh$c64u$1@nntp6.u.washington.edu> <3A395197.CB2974D5@home.com>
Message-ID: <L_r_5.7453$n51.136546@e420r-atl1.usenetserver.com>

James A. Robertson <jarober at home.com> wrote in message
news:3A395197.CB2974D5 at home.com...
> "Russell E. Owen" wrote:
> >
>
> >
> > My point is, that Cincom's pricing policies look self-destructive to me.
> > I cannot in good conscience make my observatory rely on a language that
> > I fear will continue to be a marginal player, at best.
> >
>
> How so?  We went way over our sales targets last year, and things are
> looking pretty good for the upcoming year.  Cincom has devoted a lot of
> resource into Smalltalk; we are committed to it.
>
Unfortunately, sales targets are arbitrary numbers.  What evidence do you
have that revenues might have been increased if you had reduced the price?
I'm sure you can appreciate that if you sold for one-third the price and
made six times as many sales, profits would be up.  Most sales organizations
I've worked with don't have a model which shows profits as a function of
price.

The fact that I've seen two postings on this thread complaining that your
prices are too high suggests sales would increase with lower pricing.  Of
course the effect on profit is unknown.

The fact that an academic version is available redeems your policy somewhat,
but I suspect your target market may be nervous about the company's survival
because it feels your pricing policies are unrealistic.

> > I'm also reluctant to write an application with a "freebie" splash
> > screen that comes up every time. And I'm reluctant to use free versions
> > of commercial stuff we cannot afford, because it doesn't support the
> > company and because their pricing policies are risky -- if they dump the
> > free version we're in deep trouble.
>
> If an asteroid hits the planet, you're in trouble too.  Nothing is
> certain.  If Sun's hardware sales slip at some point, Java is uncertain
> as well.  They aren't making money on Java itself.
>
Well, that's probably why this discussion is being held on the Python group!
Clearly Sun's demise would impact a lot of people and upset a lot of plans,
but they do have a certain financial momentum which means a crash would
probably be heralded by several years of poor performance.

Also, Java has a number of compatible compilers, so even if Sun died the
language would conceivably survive.  It sounds as though the degree of
SmallTalk compatibility is lower, effectively making the Cincom
implementation proprietary.

Sun's revenues are measured in the billions - they became a $1bn company in
year six, if I remember rightly.  Whereas Cincom's web site says "And during
the past three decades, our annual sales have grown to $185 million".  Not
quite so impressive.

Is there NOBODY in the company who suspects this slow growth might be due to
overpricing?  I'm not denigrating your figures in absolute terms (I achieved
much less in my own company), but other companies have grown much larger
revenues in the same period, and thereby become a better long-term bet.

> >
> > If there were compatible implementations I'd not worry so much about the
> > stability of one company, but I'm reluctant to hitch our star to Cincom.
> >
> > Smalltalk is my favorite language. However, I am the only one at the
> > observatory who knows it and feels it is worthwhile. As I fight Python's
> > disgusting GUI I keep wavering, and I may yet give up and switch to
> > Smalltalk. The worst they can do is fire me. But I love my job.
> >
> > -- Russell
>
> --
> James A. Robertson
> Technical Product Manager (Smalltalk), Cincom
> jarober at mail.com
>
> <Talk Small and Carry a Big Class Library>
>
regards
 Steve





From erno-news at erno.iki.fi  Thu Dec 14 22:53:22 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 15 Dec 2000 05:53:22 +0200
Subject: [Q] Are Exceptions used that much in practice?
References: <3A35D546.A4731B5B@enteract.com> <914mfd02ibt@news2.newsguy.com> <3A36F20B.7B766D0@cosc.canterbury.ac.nz> <uaea0ghlx.fsf@ctwd0143.fitlinxx.com> <jon.976778599@bezek.dstc.monash.edu.au> <3A397B7E.A4D2F60D@cosc.canterbury.ac.nz>
Message-ID: <ku66km5oi5.fsf@lasipalatsi.fi>

In article <3A397B7E.A4D2F60D at cosc.canterbury.ac.nz>, Greg Ewing
<greg at cosc.canterbury.ac.nz> writes:

| Yeah, except that the error numbers it gives you
| seem to be platform-dependent.

the idea is to compare them to the symbolic constants in the errno
module.

  -- erno


From amk at mira.erols.com  Tue Dec 12 10:12:38 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 12 Dec 2000 15:12:38 GMT
Subject: Request for poll() help
Message-ID: <slrn93cgru.j9.amk@207-172-113-220.s220.tnt5.ann.va.dialup.rcn.com>

I have 3 bugs filed against me relating to the poll() support in the
select module.  Unfortunately all of them are on platforms that I don't
have access to:

Debian 2.2 (I suspect using Linux kernel 2.0.x is the problem here):
 http://sourceforge.net/bugs/?func=detailbug&bug_id=115221&group_id=5470

Linux 2.0.38, glibc 2.0.7:
 http://sourceforge.net/bugs/?func=detailbug&bug_id=115617&group_id=5470

Redhat 5.2 on SPARC:
 http://sourceforge.net/bugs/?func=detailbug&bug_id=116547&group_id=5470

If someone out there is still running a 2.0 kernel and wants to help
debug this, the assistance would be much appreciated.  Or, if I could
be pointed at a compile farm, freenet, or other place where I could
get a login on a 2.0 machine, that would be equally useful.

--amk 


From no_spam_eirik at stockpoint.no  Thu Dec 14 03:02:05 2000
From: no_spam_eirik at stockpoint.no (Eirik Mangseth)
Date: Thu, 14 Dec 2000 08:02:05 GMT
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <tcmay-FCA22F.14073713122000@news.got.net> <918v7a02scn@news1.newsguy.com> <tcmay-04E80C.16181313122000@news.got.net>
Message-ID: <18%Z5.984$VM5.179625984@news.telia.no>

"Tim May" <tcmay at got.net> wrote in message
news:tcmay-04E80C.16181313122000 at news.got.net...
> In article <918v7a02scn at news1.newsguy.com>, "Alex Martelli"
> <aleaxit at yahoo.com> wrote:
>
> > "Tim May" <tcmay at got.net> wrote in message
> > news:tcmay-FCA22F.14073713122000 at news.got.net...
> >     [snip]
> > > "And there seem to be very *few* good examples of where multiple
> > > inheritance, a la C++, is _actually_ used. Most examples seem to be
> > > contrived. "
> >
> > Have you looked into Coplien's "strangely recurring template
> > pattern", aka "the pattern without a name", and frameworks
> > developed around it, such as Microsoft's ATL and WTL?  The
> > multiplicity of inheritance in such frameworks is positively
> > dazzling to people who come from Cargill's old opinion about
> > the lack of usefulness in multiple inheritance.
>
> No, I haven't. And I thank you in all sincerity for pointing this out. I
> will look for it tonight.
>
> Being interested in ontology (things, objects, hierarchies, structures)
> even more so than in advocating any one language over another, I would
> be fascinated to see some good examples of MI in actual use.

Take a look at e.g. EiffelBase or any well designed Eiffel system and
you'll see how useful MI can be with the appropriate language mechanisms
in place to resolve any ambiguities that may arise. Another good source
is Object-Oriented Software Construction 2nd Edition, chapters 14, 15 and
16. The latter is called "Inheritance techniques". To someone who's used
to a good implementation of MI, having only SI is just to constraining.
But of course, if one enjoys writing delegates for those situations where
MI makes a "better" solution, then be my guest. Just don't force others
to do the same thing. And remember, just because C++ screwed up
MI doesn't imply that MI is bad in general. Too many people pass
judgement om MI based on their own or others experience with
MI in C++.

Eirik Mangseth
Stockpoint Nordic AS
Lysaker, Norway

"If I can't Eiffel in heaven, I won't go"

>
> > course having just 1 is still better than nothing:-).  But there
> > are plenty of good examples of multiple inheritance (of
> > implementation) being used effectively, in all languages
> > (that allow it) mentioned so far, and, I'd bet, in more yet
> > (Common Lisp, for example -- wasn't CLOS the specific
> > environment in which the concept of 'mixin inheritance' was
> > originally developed?).
>
> Actually, it was in Zetalisp, the dialect running in the old Symbolics
> Lisp Machines (later on other machines). Mostly the work of a young
> programmer named Howard Cannon (wonder what became of him?).
>
> --Tim May
>
> --
> Timothy C. May         tcmay at got.net        Corralitos, California
> Political: Co-founder Cypherpunks/crypto anarchy/Cyphernomicon
> Technical: physics/soft errors/Smalltalk/Squeak/agents/games/Go
> Personal: 1951/UCSB/Intel '74-'86/retired/investor/motorcycles/guns
>
>




From pblanchette at pixelsystems.com  Wed Dec 13 14:08:17 2000
From: pblanchette at pixelsystems.com (Patrick Blanchette)
Date: Wed, 13 Dec 2000 19:08:17 GMT
Subject: embeding problem 2: Py_Finalize didn't free all the memory
Message-ID: <3A37C910.3040004@pixelsystems.com>

I,
   I'm embeding python in a WinNT application.  This application can run 
a python script.  Before executing a python script, the application 
calls the Py_Initialize function and executes the script.  After the 
script execution, the application calls the Py_Finalyze function.  I 
notice that the application use more memory than before the script 
execution, I suppose that the Py_Finalyze function didn't not cleanup 
the python interpreter completely.  If I try to execute another script, 
an error happens!
   If I call Py_Initialize and Py_Finalyze only 1 time, everything works 
correctly.  The problem is that my application use more memory after 
each script execution. 
   I tried to call Py_NewInterpreter() function before executing a 
script and the Py_EndInterpreter() at the end of each script but  the 
memory usage still grow rapidly.

   What should I do to keep the memory usage low?  My application is 
supposed to be able to run continuously without restart.

Thanks in advance



From be at et.dtu.dk  Wed Dec 20 04:54:09 2000
From: be at et.dtu.dk (Brian Elmegaard)
Date: Wed, 20 Dec 2000 10:54:09 +0100
Subject: Customizing IDLE
Message-ID: <3A4081C1.DDBBE08C@et.dtu.dk>

Hi,

I would like to customize IDLE and use with another language. I mainly need to
change the keywords list and the compile command.

Can this be done easily, i.e., which files to change, and may I do it?

-- 
Brian Elmegaard (be at et.dtu.dk) http://www.et.dtu.dk/energysystems 
Dept. of Energy Engineering, Technical University of Denmark, 
Phone +45 4525 4169  Fax +45 4593 5215
:) http://www.rk-speed.dk http://fiduso.dk http://sunsite.auc.dk/dk-tug


From sholden at holdenweb.com  Sun Dec 31 11:16:29 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 31 Dec 2000 11:16:29 -0500
Subject: Newbie question ... reading modules from directories in Windows
References: <5Md36.46251$59.13170408@news3.rdc1.on.home.com>
Message-ID: <6%I36.196$9s1.38186@e420r-atl2.usenetserver.com>

Peder Jakobsen <jakobsen at home.com> wrote in message
news:5Md36.46251$59.13170408 at news3.rdc1.on.home.com...
> Just a small question:
>
> I have added a "C:\Python20\sandbox" directory to my PYTHONPATH
>
> Now I'm trying to import modules from that directory whole using the
> interactive Python prompt, but it only works if I start python.exe in the
> directory where the module resides.
>
> I thought by adding the directory where the .py files exist to the Path,
> Python would be able to detect them???
>
> Thanks in advance,
>
> Peder Jakobsen, Ottawa
>
>
Did you set an envvironment variable, or did you modify the registry?  I
suspect what you need to do is modify the
\HKEY_CURRENT_USER\Software\Python\PythonCode\2.0\PythonPath registry entry.

However, it's a lot easier to establish a sitecustomize.py moduile in your
library, and add any required values to sys.path in there: that way it will
haooen automatically every time Python starts.

regards
 Steve






From pdsmith at iinet.net.au  Sun Dec 24 20:29:30 2000
From: pdsmith at iinet.net.au (pdsmith at iinet.net.au)
Date: Mon, 25 Dec 2000 01:29:30 GMT
Subject: comp.lang.python.announce
References: <91aha7$791$1@reader1.imaginet.fr> <Ka5_5.3904$Kd1.451379@newsb.telia.net>
Message-ID: <3a46a2d2.6567084@news.iinet.net.au>

Isn't there a way of receiving the messages via email? Check the
python.org site.


From DanielK at aracnet.com  Sun Dec 31 11:14:27 2000
From: DanielK at aracnet.com (Daniel Klein)
Date: Sun, 31 Dec 2000 08:14:27 -0800
Subject: array of objects
References: <92n1s7$tq9$1@nnrp1.deja.com> <siE36.2779$Qb7.318673@newsb.telia.net>
Message-ID: <FRI36.76$KD3.30489@typhoon.aracnet.com>

How then is list comprehension different (and/or better) than say:

obj = []
for x in range(N): obj.append(test())

Happy (real) New Millenium,

Daniel Klein
Portland, OR USA

"Fredrik Lundh" <fredrik at effbot.org> wrote in message
news:siE36.2779$Qb7.318673 at newsb.telia.net...
> Anders Olme wrote:
> > How do I create a array of object from my own classes?
> >
> > If I have a class like this:
> >
> > class test:
> > def dosomething(self):
> > return 1234
> >
> > obj= test()
> >
> > obj= [test(),test()]
> >
> > The last line works but is there a more elegant way?
>
> Not if you need exactly two elements ;-)
>
> To create a list containing N instances, you can use a
> list comprehension:
>
>     obj = [test() for x in range(N)]
>
> which is short for
>
>     obj = []
>     for x in range(N):
>         obj.append(test())
>
> If the test class is "immutable" (that is, if it cannot be modified
> in place), you can also use "sequence multiplication":
>
>     obj = [test()] * N
>
> this creates a list with N references to the same instance.
>
> </F>
>
>




From lk at veriomail.com  Mon Dec 18 20:51:28 2000
From: lk at veriomail.com (Lawrence Kesteloot)
Date: Tue, 19 Dec 2000 01:51:28 GMT
Subject: An "exit" function for C modules?
Message-ID: <Aaz%5.2310$1%2.101471@sjc-read.news.verio.net>

I'm writing a C module where I need to keep a global
pointer to a Python object (for API convenience).

When the script finishes, I never get the chance to
decref the pointer, so it never gets destroyed.  This
is a problem with streaming output files that won't
get closed properly.  (Not text files -- this is a special
file format that requires official closing in order to be
updated properly.)

What's the best way to handle this?  I can't depend
on the user always calling a "close" function on the
stream, and the API would get much more complicated
if I didn't have these globals.  (The module is for
novice programmers.)

Should I use "atexit()"?  I'm concerned that I don't know
when that's called -- it could be after some cleanup has
been done by Python that makes decref dangerous.

It seems as though there should be an "exit<modulename>"
function like the "init<modulename>" call at an appropriate
time that allows cleanup of globals.

Lawrence







From grey at despair.rpglink.com  Sat Dec 30 13:20:00 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Sat, 30 Dec 2000 18:20:00 -0000
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <3A4D6A05.EEC97FA6@uswest.net> <slrn94qqvg.mgg.grey@teleute.rpglink.com> <ije36.96813$x6.43604873@news2.rdc2.tx.home.com> <slrn94qt4f.mgg.grey@teleute.rpglink.com> <CMo36.97857$x6.44474719@news2.rdc2.tx.home.com>
Message-ID: <slrn94s9qg.134.grey@teleute.rpglink.com>

On Sat, 30 Dec 2000 17:17:22 GMT, Rainer Deyke <root at rainerdeyke.com> wrote:
>In other words, you need to artifically manipulate the 'while' condition to
>always be true the first time through.  If you change the 'while' condition,
>you may have to change the manipulation.  In other words, there's an
>additional way the code could break.

    From what world do you come from where initilizing variables is bad?

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From dale at out-think.NOSPAMco.uk  Mon Dec  4 18:47:07 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Mon, 04 Dec 2000 23:47:07 +0000
Subject: Python programs (or scripts) as drop targets
References: <hv5k2tg9cdf33j8v333qpkk7jo45ht596n@4ax.com> <90h0gq$cu0$1@troll.powertech.no>
Message-ID: <1mao2tojod77d92ri841ptdvm6t3e0evqu@4ax.com>

"Syver Enstad" <syver.enstad at sensewave.com> wrote:

>> So is there anyone out there with the VC++ skills sufficiently interested
>in this idea to code the
>> necessary interfaces?
>
>If there isn't support for doing this in python I could look into it. I've
>made a couple of shell extensions before that sounds pretty similar so it
>shouldn't be too much work.
>
>Can't you contact me Dale if we can't find a way to do this in Python?
>

As far as I can tell, this isn't a big job but I think it would be neater if it was built into the
standard Python distribution. 

Mark has suggested  that it could go in win32com.shell - although I can't figure out what he means
by that - win32comext.shell, perhaps?

Is the source for this available?


--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants




From peter at engcorp.com  Mon Dec 11 08:42:35 2000
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 11 Dec 2000 08:42:35 -0500
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com>
Message-ID: <3A34D9CB.CB465246@engcorp.com>

Jaroslav Gergic wrote:
> 
> one thing I dislike in many Python tutorials is
> the style of infinite 'while' loop with 'break' statements:

[Rewrote original for compatibility with other examples:]

while 1:
   line = fh.readline()
   if not line: 
       break
   ... do something ...

> I hate it so much I prefer to write something like this:
> 
> line = fh.readline()
> while(line != ""):
>   ... do something ...
>   line = fh.readline()
> 
> OK, it is ugly to write the same line of code twice,
> but 'while true' or 'while 1' is very non-programmer construction
> it smells like VisualBasic GOTO.
> 
> Do you know better construction to avoid both - duplicate lines
> and horrible 'while 1' construction?
> 
> Something like:
> 
> while( "" != (line = fh.readline()) ):
>   ... do something ...

IMHO, you should evolve your sense of 'what is right' and accept that
the simpler style is better.  Both of the alternatives you suggest are
worse than the original!  

The former has the repetitive line of code which, while perfectly
correct, is as you say "ugly" (in that it is more likely to give trouble
during maintenance, and took a little longer to type, but that's about
the only reason).

The latter is just plain grim: C code written in Python! :)  A
confusing, compressed mess, useful only in an obfuscation contest or
where you print a lot and paper is expensive...

A clear advantage of the original is that it is much more readable and
is generally considered more "Pythonic" than the others.  I used to have
the same objections you do to this style, but after a few months of
using Python and considering the arguments in favour of the simpler
form, I learned was I believe was the error of my ways. :)  IMHO, as I
said.


From mla at 1012surf.net  Thu Dec 21 05:48:54 2000
From: mla at 1012surf.net (Michael Lausch)
Date: 21 Dec 2000 11:48:54 +0100
Subject: How to make a GUID with plain python
References: <3a413a17.3416402@192.168.0.1>
Message-ID: <cuiae9qkq21.fsf@loki.1012internet.at>

gogo at bluedynamics.com (Georg Gogo. BERNHARD) writes:

> Hello!
> 
> I am looking for a way to create a global unique identifier / UUID
> with python. I found a way to do this with ActivePython in Windows,
> but I need a solution for other platforms, too. (In ActivePython one
> can use >>> pythoncom.CreateGuid <<< to get a global unique
> identifier.)
> 
> The GUIDs use the MAC adress of the machine they are created at in
> their algorithm. I start to think that this can not be done in python,
> but I still hope that I am wrong in that point.


use os.popen('/sbin/ifconfig -a','r') and regexp match the output for
mac addresses.

-- 
"Reality is that which, when you stop believing in it, doesn't go away".
                -- Philip K. Dick


From howard at eegsoftware.com  Wed Dec 13 12:02:28 2000
From: howard at eegsoftware.com (howard at eegsoftware.com)
Date: Wed, 13 Dec 2000 17:02:28 GMT
Subject: How to use BLT2.4 and Tcl/tk8.3 under Windows9X/NT ?
References: <39F6BBCD.C25D4FEA@mailux.berl.tek.com> <3a02381b.581985@news-server.socal.rr.com> <3A36CD0C.8B79DD33@araidesign.com>
Message-ID: <3a37aad4.3737777@news-server.socal.rr.com>

On Tue, 12 Dec 2000 17:12:45 -0800, Chris Arai <chris at araidesign.com>
wrote:


>Howard,
>
>Did you ever resolve this problem??  If you did I would like to know how, as
>I am having the same problem.
>

Well, I worked around the problem (temporarily) by copying the
BLT24.DLL file to the SAME directory as the Python 2.0 executable.

The problem appears to be that moving tcl to a subsidiary directory of
Python fixes the Tkinter version problem but breaks the Pmw loader
process which apparently does not update the path needed to find the
BLT dll.   Note:  tcl 8.3 uses the old-style loadLibrary function
which does not have the option of providing additional path
information.


Howard Lightstone
EEGSoftware
howard at eegsoftware.com


From root at [127.0.0.1]  Sun Dec 10 14:39:15 2000
From: root at [127.0.0.1] (deadmeat)
Date: Sun, 10 Dec 2000 19:39:15 GMT
Subject: pattern matching
Message-ID: <DZQY5.7555$KY1.20484@news1.rivrw1.nsw.optushome.com.au>

Is there another standard library I can use for pattern matching, other than
fnmatch? I need to be able to check for metacharacters, fnmatch doesn't
allow me to escape them...





From qrczak at knm.org.pl  Fri Dec 29 04:05:05 2000
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 29 Dec 2000 09:05:05 GMT
Subject: Erlang (was Re: Good introduction to functional programming with Python?)
References: <mailman.977984412.30211.python-list@python.org> <92f0il01a5u@news1.newsguy.com> <86wvckqw4h.fsf@localhost.my.domain>
Message-ID: <slrn94oku1.s27.qrczak@qrnik.zagroda>

28 Dec 2000 09:43:58 -0800, Brad Knotwell <knotwell at ix.netcom.com> pisze:

> FWIW, I believe the Glasgow Haskell compiler has concurrent
> extensions as well, but I've no idea if they're delivered with the
> runtime environment or not.

They are. For about two years there is no separate concurrent build,
and concurrency is always available. (Threads are managed by ghc
itself, not by the OS.)

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


From j.l.gomez-dans at sheffield.ac.uk  Wed Dec 13 06:27:08 2000
From: j.l.gomez-dans at sheffield.ac.uk (=?iso-8859-1?q?Jos=E9_Luis_G=F3mez?= Dans)
Date: 13 Dec 2000 11:27:08 +0000
Subject: Tutorial in Numerical Python?
Message-ID: <87k8947e9f.fsf@sheffield.ac.uk>

Hi,
        I am wondering whether such a thing exists. Is there a nice
tutorial for NumPy that guides you through the basic steps? Say, a
little hands-on introduction to Linear Algebra, Transforms and
possibly the inclusion of some of the graphic modules?

        There are some willing students to be converted around these
parts, and a hands-on tutorial would be great. Unfortunately, I am a
beginner in these waters myself, but can see the opportunity for this.

        Thanks in advance,
Jos?
-- 
Jos? L G?mez Dans			PhD student
Tel: +44 114 222 5582			Radar & Communications Group
FAX; +44 870 132 2990			Department of Electronic Engineering
					University of Sheffield UK


From olafb+usenet at pvv.org  Thu Dec 14 05:34:45 2000
From: olafb+usenet at pvv.org (Olaf Trygve Berglihn)
Date: 14 Dec 2000 11:34:45 +0100
Subject: rpmmodule.so and python 2.0
Message-ID: <dn1hf47mgu2.fsf@verden.pvv.ntnu.no>

I've compiled rpmmodule against the new python api. But when I try to
user the interface to the rpm database, I get the following error:

Python 2.0 (#1, Dec 13 2000, 12:05:47)
[GCC 2.95.2 19991024 (release)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> import rpm
>>> db = rpm.opendb( 0, '/' )
cannot open file //var/lib/rpm/nameindex.rpm: Invalid argument
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
rpm.error
>>>


But if I rebuild the database, it works:

>>> rpm.rebuilddb( '/' )
cannot open file //var/lib/rpm/nameindex.rpm: Invalid argument
0
>>> db = rpm.opendb( 0, '/' )
>>> db.findbyname( 'bash' )
[3450392]
>>>


If I now exit to the shell:

$ rpm -q bash
cannot open file //var/lib/rpm/nameindex.rpm: Invalid argument
rpmQuery: rpmdbOpen() failed
$


But if I rebuild:

$ rpm --rebuilddb
cannot open file //var/lib/rpm/nameindex.rpm: Invalid argument
$ rpm -q bash
bash-2.03-11tr
$


Any suggetions why this is so, and how to fix it? I've recompiled the
rpmmodule against python 2.0 libraries.

Olaf
-- 
Olaf Trygve Berglihn <olafb+usenet at pvv.org>


From moshez at zadka.site.co.il  Mon Dec 18 23:11:01 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Tue, 19 Dec 2000 06:11:01 +0200 (IST)
Subject: while true: !!!
In-Reply-To: <slrn93snve.tvg.grey@teleute.rpglink.com>
References: <slrn93snve.tvg.grey@teleute.rpglink.com>, <B43D149A9AB2D411971300B0D03D7E8B0FBCBB@natasha.auslabs.avaya.com> <v04220801b663769e1800@[192.168.11.33]> <mailman.977147642.18708.python-list@python.org> <91lclt02qr3@news1.newsguy.com>
Message-ID: <20001219041101.8B3B5A84F@darjeeling.zadka.site.co.il>

On Mon, 18 Dec 2000 19:05:18 -0000, grey at despair.rpglink.com (Steve Lamb) wrote:
>     I have tough times in C, Pascal and other such
> languages because of one consideration they make that Perl and Python do not.
> Data is data
....
[in Python]
> It's a string, not a string[80] that I might one day
> want to stuff 81 characters into.  :/

I think you're dissing C unfairly here. If you deal a lot with strings,
why not pick up a good library, like glib, to help you. Glib will
make C seem almost like Python <wink>

-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From fredrik at effbot.org  Wed Dec 13 12:32:02 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 13 Dec 2000 17:32:02 GMT
Subject: Traceback and source code
References: <slrn93f68c.m2r.jajvirta@sirppi.helsinki.fi>
Message-ID: <moOZ5.3824$Kd1.435556@newsb.telia.net>

Jarno J Virtanen wrote:
> I'd be interested to know how the traceback module is syncronized
> with the actual source code (file).

It's cached by the "linecache" module.

To reset the cache, use linecache.clearcache()

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From ivnowa at hvision.nl  Mon Dec 11 09:48:11 2000
From: ivnowa at hvision.nl (Hans Nowak)
Date: Mon, 11 Dec 2000 14:48:11 GMT
Subject: Modifying running code? [HACK ALERT]
References: <912m65$pu1$1@news.rz.uni-karlsruhe.de>
Message-ID: <3A34E7B0.495@hvision.nl>

Bjoern Giesler wrote:
> 
> Hi,
> 
> I'm writing a kind of mini runtime environment that lets me debug / correct
> buggy Python code "on the fly". This is supposed to work as follows:
> 
> 1) an exception gets thrown by the buggy code and is caught by a toplevel
> exception handler
> 
> 2) the toplevel handler figures out the function that caused the exception
> 
> 3) the function source is loaded into an editor and can be corrected by the
> user
> 
> 4) if the code has been saved, it is recompiled and inserted instead of the
> original function
> 
> 5) the program is then resumed at the point where the original function had
> been called
> 
> Can this be done in Python at all? Can it be done in regular Python or only
> in Stackless Python (which supports continuations and would therefore, I
> believe, make step 5 easy)? Is the whole thing complete bull doodoo? Any
> hints or suggestions?
> 
> Thanks in advance for any input,
>                 --Bj?rn

This sure sounds like an interesting task... I'm not sure if it's
possible... the reference manual says,

"Python uses the ``termination'' model of error handling: an exception
handler can find out what happened and continue execution at an outer
level, but it cannot repair the cause of the error and retry the failing
operation (except by re-entering the offending piece of code from the
top)."

I think/hope there are ways around this, like wrapping every function
that can be debugged in a wrapper class that catches the exception, and
can recall the wrapped function... not very convenient if you have a lot
of functions like this... if it's possible at all. Editing the code will
be a problem too; you will need to point the program to a string or text
file that holds the function code, unless you're planning to edit
bytecode. :(

--Hans Nowak


From loewis at informatik.hu-berlin.de  Fri Dec  8 15:53:27 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 08 Dec 2000 21:53:27 +0100
Subject: SSL how to?
References: <yZ3Y5.3016$9B.89007@news1.oke.nextra.no>
Message-ID: <p6qu28eiqig.fsf@informatik.hu-berlin.de>

"Syver Enstad" <syver at NOSPAMcyberwatcher.com> writes:

> Anyone know if SSL is painless to use under python. Does httplib support it?

It's not painless. It is only available if it was selected at compile
time, the standard Windows distributions don't include it, and it
won't compile on Windows out-of-the-box (although there are patches).

Yes, httplib supports SSL in Python 2.

Regards,
Martin



From ssthapa at harper.uchicago.edu  Fri Dec  8 11:51:02 2000
From: ssthapa at harper.uchicago.edu (Suchandra Thapa)
Date: Fri, 08 Dec 2000 16:51:02 GMT
Subject: parallel port access?
References: <3A268378.4DA1B831@sympatico.ca> <3A272881.6F444750@engcorp.com> <3A2D6761.814ABE8E@bioeng.ucsd.edu> <slrn92su1q.12g.ssthapa@localhost.localdomain> <udhf4gk079.fsf@box.home.de>
Message-ID: <slrn92vbu2.2jp.ssthapa@localhost.localdomain>

Jochen K?pper <jochen at pc1.uni-duesseldorf.de> wrote:
>>>>>> Suchandra Thapa wrote on Thu, 07 Dec 2000 08:26:19 GMT:
>
>Well, on many unices there is the common idea of a printer beeing a
>file:-) This means you have a driver sitting behind it, and simply
>writing to that file sends your data to the port.
>Direct access to the ports in general requires kernel-level priorities
>or root-access and going through ioperm or such...

    Although, the printer may be  just a file on unices, 
I see generic access to the parallel port as being more than just printing.
I take generic access to the parallel port to mean that you are able to
change the status of individual lines so you can interface with 
various devices that don't follow the standard printer methods (I'm not
sure this exists witht the variety of printers out there).  If I were
to use the parallel interface from the python library, I would certainly
hope for this type of functionality since printing to files can be done
without the interface by opening a connection to lptx on windows/nt or
opening a pipe to lpr on unix.
    Also a printer does not correspond to a parallel port.  The printer
may in actuallity be a networked printer down the hall or something
in another location entirely.  
    
>I don't see a need to write a driver for each platform, only DOS
>probably, which doesn't bring one by itself [1].
>
>So I would say you'll need a generic wrapper function for sending data
>to the OS, that's it. On Unix I'd say all the problem reduces to find
>out wether there is a printer-file and what it is called.
>
>[1] Then a parport is such a simple device, for most applications you
>really don't need a real driver...

    It depends on what mode you are using and what device you are communicating
with.  If you look at timing diagrams for ECP or EPP operations, there
are fairly strigent requirements on the how quickly certain lines have to 
change.  E.g. if you raise line 1 (ACK) to +5 V to indicate a data transfer 
then you must raise lines 2-9 (data) to the values being transfered
within x milisecs or the device will raise one or more of the control lines 
to indicate an error and you need to deal with that then.  Also devices
like scanners and zip drives have more complex protocols than printers.

    For windows nt a kernel driver must be written since the os doesn't
provide the required functionality and you can't go around the os and 
access the port directly.  With windows 98, you can access the hardware but 
it's probably better to write a driver and use that. I'm not sure what the
requirements are for various unixes, but I suspect a driver would be required
to get the level of control I've indicated.

-- 
------------------------------------------------------------------
			    |
Suchandra Thapa             | "There are only two kinds of math books. 
s-thapaNO at SPAMuchicago.edu  | Those you cannot read beyond the first 
			    | sentence, and those you cannot read 
			    | beyond the first page."
			    |                       -C.N. Yang
------------------------------------------------------------------


From nospam at no.spam  Wed Dec 20 19:26:22 2000
From: nospam at no.spam (Carsten Geckeler)
Date: Thu, 21 Dec 2000 01:26:22 +0100 (MET)
Subject: Extracting attributes from modules
In-Reply-To: <91rej7$td7$1@nnrp1.deja.com>
Message-ID: <Pine.LNX.4.10.10012210122320.1436-100000@erde.kosmos.all>

On Wed, 20 Dec 2000 plowboy_lifestyle at my-deja.com wrote:

> Hi,
>   I use python as an embedded scripting tool in the 3D
> modeling application called Blender and I'm looking for some
> clarification from a pro.
> I have this difficulty when I try learn about the various features of
> the embedded API from my scripts. A typical blender script looks
> something like this:
> 
> import Blender
> from Blender import *
> 
> print dir()
> 
> this program is supposed to give me a list of the objects in the name
> space including everything from the module blender. The output looks
> something like this (hopefully I can explain all this without being
> application specific):
> 
> ['Blender','Camera','Object','World',...etc.]
> 
> so I explore further with
> 
> print dir(Object)
> 
> and I get:
> 
> ['Get','GetSelected','Update','__doc__','__name__']
> 
> if I use the get function I find out that it returns an object from
> within the modelling program, like a sphere or a mesh, whatever I'm
> using and I can manipulate it by setting various variable like "LocX" or
> "LocY" for the various locations of the 3D objects, like so:
> 
> mySphere = Object.Get('theSphere')
> mySphere.LocX = mySphere.LocX + 5
> 
> this program will have the effect of moving the object in space positive
> 5 units on the x axis.
> 
> so this is my actual question; Even though I can find out about various
> attributes like LocX and etc. from the manuals for this application,
> there is no way for me to find them from python it seems, for example:
> 
> print dir (Object.Get)
> print dir (mySphere)
> print dir (mySphere.LocX)
> print Object.Get.__doc__
> print mySphere.__doc__
> print dir(type(mySphere))
> 
> all output an empty list or "none", even though
> 
> print myShere.LocX
> 
> works fine at printing out a floating point number.
> I'm only trying to get a list of names of attributes of the object but
> they seem hidden. Obviously I'm new to python. Can someone explain why
> this works this way?

Their probably not hidden, but mySphere is "only" an instance, so there
may be attributes or methods of the class or the base classes.  These
attributes don't show up with the dir() command.  Try the following:

print dir(mySphere)
print dir(mySphere.__class__)
for bc in mySphere.__class__.__bases__:
	print dir(bc)

Cheers, Carsten
-- 
Carsten Geckeler:  carsten dot geckeler at gmx dot de
To get proper email-address replace `dot' and `at' by the corresponding symbols.




From moshez at zadka.site.co.il  Mon Dec 25 09:34:44 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: 25 Dec 2000 14:34:44 -0000
Subject: another newbie question - regex
In-Reply-To: <slrn94eju1.1fk.jp@localhost.localdomain>
References: <slrn94eju1.1fk.jp@localhost.localdomain>
Message-ID: <20001225143444.30585.qmail@stimpy.scso.com>

On 25 Dec 2000, jp at ulgo.koti.com.pl (Jacek) wrote:

> "." - matches any character
> "*" - is 0 or more repetition 
> 
> so ".*" - matches everything?

Yep. And that's the common idiom for "everything too.".

> >>> print re.search(s,".*")
> None
> >>> print re.match(s,".*")
> None

Try

re.search(".*", s)
re.match(".*", s)

Your arguments were backwards.

In general, using re.compile will generate code that is (usually) both
clearer and faster.

everything = re.compile(".*")
everything.search(s)
everything.match(s)



From un4e at rzstud1.rz.uni-karlsruhe.de  Fri Dec 15 04:54:56 2000
From: un4e at rzstud1.rz.uni-karlsruhe.de (Bjoern Giesler)
Date: 15 Dec 2000 09:54:56 GMT
Subject: Getting a code obj's function?
References: <91co73$fpm$1@news.rz.uni-karlsruhe.de> <pBl_5.3966$O%.626490@newsc.telia.net>
Message-ID: <91cppg$fpm$2@news.rz.uni-karlsruhe.de>

Hi,

Fredrik Lundh <fredrik at effbot.org> wrote:
: Bjoern Giesler wrote:
:> ...is there a standard way to get the function object that a code object is
:> associated with? On exception, I only get the code object from the
:> traceback.

: Use co_name:

no, this doesn't work... I need the function _object_, not its _name_. I
can't use the name to find the object, since the following...

[snip]
import sys
import exceptions

def spam():
    def eggs():
        raise exceptions.Exception
    eggs()

try:
    spam()
except:
    tb = sys.exc_traceback
    while tb.tb_next:
        tb = tb.tb_next
    code = tb.tb_frame.f_code
    print code.co_name
[snip]

...produces "eggs", and not "spam.eggs" (as would be better, IMHO, but alas,
it doesn't).

I need the function object because I want to replace the function object's 
code object. :-)

Regards,
		--Bj?rn


From slhath at home.com  Tue Dec 19 00:28:41 2000
From: slhath at home.com (Scott Hathaway)
Date: Tue, 19 Dec 2000 05:28:41 GMT
Subject: xml-rpc module question
Message-ID: <dmC%5.75166$yR4.2151642@news1.rdc1.tx.home.com>

Hello,

I have taken the xml-rpc module and added the following to the call area:

def CubeIt(x):
   return x**3

def call(self, method, params):
   if method == 'CubeIt':
   return CubeIt(params)

And, I call this with the following client call:

import xmlrpclib as x
testing = x.Server('http://localhost')
print testing.CubeIt(3)

And I get the following error:

C:\Projects\core>python core_client.py
Traceback (most recent call last):
  File "core_client.py", line 7, in ?
    print testing.CubeIt(3)
  File "xmlrpclib.py", line 547, in __call__
    return self.__send(self.__name, args)
  File "xmlrpclib.py", line 630, in __request
    request
  File "xmlrpclib.py", line 585, in request
    return self.parse_response(h.getfile())
  File "xmlrpclib.py", line 601, in parse_response
    return u.close()
  File "xmlrpclib.py", line 371, in close
    raise apply(Fault, (), self._stack[0])
xmlrpclib.Fault: <Fault 1: "exceptions.NameError:There is no variable named
'CubeIt'">

Can anyone help me out?  I am sure that I am not creating the CubeIt
function properly.  I am totally lost.

Thanks,
Scott





From dsh8290 at rit.edu  Sat Dec 30 23:51:18 2000
From: dsh8290 at rit.edu (D-Man)
Date: Sat, 30 Dec 2000 23:51:18 -0500
Subject: Regexp speed (was Re: split hangs)
In-Reply-To: <14C1lH-44b-00@mercurio.localdomain>; from davide@mercurio.localdomain on Fri, Dec 29, 2000 at 04:48:07PM +0100
References: <mailman.977998605.7293.python-list@python.org> <14C1lH-44b-00@mercurio.localdomain>
Message-ID: <20001230235118.C19500@westsidecnc.rh.rit.edu>

Try reading the suggested "Mastering Regular Expressions" book by Jefferey Freidl (published by O'Reilly I think) There are 2 main types of regex engines: NFA and DFA.  I forget exactly what they stand for, but "N" is for non-deterministic and "D" is for deterministic.

NFA's can be faster and more flexible (like allowing backreferences)
but can be slow if the regex isn't crafted right (see your example).
A DFA, however, is faster if there is a lot of text and there is no
match.  

Python (and Perl, etc) use NFA's.  NFA's start with the given text
(your file) and starts to match it.  It starts with the first
character and works 1 character at a time until a char doesn't match.
Then it backs up until it can match that character or it determines no
match.  If it finds no match it will move ahead 1 character (from the
first one in this attempt) and try again.  If you have no match at all
in a long document (especially if you use repetition in the regex)
this will be very slow.  (Also note that repetition is usually greedy,
ie the regex "/\*.*\*/" to match C comments will match your entire
file from the first /* to the last */.

DFA's (it looks like emacs uses one) instead keeps track of all
possible ways of matching the current text simultaneously.  This means
that if there is a match, it is slower than an NFA with the properly
crafted regex.  Also, DFA's don't allow backreferences.  If there is
no match in a large piece of text the DFA will be faster.


I would strongly suggest reading Jefferey Friedl's book.  I have it
(though I haven't read the entire book) and found it to be very useful
and enlightening.  He will explain in much more completely and
accurately the operation of NFA and DFA regex engines and also how to
craft regexes that are correct and efficient.

HTH,
-D

On Fri, Dec 29, 2000 at 04:48:07PM +0100, davide at mercurio.localdomain wrote:
> In comp.lang.python, you wrote:
> 
> >The tail end of the output is clear:  adding another character to the string
> >doubles the time it takes to figure out that the regexp doesn't match.  This
> >isn't a bug, it's the way Python (and Perl, and Emacs, and ...) regexps
> >work.  A full explanation is quite involved; you can find one in Jonathan
> >Friedl's excellent book "Mastering Regular Expressions" (O'Reilly).
> 
> I tried the equivalent regexp in emacs
>   <\([^<>"]+\("[^"]*"\)?\)+>
> with bigger and bigger files and it does not seem to give an exponential
> behaviour.
> 
> In everything I read regexp matching is given as a linear time problem in
> n + m (where n is the size of the testing string and m the size of the
> regular expression).

How do you measure the size of a regex?  Is ".*" inifinite or
2?

> 
> I am a python newbye and I could not find the reason for this exp 
> behaviour. The regular expressions seems pretty standard to me, 
> and so there should be no need to go beyond regular languages and DFA. 
> 
> Would you explain why this algorithm is needed in python?
> Do really the other implementation (grep, perl, emacs, ecc...)
> use the same algorithm?
> 
> Thanks,
> 
> 				Davide



From tgagne at ix.netcom.com  Wed Dec 20 22:09:51 2000
From: tgagne at ix.netcom.com (Thomas Gagne)
Date: Wed, 20 Dec 2000 22:09:51 -0500
Subject: Catching Database Exceptions (Sybase)
References: <3A40EEBF.3262040F@ix.netcom.com> <ulmtahl1h.fsf@ctwd0143.fitlinxx.com>
Message-ID: <3A41747F.B0980784@ix.netcom.com>

except Sybase.Warning, env:

That was it!  I needed to specify the module name before "Warning" and
"Error".  Not I'm catching the exception.  It worked for *both* the missing
table and the missing stored procedure.

As a side note, all errors and warning detected in the database are reported
back to the client application as either Messages or Errors.  Now I'm going to
see where the text of RAISERROR and PRINT statements ends up...

David Bolen wrote:

> Thomas Gagne <tgagne at ix.netcom.com> writes:
>
> > And miraculously, even with the table missing, it ran the "else".
> >
> > I must be missing something significant here.  Any help?
>
> You might try the same operation via some other tool (e.g., ISQL or
> equivalent).  It's my understanding from some of the DB folks here
> that depending on the type of error, errors within stored procedures
> may sometimes be simply absorbed and not reported - you just get back
> an empty result.  I know I've run some stored procedures that had bad
> references to fields that didn't exist and didn't actually get a
> database error from the execution, just no result.
>
> And to answer your follow-up post question, the Warning and Error
> classes that the DB API specifies all exceptions must be subclassed
> from are themselves DB API-specific exceptions (both are subclasses of
> the standard Python exception StandardError).
>
> So in your exception clauses try using "<module>.Warning" instead of
> just "Warning" (where <module> is your DB API compliant module).

--
.tom





From etsang at my-deja.com  Thu Dec  7 19:27:14 2000
From: etsang at my-deja.com (etsang at my-deja.com)
Date: Fri, 08 Dec 2000 00:27:14 GMT
Subject: How to keep a windows frame on top??
Message-ID: <90p9ss$chi$1@nnrp1.deja.com>

This is the function for calling my window. Can anyone tell how can I
make the windows to always stay on top when called?

def resultsData(self):

        okButton = None

        self.inputWindow = Toplevel(self.master)
        self.inputWindow.geometry("+%d+%d" % (self.master.winfo_rootx()
+180,
                                  self.master.winfo_rooty()+180))
        self.inputWindow.title("Enter data")
        Label(self.inputWindow, text ="Please enter your initials and
the test object").grid(row=0, columnspan=2, pady=6)
        Label(self.inputWindow, text="Initials").grid(row=1, sticky=W)
        Label(self.inputWindow, text="Test object").grid(row=2,
sticky=W)
        self.e1 = Entry(self.inputWindow, textvariable=self.name,
width=35)
        self.e2 = Entry(self.inputWindow, textvariable=self.iut,
width=35)
        self.e1.grid(row=1, column=1, sticky="EW", pady=1)
        self.e2.grid(row=2, column=1, sticky="EW", pady=1)
        okButton = Button(self.inputWindow, text="Ok",
command=self.getData)
        okButton.grid(row=3, column=1, sticky=W, pady=6)
        if self.automatic == FALSE:
	   self.inputWindow.wait_window()
	else:
	   self.getData()


Sent via Deja.com http://www.deja.com/
Before you buy.


From teyc at bigfoot.com  Tue Dec 19 17:26:07 2000
From: teyc at bigfoot.com (Chui Tey)
Date: Wed, 20 Dec 2000 08:26:07 +1000
Subject: Python interpreter in python?
References: <mailman.977254923.336.python-list@python.org>
Message-ID: <91ond7$kor$1@bunyip.cc.uq.edu.au>

I was planning play around with the ceval.c to get the python debugger to
reexecute certain lines (after stepping past them), which is very handy when
debugging tortuous routines in Zope. If there is a bytecode machine in
Python, we can make it so that it will play with pdb. Then we can prototype
changes to ceval in python first before digging around in c code.

Chui

Steven D. Majewski <sdm7g at virginia.edu> wrote in message
news:mailman.977254923.336.python-list at python.org...
>
> On Tue, 19 Dec 2000, D-Man wrote:
>
> >
> > The problem with such a thing is that to run your interpreter you
> > would need an existing interpreter (written in C/Java or other
> > language).  The only languages that can be implemented in themselves
> > are compiled languages like C and C++.
> >
>
> That is absolutely untrue!
>
> Lisp and Scheme are frequently implemented in themselves -- a
> 'metacircular interpreter' is a fairly typical student exercise.
>
> I don't know of a Python-in-Python, but I don't know of any reason
> you couldn't do one.
>
> The only question is: how deeply do you want to emulate it? i.e.
> are you allowed to use the existing parser module or do you want
> to write one from scratch?
>
> -- Steve Majewski <sdm7g @ Virginia.EDU>
>
>
>
>




From peter at engcorp.com  Fri Dec 29 09:34:48 2000
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 29 Dec 2000 09:34:48 -0500
Subject: Python vs. PHP (& Java?)
References: <69BC4C6AE83ED311BC9400805FA7B14AF84D2C@il93exp01.css.mot.com> <mailman.978059571.14515.python-list@python.org> <3A4C89AA.7E741D91@stroeder.com>
Message-ID: <3A4CA108.503B4DE5@engcorp.com>

Michael Str?der wrote:
> > On Thu, 28 Dec 2000, Kemp Randy-W18971 <Randy.L.Kemp at motorola.com> wrote:
> > > 2.  Could we argue that an artificial language, such as Esperanto, is better
> > >     to use around the world then a natural language, like English?
> 
> <OFFTOPIC>
> Esperanto is pretty much derived from romanian languages. Rather
> hard to learn for many people. How about http://www.lojban.org ?
> </OFFTOPIC>

Tee-hee!  That rather ignores all empirical evidence from the last 113
years or so, now doesn't it?

By the way, you probably meant "Romance" languages, not "romanian".  Which
is another way of saying "Latin-derived", mostly, and is a somewhat
truthful claim when applied to Esperanto, although I always felt it misled
more than explained, given Esperanto's grammar.

Anyway, not that number of speakers demonstrates ease of learning in any
way, but how many people fluently speak Loglan/Lojban?  Had it reached ten
yet?


From Daniel.Kinnaer at AdValvas.be  Mon Dec 18 12:01:33 2000
From: Daniel.Kinnaer at AdValvas.be (Daniel)
Date: Mon, 18 Dec 2000 17:01:33 GMT
Subject: search for pattern in StringList
References: <3a3e1935.5461298@news.skynet.be> <91lap002od4@news1.newsguy.com>
Message-ID: <3a3e4184.15782077@news.skynet.be>

On Mon, 18 Dec 2000 16:33:23 +0100, "Alex Martelli"
<aleaxit at yahoo.com> wrote:

>"Daniel" <Daniel.Kinnaer at AdValvas.be> wrote in message
>news:3a3e1935.5461298 at news.skynet.be...
>> As a Python newbie, I was trying a very simple exercise, but got
>> stuck. Is there someone who can help me complete this code, please?
>>
>> Thanks.
>> Daniel
>>
>>
>> MyList = ['an','been','call','door']
>>
>> print "Scanning a StringList"
>>
>> for i in range(0, len(MyList)):
>>     print i, MyList[i]
>
>Note that
>    for i in range(len(MyList)):
>is equivalent (0 is the default starting value for range).
>
>> #try to show every item in MyList which contains an 'a'
>> #that is 'an' and 'call'.   In Pascal :
>> #for i:=0 to MyList.Items.Count -1 do
>> #    if pos('a',MyList.Items[i])>0 then writeln(MyList.Items[i])
>> #in Python : ???
>
>In Python, it can be a little bit more "conceptually concise",
>thanks to the for-in statement (one can iterate on the
>items of a sequence directly, rather than iterating on
>the indices and using indexing in the body, when the
>index itself isn't needed):
>
>for item in MyList:
>    if item.find('a')>0:
>        print item
>
>
>It's not *necessary* to break the 'if', since its body
>is a single, simple statement -- you *could* also write:
>
>for item in MyList:
>    if item.find('a')>=0: print item
>
>but the normal Python convention is to end the line at
>any ':' (which concludes such statements as for, while,
>if, def, ... -- 'compound statements', basically), and
>then (this is not just convention, but a key Python
>lexical rule) you have to indent the statement's body
>uniformly.
>
>
>Alex
>
>
>
Thanks guys!  Indeed, Python seems to be a powerful and  'easy'
language (if you know how :-)

Daniel



From pshinners at mediaone.net  Thu Dec  7 02:35:16 2000
From: pshinners at mediaone.net (Pete Shinners)
Date: Thu, 07 Dec 2000 07:35:16 GMT
Subject: problems with installer 20_3i
References: <_o%V5.47439$yR4.1041961@news1.rdc1.tx.home.com> <8FFED7BE9gmcmhypernetcom@199.171.54.155>
Message-ID: <U4HX5.2776$jt2.260296@typhoon.we.rr.com>

"Gordon McMillan" <gmcm at hypernet.com> wrote
> Nothing. Barry Scott just pointed out to me that the Python 2.0 imputil
> does not by default install any method of importing builtins (even though
> it installs, by default, a file system importer).
>
> Untested: at the top level of archive_rt (after the imports), put:
>
> sys.path.append(imputil.BuiltinImporter())

i've been trying to get this going with my own stuff.
i used the "Freeze.py" script to build my EXE. i made this change
to archive_rt.py. i'm no longer getting the problem with the 'time'
module, but having trouble using one of my extension modules.

"ImportError: no module named pygame.base"

i looked in the "Builder.log", and at the bottom, this module
is definitely getting put into the EXE...

Final Table of Contents for aliens.exe:
[('win32api', 'c:\\python\\win32\\win32api.pyd', 'b'),
 ('pygame.key', 'c:\\python\\pygame\\key.pyd', 'b'),
 ('pygame.cdrom', 'c:\\python\\pygame\\cdrom.pyd', 'b'),
 ('pygame.surface', 'c:\\python\\pygame\\surface.pyd', 'b'),
 ('pygame.display', 'c:\\python\\pygame\\display.pyd', 'b'),
 ('PyWinTypes20.dll', 'C:\\WIN\\SYSTEM\\PyWinTypes20.dll', 'b'),
 ('pygame.rect', 'c:\\python\\pygame\\rect.pyd', 'b'),
 ('pygame.mixer', 'c:\\python\\pygame\\mixer.pyd', 'b'),
 ('pygame.image', 'c:\\python\\pygame\\image.pyd', 'b'),
 ('pygame.joystick', 'c:\\python\\pygame\\joystick.pyd', 'b'),
 ('pygame.font', 'c:\\python\\pygame\\font.pyd', 'b'),
 ('pygame.constants', 'c:\\python\\pygame\\constants.pyd', 'b'),
 ('pygame.event', 'c:\\python\\pygame\\event.pyd', 'b'),
 ('pygame.time', 'c:\\python\\pygame\\time.pyd', 'b'),
 ('pygame.base', 'c:\\python\\pygame\\base.pyd', 'b'),
 <<<snip>>>


this whole installer is still a 'black-art' to me, so it's tough
for me to figure out any problems.

in need of clues :]





From nicmila at idoox.com  Tue Dec  5 02:53:14 2000
From: nicmila at idoox.com (Miloslav Nic)
Date: Tue, 05 Dec 2000 08:53:14 +0100
Subject: Converting Perl Regex to Python!
References: <90h0a8$jgm$1@nnrp1.deja.com>
Message-ID: <3A2C9EEA.1EE1C2D9@idoox.com>

Examples in Zvon Regular expression reference:
http://zvon.org/other/reReference/Output/index.html

 are generated with Python, on the basic level with the re module 
there is no difference

thelocust at my-deja.com wrote:
> 
> I'm working on porting a program from Perl to Python, and i'm to a
> point where I need to convert their Perl regex stuff over to python.
> Now, I understand what is going on in the Perl if statement below, but
> i haven't seen to many good examples of regex stuff for Python, so i'm
> hurting for an translation
> 
> The string i need to grab the info from looks like this:
> 
> MSG notify timecode [00:00:00:00]
> 
> if ($msg =~ /^MSG notify timecode \[(\d\d:\d\d:\d\d):\d\d\]/) {
>     $cajun->{state_clock}=$1;
>     }
> 
> So, how exactly do i convert this to python?
> 
> Sent via Deja.com http://www.deja.com/
> Before you buy.

-- 
******************************************
<firstName> Miloslav </firstName>    
<surname>   Nic      </surname>     

<mail>    nicmila at idoox.com    </mail>   
<support> http://www.zvon.org  </support>
<zvonMailingList> 
    http://www.zvon.org/index.php?nav_id=4 
</zvonMailingList>


From paul.moore at uk.origin-it.com  Tue Dec 19 07:49:23 2000
From: paul.moore at uk.origin-it.com (Paul Moore)
Date: Tue, 19 Dec 2000 13:49:23 +0100
Subject: Events in win32com
Message-ID: <Y1g=Op6IfWSoJn1ISFtOyuvqwBb3@4ax.com>

If I create an object using

obj = win32com.client.Dispatch(progid)

can I see events from the object? In WSH the CreateObject method has
an extra parameter which is the prefix for subroutine names which are
called on an event, something like

ie = WScript.CreateObject("InternetExplorer.Application", "ie_")

Sub ie_OnQuit()
End Sub

While I don't like the "specially named subroutine" approach, I do
need to be able to sink events in Python. Is this possible, or should
I resort to using Python via WSH?

(BTW, assuming I can sink events, will Python handle ByRef event
parameters - for example, a Cancel parameter which is set to True in
the handler to stop the event occurring).

Thanks,
Paul Moore



From jasonic at nomadicsltd.com  Sat Dec  9 12:18:15 2000
From: jasonic at nomadicsltd.com (Jason Cunliffe)
Date: Sat, 9 Dec 2000 12:18:15 -0500
Subject: corba, soap, com, etc...
References: <pKsY5.58762$yR4.1496678@news1.rdc1.tx.home.com>
Message-ID: <t34q6aso9mps3b@corp.supernews.com>

Scott

You may want to look at Pyro for rapid prototyping, inspiration and more..

http://www.xs4all.nl/~irmen/ap/pyro.html

<quote>
Pyro is an acronym for PYthon Remote Objects. It is a basic Distributed
Object Technology system written entirely in Python, and for use in Python
only. With this, it closely resembles Java's Remote Method Invocation (RMI).
It has less similarity to CORBA - which is a system- and language
independent Distributed Object Technology and has much more to offer than
Pyro or RMI. But Pyro is small, simple and free! However, read the
disclaimer and copyright notice.
</quote>

- Jason
___________________________________________________________
Jason CUNLIFFE = NOMADICS['Interactive Art and Technology']


Scott Hathaway <slhath at home.com> wrote in message
news:pKsY5.58762$yR4.1496678 at news1.rdc1.tx.home.com...
> Hello Everyone,
>
> I need some advice.  I want to write an api engine in python that will be
> used as the back end to a gui.  The gui will not be written in python.  I
> need a way to expose the api I create that is cross platform.  On Windows,
I
> would use COM.  Since I need cross platform, what are my options and what
> are the advantages and disadvantages.  I know about CORBA, but I do not
know
> if CORBA can be used from python.  I have also heard about SOAP, but is it
> only web based?
>
> If there is a corba or soap interface to python, can you please point me
to
> it?
>
> Your input is greatly appreciated!
>
> Thanks!
> Scott





From joconnor at cybermesa.com  Fri Dec 22 19:41:12 2000
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Sat, 23 Dec 2000 00:41:12 +0000
Subject: A Suggestion for Python Dictionary/Class
Message-ID: <3A43F4A8.5CE531C@cybermesa.com>

> Therefore, is there any way for me to define a class, with some definite
> members just like in C/C++/Java, so that no new members can be defined? 


You might want to define all your instance variables in __init__() and
the overload __setattr__ to not allow anyone to set a value unless
already defined

class MyClass:
	def __init__(self):
		self.member = 0

	def __setattr__(self, name, value):
		if self.__dict__.has_key(name):
			self.__dict__[name]=value
		else:
			raise: "Error - no instance variable named "+ name

This will flag a runtime error if you try

o = MyClass()
o.menber = 1


Take care,

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

"God himself plays the bass strings first when He tunes the soul"



From SBrunning at trisystems.co.uk  Wed Dec  6 10:32:42 2000
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Wed, 6 Dec 2000 15:32:42 -0000 
Subject: How can i use a DLL in python ?
Message-ID: <31575A892FF6D1118F5800600846864D5B1557@intrepid>

> From:	^Socrates^ [SMTP:joseamartin at venmex.com]
> I am trying to port the fdftk.dll  and fdftk.lib from adobe acrobat to
> work
> in python
 
I may be way off base here, but what is it that you want to do? If your
intention is to generate PDFs, then I suggest that you look at ReportLab
(<http://www.reportlab.com>) before you get too carried away with DLL
calls...

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From hanche at math.ntnu.no  Wed Dec 13 17:52:06 2000
From: hanche at math.ntnu.no (Harald Hanche-Olsen)
Date: 13 Dec 2000 23:52:06 +0100
Subject: Python-like compiled language
References: <913jr3$dmv$1@sylvester.vcn.bc.ca> <3A35835B.AAA92005@bioeng.ucsd.edu> <kuzoi178kg.fsf@lasipalatsi.fi> <3A36B288.FD0524DF@bioeng.ucsd.edu> <m37l552ntb.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <pco3dfslysp.fsf@thoth.home>

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

| Curtis Jensen <cjensen at bioeng.ucsd.edu> writes:
| 
| > If C/C++ isn't the closest, then what is?
| 
| The Lisps of the world are very similar in some ways (variables being
| bindings, dynamic typing, interactive, introspective (well, CL, at
| least)), but very different in others (syntax, the macro system).

Oh, but Common Lisp and Python do have one important syntax feature in
common:

- non-Lispers hate all those parentheses,
- non-Pythoneers hate that syntactically significant white space.

Apart from that, CL has CLOS, which gives you object orientation.
And all good CL systems can compile to native code.

not-to-mention-full-access-to-the-parser-ly y'rs,

-- 
* Harald Hanche-Olsen     <URL:http://www.math.ntnu.no/~hanche/>
- Yes it works in practice - but does it work in theory?


From aleaxit at yahoo.com  Fri Dec 22 05:23:48 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 22 Dec 2000 11:23:48 +0100
Subject: Events in win32com
References: <Y1g=Op6IfWSoJn1ISFtOyuvqwBb3@4ax.com> <FY0=OlZFLy71DtnK62u5TEPlhOts@4ax.com> <3A40060B.10504@ActiveState.com> <cYxAOmxi9FgNERgBo29zuCl5c2h1@4ax.com> <ik724tc2v3cpcqfbt7i1ookvaona0mk5ql@4ax.com> <3A418D02.2040302@ActiveState.com> <91t2mf02dob@news2.newsguy.com> <3A42C1F7.7040407@ActiveState.com>
Message-ID: <91va46025pf@news2.newsguy.com>

"Mark Hammond" <MarkH at ActiveState.com> wrote in message
news:3A42C1F7.7040407 at ActiveState.com...
    [snip]
> The module is _always_ created at makepy time, and references to all
> subordinate items created at that time.

You mean that makepy doesn't use bForDemand, or am I misreading
you?


> > But how do I instantiate and call genpy.Generator, or arrange for
> > it to be instantiated and called, to ensure the HTMLDocumentEvents2.py
> > gets built in the bForDemand setting...?  Is there a canonical/
> > recommended way...?
>
> gencache.GetClassForProgID() and GetClassForCLSIDID - however, the

But HTMLDocumentEvents2 doesn't have a ProgID -- it's an interface,
not a coclass, much less a creatable one.  So how to best find out
the CLSID (IID, really -- but it's still a GUID anyway, so I guess
that would work just as well) from the interface-name, which is
what one normally gets out of the docs?  Linear search through the
keys of CLSIDToPackageMap in the module...?  That's over 400
entries for MSHTML -- it could be a bit slow.  Would it be better
to search for it manually and hardwire it as a constant in one's
Python code?

> "EnsureModule" will already need to have been run.  There is no
> single-step generate-module-for-demand-and-get-a-specific-class.

I don't really see this as a problem in this context -- it's
pretty easy to roll up one's own.  Anyway, this version works
just fine (and thanks for your advice!):


import sys, win32com.client.gencache, pythoncom

def EnsureModuleForTypelibOfObject(obj, bForDemand=1):
    try:
        ti = obj._oleobj_.GetTypeInfo()
        clsid = ti.GetTypeAttr()[0]
        tlb, index = ti.GetContainingTypeLib()
        tla = tlb.GetLibAttr()
        return win32com.client.gencache.EnsureModule(
            tla[0], tla[1], tla[3], tla[4], bForDemand=bForDemand)
    except pythoncom.com_error:
            raise TypeError, "This COM object can not automate the makepy
process - please run makepy manually for this object"

ie = win32com.client.gencache.EnsureDispatch("InternetExplorer.Application")
ie.Navigate('c:/index.htm')
ie.Visible=-1

html_module = EnsureModuleForTypelibOfObject(ie.Document)

itf = win32com.client.gencache.GetClassForCLSID(
    '{3050F613-98B5-11CF-BB82-00AA00BDCE0B}')

class Handler(itf):
    def __init__(self, obj):
        self.seen = 0
        itf.__init__(self, obj)
    def Ononclick(self, ev):
        self.seen += 1
        ev = win32com.client.gencache.EnsureDispatch(ev)
        el = ev.srcElement
        print 'clic on %s(%s) "%s", at (%d,%d)' % (
            el.tagName, el.id, el.innerText,
            ev.clientX, ev.clientY)

h = Handler(ie.Document)

while h.seen < 3:
    pythoncom.PumpWaitingMessages()

ie.Quit()


The hard-wiring of '{3050F613-98B5-11CF-BB82-00AA00BDCE0B}' is
the only part of this that seems like a doubtful idea -- though
maybe it's actually better than it 'feels', since IID's *are*
'hardwired forever' in COM's architecture.  But it sure would
be nice to be able to say 'HTMLDocumentEvents2' (in the context
of the typelibrary in which it is thus named, i.e. the
html_module object in this case).  Maybe I'm just worrying
uselessly!  The variant...:

html_module = EnsureModuleForTypelibOfObject(ie.Document)

import time
start = time.clock()
itfName = 'HTMLDocumentEvents2'
for guid, name in html_module.CLSIDToPackageMap.items():
    if itfName==name:
        itfID = guid
        break
else:
    raise KeyError, "%s not found in %s" % (itfName, html_module)
stend = time.clock()

print "GUID found, time %f" % (stend-start)

itf = win32com.client.gencache.GetClassForCLSID(itfID)


seems to also work just fine, with less than 6 milliseconds
"wasted" looking for the interface's GUID (which happens
to be around the middle of the package map's 400 entries,
so this should be rather typical).  So one could just, worst
case, package this up as a name-to-GUID-within-module helper
function, say:

def name2GUID(module, classname):
    for guid, name in html_module.CLSIDToPackageMap.items():
        if class==name: return guid
    return None

and live happily ever after (without real performance-driven
need to actually provide a reverse-mapping data structure).


One thing I really don't understand is why the above script
doesn't work right any more if I comment out the statement:

        ev = win32com.client.gencache.EnsureDispatch(ev)

in Ononclick -- the srcElement &c properties of the event
object are then not found any more (the resulting exception
is correctly caught in upper levels, so the script runs
until three clicks have been recorded, but I don't get
the print at each click any more).  Shouldn't the "pure
dispatch" ev object be _equivalent_ to the genpy-covered
one which I get through this statement, at least for
such simple stuff as getting object-properties...?


> I would be more than happy for us to step back and look at the
> makepy/gencache exposed API - I agree it sucks, as it "just grew" and
> was really never designed for direct use by client code - it just
> happened that way.

That's the sign of successful software -- it grows and develops
beyond what its creator had originally conceived!  And win32com
most definitely IS successful -- it spans the powerful worlds of
COM and Python to deliver truly awesome power... a whole even
greater than the already-impressive sum of its parts.

I also fully agree that some refactoring of the API's exposed
to the application programmer would be great.  The HTML/XML
DOM's, and Office applications' object models, stress any
Automation framework's design parameters, being so huge and
rich (the bForDemand innovation was GREAT from this POV!).

Examining typical IE/HTML/XML use cases, we might come up
with some reasonably palatable packaging.  The 'meat' IS
all there, after all.


> I would be happy for discussion either here, or on the Python COM
> developers list at
> http://mailman.pythonpros.com/mailman/listinfo/pycom-dev - that is very
> quiet tho - not much COM work is going on at the moment.

I guess pycom-dev would be more appropriate, to avoid boring
all of the c.l.p readers ... I'll wait for you to 'open the
dances' there!-)


Alex





From syver at NOSPAMcyberwatcher.com  Fri Dec 15 23:08:45 2000
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Sat, 16 Dec 2000 05:08:45 +0100
Subject: reloading modules question
References: <90scdf$hpe$1@troll.powertech.no> <ouk43t494isd9m8fk92udjjbg2pb6crq7q@4ax.com> <910om3$2vk$1@troll.powertech.no> <kg9a3t0s6ah2c7t3pf94o3vl4ls2hespuc@4ax.com> <916m48$648$1@troll.powertech.no> <mjre3t8ta33bj9agc7l4qktmgisibtlvif@4ax.com>
Message-ID: <91eq19$1rh$1@troll.powertech.no>

"Dale Strickland-Clark" <dale at out-think.NOSPAMco.uk> wrote in message
news:mjre3t8ta33bj9agc7l4qktmgisibtlvif at 4ax.com...
> It might be relevent to note that Applciation Protection is set to High on
our sites.

I'll have to check that out, that might be it. Thanks.




From mwh21 at cam.ac.uk  Wed Dec 13 05:37:16 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 13 Dec 2000 10:37:16 +0000
Subject: Thank you, Fredrik. (was Re: Label-Value)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <WntZ5.3566$Kd1.414873@newsb.telia.net> <Pine.GSO.4.21.0012121550530.19682-100000@y.glue.umd.edu> <9009628F0PaCmAnRDLM@161.48.76.20> <ARGZ5.3707$Kd1.429323@newsb.telia.net>
Message-ID: <m34s0838v7.fsf@atrus.jesus.cam.ac.uk>

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

> Fred Pacquier wrote:> As Fredrik says, this info certainly exists elsewhere and could be pieced
> > together from existing material. But having it summed up clearly and in few
> > words in one place can be a revelation, so this definitely needs to go in
> > python-snippets or FAQTS or somewhere...
> 
> note to snippet/faqts maintainers (and other interested parties):
> 
> I've fixed a couple of typos, and posted the article here:
> 
>     http://effbot.org/guides/python-objects.htm
> 

Now I want "M-x gnus-summary-reply-with-effbot-link"!

Good work,
M.

-- 
  You sound surprised.  We're talking about a government department
  here - they have procedures, not intelligence.
                                            -- Ben Hutchings, cam.misc


From tim at degree.ath.cx  Tue Dec  5 17:13:08 2000
From: tim at degree.ath.cx (Tim Hammerquist)
Date: Tue, 05 Dec 2000 22:13:08 GMT
Subject: Script vs Program
References: <y39V5.524$OC.159868@typhoon.aracnet.com> <bHnV5.96503$U46.3094705@news1.sttls1.wa.home.com> <jDFV5.102058$U46.3208014@news1.sttls1.wa.home.com> <3A27BAB5.CF6D060D@san.rr.com> <TI0X5.116132$U46.3743208@news1.sttls1.wa.home.com>
Message-ID: <slrn92qqjl.tl0.tim@degree.ath.cx>

GMT, Greg Jorgensen <gregj at pobox.com> wrote:
> Actually JavaScript used to be called LiveScript, and it was the server-side
> scripting language of the Netscape LiveWire web server. Netscape renamed the
> language and decided it would be cool to enable it in the web client.
> Microsoft was still trailing Netscape at the time so they put JavaScript
> into Internet Explorer to stay compatible with Netscape.
> 

To my understanding, Netscape included "JavaScript" in its web clients,
whereas MSIE, ever stubborn, included it's own modified but similar
"JScript."  JavaScript has been standardized by ECMA.  The standard is
available http://www.ecma.ch/ under the code ECMA-262.  

-- 
-Tim Hammerquist <timmy at cpan.org>

Love is the triumph of imagination over intelligence.
	-- H. L. Mencken


From aahz at panix.com  Thu Dec 21 20:50:42 2000
From: aahz at panix.com (Aahz Maruch)
Date: 21 Dec 2000 17:50:42 -0800
Subject: while true: !!!
References: <mailman.977446988.16433.python-list@python.org>
Message-ID: <91uc1i$j8p$1@panix2.panix.com>

In article <mailman.977446988.16433.python-list at python.org>,
Tim Peters <tim.one at home.com> wrote:
>
>then-again-i-confuse-easily<wink>-ly y'rs  - tim

That's true -- I usually feel confused after reading on of your posts.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

I don't really mind a person having the last whine, but I do mind
someone else having the last self-righteous whine.  --Aahz


From barrett at stsci.edu  Mon Dec  4 16:50:01 2000
From: barrett at stsci.edu (Paul Barrett)
Date: Mon, 04 Dec 2000 16:50:01 -0500
Subject: A better python interface for buffer objects
Message-ID: <3A2C1188.D384E47C@stsci.edu>

I've just started working with buffer objects and noticed that they do
not appear to be well supported at the python level, particularly for
writeable buffer objects which I need. The string object exports a
buffer interface, but only for read-only access. (Though if you test for
the bf_getwritebuffer slot, it appears to be defined, indicating the
string method is writeable. Only when you invoke this function does it
return an error indicating that it's a read-only buffer.  Is there a
reason for implementing it this way, since I find this rather
confusing?)  Since I can't use a string object as a writeable buffer, is
there some other object that I can use as a writeable buffer, not
include that mmap object? What I'd like to do is just create a writeable
buffer directly indicating the desired amount of memory, for example

>>> buff = buffer(1024, 'w')

I'm willing to supply patches to Python to support his type of
functionality, if it can be agreed upon.

 -- Paul

--
Dr. Paul Barrett       Space Telescope Science Institute
Phone: 410-338-4475    ESS/Science Software Group
FAX:   410-338-4767    Baltimore, MD 21218





From aahz at panix.com  Sun Dec 31 15:36:32 2000
From: aahz at panix.com (Aahz Maruch)
Date: 31 Dec 2000 12:36:32 -0800
Subject: Is 100,000 entries big for a dictionary?
References: <92iug4$vqb$1@nnrp1.deja.com> <92j7d4$6qd$1@panix2.panix.com> <0p936.2528$Qb7.277261@newsb.telia.net>
Message-ID: <92o5cg$ooc$1@panix2.panix.com>

In article <0p936.2528$Qb7.277261 at newsb.telia.net>,
Fredrik Lundh <fredrik at effbot.org> wrote:
>
>The implementation limit is currently one billion objects.

It occurs to me, in light of the earlier thread on "is", that the limit
is likely to hit even faster than that, depending on how id() is
implemented.  That is, a non-artificial program that tries creating a
billion entries in a hash table is going to have enough other objects
that the id() space will fill up.

Incidentally, I think you were insufficiently pedantic in that thread
when you suggested "id(a) == id(b)".  It really should be

import __builtin__
if __builtin__.id(a) == __builtin__.id(b):
    ....


;-)

-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

Ninth Virtual Anniversary: Today!


From sill at localhost.kitenet.net  Fri Dec  8 17:44:32 2000
From: sill at localhost.kitenet.net (Oldayz)
Date: Fri, 08 Dec 2000 22:44:32 GMT
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu> <slrn92tu7t.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com> <3A2FE630.5C113AB@bioeng.ucsd.edu> <3A30879B.C518D9FE@engcorp.com>
Message-ID: <kvdY5.50242$II2.4521078@newsread2.prod.itd.earthlink.net>

On Fri, 08 Dec 2000 02:02:51 -0500, Peter Hansen <peter at engcorp.com> wrote:
>Curtis Jensen wrote:
>> 
>> A while back, someone wrote in and asked what a text editor was.  The
>> answers to this were polite and informative, and non-demeaning.  I'm
>> curious why this one and why others get ripped on, and others don't.
>> Luck of the draw?
>
>My first thought was like Andrei's but I actually followed those links,
>as Fredrik did, and realized it wasn't really true that Python info was
>easier to find than the other with a quick google search.
>
>So I went back to the original posting to reread it to see if I could
>understand why we reacted as we did, and whether the reaction was
>warranted.  Let's see, he said:
>
>   "I am doing some research and I would be grateful if 
>    anybody could give me some history on the origins of 
>    Python."
>
>Hmmm... My immediate gut reaction to that is "you say you are *doing
>some research* but you show absolutely no evidence of having done any
>before asking us for help".  In other words, his time is more valuable
>than ours and it isn't worth his even *trying* to find something, since
>we're sitting here waiting for his questions because we have nothing
>better to do.
>
>Had the original poster said even "I am trying to research the origins
>of Python and haven't been able to find any links that help" I suspect
>he would have received little from anyone but helpful links to the real
>thing.
>
>Those of us who have made even the mildest attempt to actually *do* such
>research have found the information rather easy to find.  Maybe not at
>the top of a search engine result, but easily found nonetheless.  I just
>tried going to www.python.org, for example, and within about thirty
>seconds found that clicking on "Documentation", then "FAQ" led me to
>http://www.python.org/doc/FAQ.html#1.15, which contains a good start.
>
>In other words, if the original poster didn't just want us to do his
>research for him, as implied, he shouldn't have suggested he hadn't done
>any himself.  Or perhaps it was simply a poor understanding of the
>sentiment of many people (some of us old-timers?) who have long since
>become tired of "do my homework" type postings.

I couldn't have said it better (and in fact, I didn't ;-) ). I guess
there is a chance that he did do research but simply forgot to mention
it, but I don't think that's too likely, especially if research has
been extensive, for then the guy would express his surprise that the
info is so hard to find!

-- 

	Andrei


From root at rainerdeyke.com  Thu Dec 14 13:28:12 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Thu, 14 Dec 2000 18:28:12 GMT
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <91a4ht02e59@news2.newsguy.com>
Message-ID: <0j8_5.25310$x6.13684532@news2.rdc2.tx.home.com>

"Alex Martelli" <aleaxit at yahoo.com> wrote in message
news:91a4ht02e59 at news2.newsguy.com...
> I'd go further, and say that implementation-inheritance fans
> have never shown a compelling reason why any inheritance AT
> ALL is 'necessary' -- and we never will (I'll class myself among
> such fans:-), because it is *NOT* _necessary_.  Many complex
> LISP applications were programmed without actually using CLOS;
> C was used to program complex, successful applications, and has
> no inheritance at all; Visual Basic, up to version 6, did not
> allow implementation inheritance.  Whatever you can do by
> inheriting implementation, you can also do by encapsulating
> and delegating, and in some languages it's almost as easy
> (in Python, for example, a suitable __getattr__ method may
> let you 'transparently delegate' almost as conveniently as
> inheritance does).

While nothing is truly necessary in the absolute sense, there are cases
where implementation inheritance (and by extension multiple implementation
inheritance) is the only way to reuse another class, i.e. delegation is not
a viable substitute.

class A:
  def f(self): return self.g() + 5
  def g(self): return 5

class B(A):
  def g(self): return 23

my_object = B()
print my_object.f()


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From paulp at ActiveState.com  Tue Dec 12 20:47:32 2000
From: paulp at ActiveState.com (Paul Prescod)
Date: Tue, 12 Dec 2000 17:47:32 -0800
Subject: [Q] Are Exceptions used that much in practice?
References: <3A35D546.A4731B5B@enteract.com> <3A36596A.89EEA210@alcyone.com> <3A36C6E3.26181E3F@enteract.com>
Message-ID: <3A36D534.D4C844F2@ActiveState.com>

Jerome Mrozak wrote:
> 
> ...
> 
> The point of the question is this -- is code like this sample a way of
> exercising the try-catch, or is this an example of something I end up
> doing all of the time because it is so much easier than the alternative?

I've read both your emails a few times and I still don't really
understand the question. 

Python users like exception handling. They tend to like almost
everything about Python's exception handling. Python users implement
code that throws exceptions and Python users implement code that catches
exceptions. Arguably Python forces them to do the latter but not the
former.

Does that help?

> The long-range goal of my question is to determine for my satisfaction
> whether it is worth while for me to spend my limited time pursuing
> Python as a compliment to (whatever other skills I might have) or to go
> with the apparently much greater flow and use those resources on Perl.
> The 'exception' thing could be a determiner, esp. with nothing similar
> in Perl, but only if I can use it in a *positive* way instead of just to
> test a return value.  Heck, Perl has 'die' for that kind of testing.

Over the years, Perl has added features to make its exception handling
more and more like Python's. I expect this trend will continue with Perl
6. 

http://genome-www.stanford.edu/perlOOP/exceptions.html
http://archive.develooper.com/perl6-language-flow%40perl.org/msg00044.html

I would not suggest that this is a basis for making the decision between
the two languages. It's not exactly a killer feature.

 Paul Prescod



From fgeiger at datec.at  Sat Dec 23 04:45:57 2000
From: fgeiger at datec.at (Franz GEIGER)
Date: Sat, 23 Dec 2000 10:45:57 +0100
Subject: Exception Handling and Inheritance
Message-ID: <921sdr$mlq$1@newsreaderg1.core.theplanet.net>

I'd like to specialize Exceptions. So I derived a TException from Exception
and added all the things I think I need to have.

But now there is a problem: I have created a second branch in the
inheritance tree:
Exception
    StandardError
        IOError
        ...
    TException
        TDataException
        ...

When I want to catch TExceptions I write
try:
    obj.meth(anyStuff)
except TException, e:
    print >> sys.stderr, "Panic: %s!" % e.Formatted()

But what about Exceptions? They are not caught by my TException handler! And
Exception does not have my Formatted method.

Do I really have to catch them in a 2nd handler? Or can I add some
functionality somehow to Exception instead of implementing it in a derived
TException?

Many thanks in advance and best regards
Franz GEIGER





From MarkH at ActiveState.com  Tue Dec  5 00:18:44 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Tue, 05 Dec 2000 05:18:44 GMT
Subject: Python programs (or scripts) as drop targets
References: <hv5k2tg9cdf33j8v333qpkk7jo45ht596n@4ax.com> <90h0gq$cu0$1@troll.powertech.no> <1mao2tojod77d92ri841ptdvm6t3e0evqu@4ax.com>
Message-ID: <3A2C7968.4050205@ActiveState.com>

Dale Strickland-Clark wrote:

> Mark has suggested  that it could go in win32com.shell - although I can't figure out what he means
> by that - win32comext.shell, perhaps?

Yes - it is imported as "win32com.shell", but exists in a different tree 
as it is not part of the "COM core".  Such a distinction is a little 
pointless...


> Is the source for this available?

With the rest of the sources.  See my win32all pages.

Mark.



From bradh at mediaone.net  Sat Dec  9 21:14:20 2000
From: bradh at mediaone.net (Brad Howes)
Date: Sun, 10 Dec 2000 02:14:20 GMT
Subject: Python wrapper for C++ STL?
References: <3A31760E.D157EB54@erols.com> <C0gY5.126062$U46.4232859@news1.sttls1.wa.home.com> <90s3fk$j8p$1@nnrp1.deja.com> <fUgY5.126218$U46.4241063@news1.sttls1.wa.home.com> <90s9l9$njd$1@nnrp1.deja.com> <KVjY5.126920$U46.4261650@news1.sttls1.wa.home.com> <90t9n902p8r@news2.newsguy.com> <90u4dc$3h$1@nnrp1.deja.com>
Message-ID: <m3elzhc9aa.fsf@h0050047c4794.ne.mediaone.net>

shindich at my-deja.com writes:

> And I am all for going away from "C" standard! Unfortunately, C++ is not
> a binary standard.

Nor is there one for C. Hence the fun people have with the C preprocessor to
support the wide variety of architectures out there, past and present.

Brad


From aahz at panix.com  Fri Dec 29 11:14:49 2000
From: aahz at panix.com (Aahz Maruch)
Date: 29 Dec 2000 08:14:49 -0800
Subject: Newbie question : What is a thread?
References: <92hod8$298$1@nnrp1.deja.com>
Message-ID: <92id9p$k72$1@panix2.panix.com>

In article <92hod8$298$1 at nnrp1.deja.com>,  <judi_ahivodji at yahoo.fr> wrote:
>
>I'm a beginner in Python an I would like someone to tell me what is a
>thread and/or where I can find other deep explanations about it.

A thread is a unit of program control that is used to keep track of
which statement is being executed.  It's similar to a "process" (and a
process is essentially equivalent to "one running program").  However,
each process can have multiple threads inside it, and the threads all
share that one process's memory.

Threads are useful because of that shared memory (making communication
between parts of a large program somewhat simpler than inter-process
communication).  Threads are also usually faster to context-switch than
processes are, because the computer has to store less information for a
thread than a process.

For a brief not-tutorial, see http://starship.python.net/crew/aahz/
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

Ninth Virtual Anniversary: 2 days and counting


From SBrunning at trisystems.co.uk  Tue Dec 12 08:57:44 2000
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Tue, 12 Dec 2000 13:57:44 -0000
Subject: gcc and Python 2.0 on windows
Message-ID: <31575A892FF6D1118F5800600846864D5B15CC@intrepid>

> From:	John J. Lee [SMTP:phrxy at csv.warwick.ac.uk]
> It seems that the Python 2.0 headers have been patched to allow the use of
> gcc on windows following Paul Sokolovsky's patch to 1.5.2, but only in
> config.h, not in object.h.  I haven't noticed any reference to these
> changes in the C extension tutorial or reference.  I managed to compile
> and run something with the plain Python 2.0 headers, so I presume it's
> fully working - but I don't know what the changes to object.h in Paul's
> patch are supposed to do - something to do with __declspec(dllimport).
> Are these changes redundant now?
 
See <http://starship.python.net/crew/kernr/mingw32/Notes.html>.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk





-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From fredrik at effbot.org  Thu Dec 14 09:54:34 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Thu, 14 Dec 2000 14:54:34 GMT
Subject: comp.lang.python.announce
References: <91aha7$791$1@reader1.imaginet.fr>
Message-ID: <Ka5_5.3904$Kd1.451379@newsb.telia.net>

Gilles Lenfant wrote:
> My ISP's news server does not host "comp.lang.python.announce".
> It is not hosted at deja.com too.

try this link:

http://www.deja.com/=dnc/topics_if.xp?search=topic&group=comp.lang.python.announce

(and complain to your ISP)

</F>




From ollir at linuxfreak.com  Mon Dec  4 10:48:23 2000
From: ollir at linuxfreak.com (Olli Rajala)
Date: Mon, 04 Dec 2000 15:48:23 GMT
Subject: Calculating factorial
Message-ID: <3a2a9496.647995@uutiset.nic.fi>

Hi!
I've learned a few time Python and programming.  I wrote a function
that calculates a factorial of N and I was wondering if it's the best
(I think it's not. =) way to do that. I'm glad to hear your comments.

<script>

def factorial(f):
    result = 1
    for number in range(1, f + 1):
        result = result * number
    return result
    
print factorial(10)

</script>

Ps. Sorry my poor English. 
-- 
Olli Rajala
A Finnish student who loves Linux.


From cg at gaia.intranet.cdegroot.com  Tue Dec 19 06:52:30 2000
From: cg at gaia.intranet.cdegroot.com (Cees de Groot)
Date: 19 Dec 2000 12:52:30 +0100
Subject: Python for REAL Projects ?
References: <3a380d1e.409238@bendigo.news.origin.net.au> <3eq_5.5732$n51.113513@e420r-atl1.usenetserver.com>
Message-ID: <91ni5u$nvj$1@gaia.intranet.cdegroot.com>

Steve Holden <sholden at holdenweb.com> said:
>Personally I wouldn't use MySQL for commercial projects, because of its lack
>of support for transactions and referential integrity constraints.  The
>former is slated for implementation at a later date, but the latter appears
>to be regarded as a feature by MySQL users - at least according to comments
>I have seen to the effect that "these can easily be maintained by the
>application" (in New Riders' "MySQL").  I don't know whether this is typical
>of the MySQL community.
>
Probably. But then, I think mySQL is used for 99.9% in web environment,
where you often have a different transaction model and mySQL's concept
of atomic operations work quite well. You need to be a little bit more
aware of some issues, but the payoff in performance is often worth
it for websites with unpredictable loads.

-- 
Cees de Groot               http://www.cdegroot.com     <cg at cdegroot.com>
GnuPG 1024D/E0989E8B 0016 F679 F38D 5946 4ECD  1986 F303 937F E098 9E8B


From scarblac at pino.selwerd.nl  Mon Dec 18 07:06:01 2000
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 18 Dec 2000 12:06:01 GMT
Subject: Error evaluating numbers starting in zero
References: <mailman.977137021.11074.python-list@python.org>
Message-ID: <slrn93rv86.45o.scarblac@pino.selwerd.nl>

Mike Brenner <mikeb at mitre.org> wrote in comp.lang.python:
> >> I have a python checkbook program that evaluates date strings, such
> >> as "01/31/99". Most dates are fine, but if I have a date such as
> >> "01/08/99", python runs into an error when it tries to eval("08").
> 
> > Use int() instead of eval().
> 
> 
> So, that may also explain the failure with 
> 
> 	i = 08
> 
> and with reading in numbers,
> the python interpretor must be using EVAL instead of INT?

No, the failure is that this is a syntax error. In fact, it's the
other way around - eval() uses the python interpreter. That's what eval
does, evaluate an expression like the interpreter does. int() converts
things to integers.

If a number starts with 0, it is in octal and not in decimal. So i = 08 is a
syntax error.

-- 
Remco Gerlich


From jepler at inetnebr.com  Mon Dec 11 14:41:06 2000
From: jepler at inetnebr.com (Jeff Epler)
Date: Mon, 11 Dec 2000 19:41:06 GMT
Subject: [1.5, 2.0] parser.tuple2ast error on **kw arguments
Message-ID: <slrn93abeg.8jh.jepler@potty.housenet>

This looks like a bug in the tuple2ast function of the parser module
in 1.5.2 and 2.0.  I discovered this bug while extending the Python
2.0 test_parser module with more test cases.  Unfortunately, I haven't
yet investigated the bug further, nor produced a patch.

Python 1.5.2 (#1, Sep 17 1999, 20:15:36)  [GCC egcs-2.91.66 19990314/Linux (egcs- on linux-i386
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import parser
>>> parser.tuple2ast(parser.expr("lambda x, *y, **z: 0").totuple())
Traceback (innermost last):
  File "<stdin>", line 1, in ?
parser.ParserError: Expected node type 16, got 36.
>>> lambda x, *y, **z: 0
<function <lambda> at 80c6858>

Python 2.0 (#1, Dec 10 2000, 12:10:38)
[GCC egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> import parser
>>> parser.tuple2ast(parser.expr("lambda x, *y, **z: 0").totuple())
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
parser.ParserError: Expected node type 262, got 16.
>>> lambda x, *y, **z: 0
<function <lambda> at 0x81b7004>

PS I enclose my extended list of tests.  Some are commented out
because I am currently testing my (spark-derived) parser against
the 1.5.2 subset.  I hope that a larger parser test suite might be
included in a future release of Python, perhaps drawing on these
test cases.
PPS My revised roundtrip() function, not included, compares the output
of my parser to the totuple() form of the AST produced by Python itself,
thus ensuring a high degree of confidence in the compatibility of the
two.


print "Expressions:"

test_expr('r"hi"')
test_expr("'hi'")
test_expr("'hi' 'there'")
test_expr("''' 'hi' 'there' '''")
test_expr('"hi"')
test_expr('"hi" "there"')
test_expr('""" "hi" "there" """')
test_expr("foo(1)")
test_expr("foo(1,)")
test_expr("foo(1, 2)")
test_expr("foo(1, 2,)")
test_expr("foo(1, a=2)")
test_expr("foo(1, a=2,)")
test_expr("foo(1, a=2, b=2,)")

test_expr("(1, 2, 3)")
test_expr("[1, 2, 3]")
#test_expr("[x**3 for x in range(20)]")
#test_expr("[x**3 for x in range(20) if x % 3]")
#test_expr("foo(*args)")
#test_expr("foo(*args, **kw)")
#test_expr("foo(**kw)")
test_expr("foo(key=value)")
#test_expr("foo(key=value, *args)")
#test_expr("foo(key=value, *args, **kw)")
#test_expr("foo(key=value, **kw)")
#test_expr("foo(a, b, c, *args)")
#test_expr("foo(a, b, c, *args, **kw)")
#test_expr("foo(a, b, c, **kw)")
test_expr("foo + bar")
test_expr("foo - bar - baz")
test_expr("foo * bar")
test_expr("foo * bar / baz")
test_expr("foo ^ bar")
test_expr("foo ^ bar ^ baz")
test_expr("foo | bar")
test_expr("foo | bar | baz")
test_expr("foo & bar")
test_expr("foo & bar & baz")
test_expr("foo > bar > baz")
test_expr("foo < bar")
test_expr("+foo")
test_expr("-foo")
test_expr("not foo")
test_expr("foo not in bar")
test_expr("foo in bar")
test_expr("foo is bar")
test_expr("foo is not bar")
test_expr("x[0]")
test_expr("x[0:]")
test_expr("x[:0]")
test_expr("x[:]")
test_expr("x[0:0]")
test_expr("x[0:0:0]")
test_expr("x[:0:0]")
test_expr("x[0::0]")
test_expr("x[0:0:]")
test_expr("x[::0]")
test_expr("x[0::]")
test_expr("x[:0:]")
test_expr("x[::]")
test_expr("lambda: 1")
test_expr("lambda x: 1")
test_expr("lambda x,: 1")
test_expr("lambda x,y: 1")
test_expr("lambda x,y,: 1")
test_expr("lambda x=1: 1")
test_expr("lambda x=1,: 1")
test_expr("lambda x=1,y=2: 1")
test_expr("lambda x=1,y=2,: 1")
test_expr("lambda x,y=2: 1")
test_expr("lambda x,y=2,: 1")
test_expr("lambda x,y,: 1")
test_expr("lambda x,*y: 1")
test_expr("lambda x,**y: 1")
#test_expr("lambda x,*y,* *z: 1")  # Fails in 1.5.2 due to bug in ast2tuple
#test_expr("lambda x,*y,**z: 1")  # Fails in 1.5.2 due to bug in ast2tuple

print
print "Statements:"
test_suite("return")
test_suite("return 1")
test_suite("return 1,2")
test_suite("raise")
test_suite("raise a")
test_suite("raise a, b")
test_suite("raise a, b, c")
test_suite("del x")
test_suite("del x[0]")
test_suite("del x[:]")
test_suite("x=0")
test_suite("x[0]=0")
test_suite("x[:]=0")
test_suite("import x")
test_suite("import x.y")
test_suite("import z, x")
test_suite("import x.y, x.z")
test_suite("from x import z")
test_suite("from x import z, a")
test_suite("from x.y import z")
test_suite("global a")
test_suite("global a")
test_suite("global a,b")
test_suite("global a,b")
test_suite("pass")
test_suite("print")
test_suite("print 1")
test_suite("print 1,")
test_suite("print 1, 2")
test_suite("if 1: pass")
test_suite("if 1:\n\tpass\n")
test_suite("if 1:\n\tpass\n\tpass\n")
test_suite("for i in (): pass")
test_suite("for i in ():\n pass\nelse:\n pass\n")
test_suite("while 1:\n pass\nelse:\n pass\n")
test_suite("while 1:\n continue\nelse:\n pass\n")
test_suite("while 1:\n break\nelse:\n pass\n")
test_suite("try: pass\nexcept: pass\n")
test_suite("try: pass\nfinally: pass\n")
test_suite("try: pass\nexcept a, b: pass\n")
test_suite("try: pass\nexcept a: pass\n")
test_suite("def foo(): pass")
test_suite("def foo(a): pass")
test_suite("def foo(a,): pass")
test_suite("def foo(a, b): pass")
test_suite("def foo(a, b,): pass")
test_suite("def foo(a, b=2): pass")
test_suite("def foo(a, b=2,): pass")
test_suite("def foo(a, *c): pass")
test_suite("def foo(a, **d): pass")
#test_suite("def foo(a, *c, **d): pass") # Fails in 1.5.2 due to parser error
#test_suite("def foo(a, *c, * *d): pass") # Fails in 1.5.2 due to parser error
test_suite("class c: pass")
test_suite("class c(a): pass")
test_suite("class c(a,): pass")
test_suite("class c(a,b): pass")
test_suite("class c(a,b,): pass")
#test_suite("print >>fp")
#test_suite("print >>fp, 1")
#test_suite("print >>fp, 1,")
test_suite("assert 1")
test_suite("assert 1, 1")
# expr_stmt
test_suite("a")
test_suite("a = b")
test_suite("a = b = c = d = e")
test_suite("a, b = b, a")
#test_suite("a += b")
#test_suite("a -= b")
#test_suite("a *= b")
#test_suite("a /= b")
#test_suite("a %= b")
#test_suite("a &= b")
#test_suite("a |= b")
#test_suite("a ^= b")
#test_suite("a <<= b")
#test_suite("a >>= b")
#test_suite("a **= b")




From robin at jessikat.fsnet.co.uk  Fri Dec 22 13:02:27 2000
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Fri, 22 Dec 2000 18:02:27 +0000
Subject: try...else
Message-ID: <O9RyIIAzc5Q6EwCS@jessikat.fsnet.co.uk>

I had expected that in try: except: else
the else clause always got executed, but it seems not for return

PythonWin 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)] on
win32.Portions Copyright 1994-2000 Mark Hammond (MarkH at ActiveState.com)
- see 'Help/About PythonWin' for further copyright information.
>>> def bang():
...     try:
...             return 'return value'
...     except:
...             print 'bang failed'
...     else:
...             print 'bang succeeded'
...
 >>> bang()
'return value'
>>> 

is this a 'feature' or bug. The 2.0 docs seem not to mention
return/continue except for try finally.
-- 
Robin Becker


From DanielK at aracnet.com  Sat Dec 30 12:08:58 2000
From: DanielK at aracnet.com (Daniel Klein)
Date: Sat, 30 Dec 2000 09:08:58 -0800
Subject: Controlling processes (popen2)
Message-ID: <Wyo36.73$KD3.19486@typhoon.aracnet.com>

First some background info: I've been learning Python seriously now for
about a month and the first real task I am faced with is to write a Python
script which controls a BASIC process. Additionally I have to do the same
thing in Java. (The BASIC process is used to interface with a third party
database [open files, read, write, etc], but this is just fyi.) The script
must be cross-platform but I am testing this on Win2000 as I don't have a
Linux box available at the moment.

The problem is that in Python I have to terminate all output to the BASIC
process with a \r (rather than \n). In Java, I can use either \n or \r.

To illustrate this, here is a simple BASIC program which compiles to a C
executable called say, TESTIT (note that a ':' is used for concatenation
here):

INPUT LINE1
INPUT LINE2
PRINT "LINE1=" : LINE1
PRINT "LINE2=" : LINE2

Here is the companion Python script which controls this process:

import popen2
instream, outstream = popen2.popen2("TESTIT.exe")
outstream.write("dead\n")
outstream.write("parrot\n")
outstream.flush()
print instream.readline(),
print instream.readline(),
outstream.close()
instream.close()

...and here is a similar one in Java:

import java.io.*;
public class Test1 {
  public static void main(String[] args) {
    try {
      Process process = Runtime.getRuntime().exec("TESTIT.exe");
      OutputStreamWriter outstream = new
OutputStreamWriter(process.getOutputStream());
      BufferedReader instream = new BufferedReader(new
InputStreamReader(process.getInputStream()));
      outstream.write("dead\n");
      outstream.write("parrot\n");
      outstream.flush();
      System.out.println(instream.readLine());
      System.out.println(instream.readLine());
      instream.close();
      outstream.close();
    }
    catch (Exception e) {
      e.printStackTrace();
    }
  }
}

The output from the Python script using the \n is:

LINE1=dead
LINE2=

Notice that it printed the literal text but not the data sent to the input
stream for the second line.

And the corresponding output from Java (which also uses \n):

LINE1=dead
LINE2=parrot

If I change the Python script to use \r, it matches the Java output. Now I
know you're going to say to use \r (and that is what I'm doing), but the
question is, why does Python behave differently?

Any ideas?

Daniel Klein
Portland, OR USA




From itamar at NOSPAMmaxnm.com  Mon Dec  4 09:45:09 2000
From: itamar at NOSPAMmaxnm.com (Itamar Shtull-Trauring)
Date: Mon, 04 Dec 2000 16:45:09 +0200
Subject: Extending Windows Shell with Python
Message-ID: <3A2BADF5.AE73CDD@NOSPAMmaxnm.com>

http://msdn.microsoft.com/library/psdk/shellcc/shell/Shell_Adv/namespace.htm
explains how to extend the Windows Explorer so that you can browse your own
namespaces. I've seen it used, for example, to allow you to browse a FTP
site through the Windows Explorer.

Is it possible to this in Python?  I assume it is, and I'm going to try
doing it myself, but if someone's already done it I'd love to have some
example code.

TIA.

-- 
Itamar Shtull-Trauring <itamar(at)shtull-trauring.org>


From tim.one at home.com  Fri Dec 15 14:55:56 2000
From: tim.one at home.com (Tim Peters)
Date: Fri, 15 Dec 2000 14:55:56 -0500
Subject: __repr__ policy problem: should module be prefixed to output?
In-Reply-To: <3A3A3A6D.2C3B1A91@sbc.su.se>
Message-ID: <LNBBLJKPBEHFEDALKOLCAECGIEAA.tim.one@home.com>

[posted & mailed -- although the post won't show up on c.l.py cuz
 the mail->news gateway is still broken]

[Per Kraulis]
> Having just released a Python extension module crng
> (http://www.sbc.su.se/~per/crng/), I have received some constructive
> criticism, one issue being save/restore functionality. My plan is to
> define a __repr__ for each type in the module.

See also the pickle module (and cPickle, its faster workalike).  A faithful
repr is nice to have, though!

[Alex Martelli]
> That may give problems with state that's in floats, so I'd rather
> have separate checkpointing save/restore mechanisms (returning a
> tuple of picklable data).  Still, the repr issue IS interesting ...

Python 1.6 changed repr(float) to print enough digits so that, on any
platform with 754-conforming C double I/O, eval(repr(x)) == x for every
finite float x.  IOW, so long as as repr(ParkMiller) converts floats to
strings using repr(), eval should reproduce the exact float state (else
you're on an oddball platform with sub-standard libc I/O).

[back to Per Kraulis]
> However, I have realized that there is a policy problem. Should __repr__
> add the module name as prefix to the output string, or not?

Just pick one and be happy; I believe most modules do not add the module
name.  As Alex explained, that choice does not imply "import *".

> Or is there some other solution?

Pickle/cPickle, but they're not human-readable.

> ...
> A suggestion is that there should be a settable module variable that
> controls this. Feels like a kludge to me...

Bleech.  repr() is also used as an interchange format, and then dealing with
a choice spreads to both ends of the channel (the sender's "freedom" becomes
the receiver's headache).  Pick just one and don't look back!




From rooneg at electricjellyfish.net  Fri Dec  8 19:37:45 2000
From: rooneg at electricjellyfish.net (Garrett Rooney)
Date: Fri, 8 Dec 2000 19:37:45 -0500
Subject: Python wrapper for C++ STL?
In-Reply-To: <3A31760E.D157EB54@erols.com>; from edcjones@erols.com on Fri, Dec 08, 2000 at 07:00:14PM -0500
References: <3A31760E.D157EB54@erols.com>
Message-ID: <20001208193745.A64496@electricjellyfish.net>

On Fri, Dec 08, 2000 at 07:00:14PM -0500, Edward C. Jones wrote:
> Has anyone written a Python wrapper for the C++ Standard Template
> Library?

out of curiosity, why would you want such a thing?

what does the STL get you that you could not more easily get from what
python already provides you?

granted it's been a while since i did C++ and the STL programming, so
maybe i'm missing something obvious.

-- 
garrett rooney                           my pid is inigo montoya.
rooneg at electricjellyfish.net             you kill -9 my parent process.
http://electricjellyfish.net/            prepare to vi.



From MarkH at ActiveState.com  Wed Dec 13 18:52:32 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Wed, 13 Dec 2000 23:52:32 GMT
Subject: Problems With socket Under NT & PythonWin 2.0
References: <910jtt$nnh$1@nnrp1.deja.com> <fMQY5.12943$hp.1457332@typhoon.nyroc.rr.com> <915k4g$kcc$1@nnrp1.deja.com> <8QzZ5.1711$7g.340689@typhoon.nyroc.rr.com> <3A37F5FA.3020209@ActiveState.com> <W1TZ5.16500$xW4.133590@news-server.bigpond.net.au>
Message-ID: <3A380A81.7050207@ActiveState.com>

Neil Hodgson wrote:

>    While out-of-proc is good for debugging, I like also having the ability
> to run things that change the main PythonWin environment. Makes it really
> flexible and allows exploration with a combination of running stuff in files
> and from the Interactive Window. Then you don't have to build throw-away
> test scripts that starts from zero.


I agree, and have been thinking about this too.  However, there is no 
reason I can see that a single, long-lived external process can't be 
used for the interactive window.

About the only thing you lose is the ability to work directly with 
Pythonwin's GUI environment - eg, creating Windows, dialogs etc directly 
from the prompt.  However, I believe this is a minority situtation.

A single long lived process would make it appear, in most cases, like it 
is really in-process.  (maybe ;-)  It would also provide a simple way to 
"reset the environment" (and oft requested feature in Pythonwin) - you 
just restart the process.  I am reluctant to add this feature to 
Pythonwin as things stand, as the best we can do is _pretend_ to reset 
the environment by wiping a few dictionaries.

Mark.



From tismer at tismer.com  Sun Dec  3 09:16:13 2000
From: tismer at tismer.com (Christian Tismer)
Date: Sun, 03 Dec 2000 16:16:13 +0200
Subject: How to extend PythonWin auto-completion - ?
References: <stf5qnosg4op6d@corp.supernews.com> <p6qpulk724m.fsf@informatik.hu-berlin.de> <stf9pf5pcvkcf4@corp.supernews.com>
Message-ID: <3A2A55AD.36CFCE3B@tismer.com>


Jason Cunliffe wrote:
> 
> > Use the source, Luke. I haven't checked it myself, but I guess the
> > completion is implemented in 100% pure Python, so it should not be too
> > hard to find the place where it's done, and improve it. Don't forget
> > to send your patches to Mark Hammond afterwards.
> 
> ok I will try...
> 
> Q1: Does anyone know where the detection source code  which determsni what
> happens  when you type "." after some text in PythonWin shell?

Actually I know.
But here is a little trick for you how to find out yourself:

When you are typing the dot in thing.stuff for instance,
there is an attribute lookup taking place right after you
were hitting the dot.

Now, how do we find out where this code lives?
Easy! Write a little class with a ridiculous __getattr__
and save the calling frame chain. Here it is:

>>> class gotcha:
... 	def __getattr__(self, name):
... 		global bingo
... 		try:
... 			1/0
... 		except:
... 			import sys
... 			ign, ign, bingo = sys.exc_info()
... 			
>>> catchit = gotcha()
>>> catchit.   # hit escape here
>>> bingo
<traceback object at 00D64E30>
>>> 

Now type ctrl-b and browse the traceback object. It has a frame
which is you interactive window. Open it's f_back frame, open
its f_code object, and you can read the filename:
D:\Python20\Pythonwin\pywin\scintilla\view.py
f_lineno tells you to go to line 413. You are exactly after
the code that tried to find properties of your object.

Neat, isn't it?

But be warned, tricky stuff here :-)

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From gerson.kurz at t-online.de  Sun Dec 31 09:54:33 2000
From: gerson.kurz at t-online.de (Gerson Kurz)
Date: Sun, 31 Dec 2000 14:54:33 GMT
Subject: Embedding Python in Windows Applications
References: <3a4e28bf.16852652@news.t-online.de> <3A4E7BC4.8040101@ActiveState.com>
Message-ID: <3a4f4581.27327344@news.t-online.de>

On Sun, 31 Dec 2000 00:25:09 GMT, Mark Hammond <MarkH at ActiveState.com>
wrote:

>Python itself always prints via "sys.stdout" and "sys.stderr".  Thus, if 
>the first thing you do is to set sys.stdout and sys.stderr to some 
>object of yours, again you have full control over where this output 
>goes.  But again, only you know what makes sense.

Thanks, but how would you do that ? For starters, I have tried the
following test code (in your excellent PythonWin):

-----------------------------------------------
import sys

class myclass:
    def __init__(self):
        self.file = open("D:\\test.log","w")
    
    def write(self,str):       
        self.file.write(str)

    def __del__(self):
        self.file.close()

old_stdout = sys.stdout
sys.stdout = myclass()
print "Hello, world"
sys.stdout = old_stdout
print "Done."
-----------------------------------------------

Now, as expected the first print statement will write to the file,
BUT, the second statement will raise an IO Error.

Traceback (most recent call last):
  File "d:\python20\pythonwin\pywin\framework\scriptutils.py", line
298, in RunScript
    debugger.run(codeObject, __main__.__dict__, start_stepping=0)
  File "d:\python20\pythonwin\pywin\debugger\__init__.py", line 60, in
run
    _GetCurrentDebugger().run(cmd, globals,locals, start_stepping)
  File "d:\python20\pythonwin\pywin\debugger\debugger.py", line 582,
in run
    _doexec(cmd, globals, locals)
  File "d:\python20\pythonwin\pywin\debugger\debugger.py", line 921,
in _doexec
    exec cmd in globals, locals
  File "D:\Python20\Script1.py", line 17, in ?
    print "Done."
  File "D:\Python20\Script1.py", line 6, in write
    
IOError: (0, 'Error')

Secondly, can you tell me which methods I have to implement in my
pseudo-stdout ? And why don't I have to derive my class from some base
file class (or do I) ?



From  at home.com  Thu Dec 28 23:02:32 2000
From:  at home.com (John Q. Public)
Date: Fri, 29 Dec 2000 04:02:32 GMT
Subject: PythonWin Question
References: <3A443EF1.125426ED@home.com> <3A4B3244.6080502@ActiveState.com>
Message-ID: <3A4C0CE1.52D4C11D@home.com>

I understand the reason for resetting stdout (having shot myself in the foot
when I overwrote stdout.write()).  Is it safe to assume that I can overwrite
sydout for the duration of a script?  This is really all I want to do, so I
can capture the output of a test and still show the tester what is happening.

Thanks!

JonathanPolley

jwpolley(a)collins(d)rockwell(d)com

Mark Hammond wrote:

> John Q. Public wrote:
>
> > While trying to add a journaling mechanism to python, I came across an
> > odd behavior from PythonWin when I tried to override sys.stdout.  For
> > both the console and IDLE versions of the interpreter, I can override
> > sys.stdout, but not for PythonWin.
> >
>
> This is a "feature".  Quite often I found people running scripts that
> themself set sys.stdout.  This would then mess-up Pythonwin after the
> script had completed - so Pythonwin resets stdout after running a script.
>
> Mark.



From gtalvola at NameConnector.com  Thu Dec 14 10:51:06 2000
From: gtalvola at NameConnector.com (gtalvola at NameConnector.com)
Date: Thu, 14 Dec 2000 15:51:06 GMT
Subject: Source Code Management
References: <fb4f3tsgffnmvme64aol1cn397te7ln349@4ax.com>
Message-ID: <91aq96$tba$1@nnrp1.deja.com>

If you don't mind paying money, I believe Perforce fits all your
requirements.  http://www.perforce.com/  We use it and have nothing
but good things to say about it.

And it has explicit support for being scripted in Python -- you can get
the results of its commands in Python marshalled dictionary format,
which makes scripting Perforce using Python pretty easy.

In article <fb4f3tsgffnmvme64aol1cn397te7ln349 at 4ax.com>,
  dale at out-think.NOSPAM.co.uk wrote:
> I'm looking for a source code management system for looking after
Python, Java, Javascript and
> possibly other languages too.
>
> Does anyone know of such a system that matches or comes close?
>
> Essential features:
> * Runs on Windows 2000
> * Works over the Internet (ftp would be good)
> * Secure (ftp level security would pobably suffice)
> * Module level version control with deltas or compressed archives
> * Package level (multiple modules) sign in/sign out
> * System level release control. (ie, this is now release x, note all
module versions for this
> release.)
>
> Desirable features:
> * Extendible - preferably scriptable with Python
> * Manage arbitrary files including documentation
> * GUI or command line interfaces
> * Low maintenance
> * Library stored as plain files and not in some odd database format
>
> That's not too much to ask, is it!  :-)
>
> Thanks
> --
> Dale Strickland-Clark
> Out-Think Ltd
> Business Technology Consultants
>


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


From Daniel.Kinnaer at AdValvas.be  Wed Dec 27 15:35:59 2000
From: Daniel.Kinnaer at AdValvas.be (Daniel)
Date: Wed, 27 Dec 2000 20:35:59 GMT
Subject: Just starting out!
References: <92buip$kcp$1@nnrp1.deja.com> <u5i26.194888$U46.5994443@news1.sttls1.wa.home.com> <3a49c756.8528401@news.skynet.be> <slrn94jj88.gfb.scarblac@pino.selwerd.nl>
Message-ID: <3a4a5130.13424557@news.skynet.be>

On 27 Dec 2000 11:08:46 GMT, scarblac at pino.selwerd.nl (Remco Gerlich)
wrote:

>
>There is an electronic book called "(the eff bot guide to) The Standard
>Python Library" that is based on over 2500 newsgroup posts by Fredrik
>Lundh, it has examples for every standard Python library.
>
>http://www.secretlabs.com/people/fredrik/librarybook.htm
>
>It's $12 (Windows only I think).
>

Remco, bedankt!

I've downloaded the electronic book and was able to solve my
NNTP-problems within an hour. The book has about 320 example scripts
from which it is easy to learn. I can greatly recommend this book to
all the Python programmers. It really is great value for only $12  !!

Happy programming

Daniel




From emile at fenx.com  Tue Dec 19 20:30:22 2000
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 19 Dec 2000 17:30:22 -0800
Subject: How to get the os.system data into a variable
References: <91o5ih$2kv$1@news.wrc.xerox.com>
Message-ID: <3A400BAE.5020902@fenx.com>

import os
info = os.popen('ls -l').readlines()

HTH

Emile van Sebille
emile at fenx.com


jicman wrote:

> Hi all!
> 
> I know that
> 
> t = os.system('ls -l')
> 
> return code and assigns it to t.  I would like to get the data, not the code, 
> that is resulted from the command into a variable.  (i.e.
> 
> -rw-r--r--   1 root     other        821 Dec  2 12:20 MyLog.txt
> drwxr-xr-x   2 root     root         512 Dec 21  1999 TT_DB
> drwxrwxrwx   3 root     root         512 Aug 18 17:05 apps
> lrwxrwxrwx   1 root     root           9 Dec 20  1999 bin -> ./usr/bin
> drwxr-xr-x   3 root     nobody       512 Nov 28 12:31 cdrom
> drwxr-xr-x   2 root     other        512 Nov 21 20:10 cgi-bin
> -rw-------   1 root     root      975040 Oct  3 11:34 core
> 
> etc.)
> 
> I know that I could do
> 
> t = os.system('ls -l > mytemp.txt')
> 
> and then read the file, but I would like to acquire that data w/o reading the 
> file.  Oh, and by the way, the command that I am sending is not really ls.  It 
> does not even belong to the unix world.  It's the output of a command from an 
> application that I have.
> 
> Any ideas?  It would be greatly appreciated.
> 
> thanks,
> 
> jos?



From fredrik at effbot.org  Fri Dec 15 04:35:17 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 15 Dec 2000 09:35:17 GMT
Subject: Getting a code obj's function?
References: <91co73$fpm$1@news.rz.uni-karlsruhe.de>
Message-ID: <pBl_5.3966$O%.626490@newsc.telia.net>

Bjoern Giesler wrote:
> ...is there a standard way to get the function object that a code object is
> associated with? On exception, I only get the code object from the
> traceback.

Use co_name:

    def spam():
        pass

    >>> spam
    <function spam at 9281b8>

    >>> spam.func_code
    <code object spam at 91cd08, file "<stdin>", line 1>

    >>> spam.func_code.co_name
    'spam'

For more ideas on what you can do with frames and code objects, check
the source code for the traceback module (in the standard library).

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From phd at phd.pp.ru  Thu Dec 14 08:50:57 2000
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 14 Dec 2000 16:50:57 +0300 (MSK)
Subject: sorting date-strings of form: Wed, 05 Jan 2000 10:07:02
In-Reply-To: <3A38CC30.8DA734F4@psi.ch>
Message-ID: <Pine.LNX.4.30.0012141649450.14664-100000@fep132.fep.ru>

On Thu, 14 Dec 2000, Jorn Verwey wrote:
> I have a document with a lot of sections which need to be ordered by date. The date is given in the
> format given in the subject: Wed, 05 Jan 2000 10:07:02. How would I go about? Being a novice I would
> rather not re-invent the wheel (it will end up square). I took notice of two packages which might
> help me (mxDateTime and normalDate), but it didn't seem straight forward. I welcome suggestions!
> TIA.

   I would recommend use mxDateTime the following way: convert all date
strings to DateTime objetcs, and sort a list of such objects. DateTime
objects has all neccessary features.

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




From dennis at codegrinder.org  Fri Dec 15 16:19:33 2000
From: dennis at codegrinder.org (Dennis Conrad)
Date: Fri, 15 Dec 2000 22:19:33 +0100
Subject: stewpid newbie tricks
References: <3A39D4F9.648D8DBA@mediaone.net> <91d14b$1t4$1@bob.news.rcn.net>
Message-ID: <91e23e$eie$03$1@news.t-online.com>

> Perhaps it should be
>
> at 16:00 /Every:m,t,w,th,f,s,su "d:\program
> files\python\mine\spider\runSpider.bat"

Jepp, should be. But you should also change the batch-file to something like

C:\Python20\python.exe "D:\Program files\python\mine\spider\mySpider.py"

because the scheduler service runs as 'System' by default and doesn't know
the PATH the users have set.

Cheers

Dennis

P.S.: Sorry Satheesh I mailed you directly, I'm too lame to use Outlook
Express ;)




From fredrik at effbot.org  Mon Dec 11 12:51:41 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Mon, 11 Dec 2000 17:51:41 GMT
Subject: Microsoft SOAP not compatible with PythonWare's SOAP?
References: <912q8s$c4i$1@news1.xs4all.nl>
Message-ID: <Nu8Z5.3453$Kd1.391596@newsb.telia.net>

Iwan van der Kleyn wrote:
> However, during testing I've not been able to retrieve any data from the
> test MS SOAP components. The soaplib seems to working fine, though. Is this
> the result of "competing" SOAP implementations?

yup.  we're working on a new version, but there's no
known ETA at this time...

(probably not before christmas, though)

for the time being, I recommend using XML-RPC instead.
implementations for COM and a dozen other platforms can
be found here:

    http://www.xmlrpc.com/

</F>




From nas at arctrix.com  Tue Dec 12 05:26:52 2000
From: nas at arctrix.com (Neil Schemenauer)
Date: Tue, 12 Dec 2000 02:26:52 -0800
Subject: Inheriting the @ sign from Ruby
In-Reply-To: <3A3658D3.4AAA9AB0@alcyone.com>; from max@alcyone.com on Tue, Dec 12, 2000 at 08:56:51AM -0800
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <3A3658D3.4AAA9AB0@alcyone.com>
Message-ID: <20001212022652.B13315@glacier.fnational.com>

On Tue, Dec 12, 2000 at 08:56:51AM -0800, Erik Max Francis wrote:
> Roy Katz wrote:
> > (I've been programming in Python for three years now
> > and I understand only that passing an int or float is call-by-value,
> > otherwise it is call-by-reference.  At the risk of naivete, I would
> > like a simple & referencing operator so that I can explicitly,
> > finally, have control over this.)
> 
> This is actually something I wish had been done explicitly as well; this
> is one of the only areas of Python that I found truly irregular and
> strange.  It would have been nice to either have pass-by-value be the
> default and have an explicit operator for referencing, or vice versa
> (doesn't matter to me which one, as long as it's consistent;
> pass-by-value as the default makes a little more sense).  Alas, it is a
> little too late for that.

You are both confused.  Roy, I find it hard to believe that you
could program in Python for three years and still have such a
warped mental model of it works.  Erik, there is no irregularity.
There is no "pass by value".  Variables do no contain values they
reference them.  Always.

Furthermore, "pass by value" would be a horrible default.
Imagine passing a multiple megabyte string or list around.  If it
is passed by value then a copy must be made for every function
call.

  Neil



From ullrich at math.okstate.edu  Sun Dec 31 08:10:30 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Sun, 31 Dec 2000 13:10:30 GMT
Subject: how do i check if a file exists with python?
References: <3a4c95dc.88792787@nntp.sprynet.com>, <fBN26.99045$yR4.2640173@news1.rdc1.tx.home.com>, <RlM26.98766$yR4.2639444@news1.rdc1.tx.home.com> <92g5v8$r5k$1@nnrp1.deja.com> <mailman.978078157.26485.python-list@python.org> <mailman.978264789.28740.python-list@python.org>
Message-ID: <3a4f2bd4.45424534@nntp.sprynet.com>

On Sun, 31 Dec 2000 17:39:21 +0200 (IST), Moshe Zadka
<moshez at zadka.site.co.il> wrote:

>On Fri, 29 Dec 2000 14:02:25 GMT, ullrich at math.okstate.edu (David C. Ullrich) wrote:
[...]
>
>> I sometimes make cumulative backups of files containing
>> important stuff. I sort of need to determine that file.001,
>> file.002 and file.003 already exist and file.004 does not
>> before deciding what to name the current backup file.
>
>That's *really* evil. What if there is another version of your
>process running???  Damn!, just lost that backup file. In that
>case, I'd go for a Maildir-like algorithm, which promises no colisions.
>Actually, in that case, usually what you want is CVS, but that's
>a matter for another day.

Really evil, oh dear.

In the actual program I was referring to there's no problem
with other processes interfering - it's a program running on
one PC with one user. (The one user is me, and I'm not
likely to have two instances of the gradebook open,
and even if I did I'd have to try really hard to get them
to interfere with each other this way - the backup happens
in response to an explicit user request and I don't see how
I could make this request close enough to simultaneously
in both windows to cause a problem.)

But there's a project coming up that's going to involve 
people making additions to a data file through a web
page, and collision avoidance is something I've been
wondering about. No doubt the right solution is an
actual database or CVS or something, but I'm too
stubborn to go with something like that except as a
last resort - doing it myself is more educational
(this is a context where losing data would be just
inconvenient, not a huge disaster.) Is it possible
to give a hint how a "Mailder-like" algorithm
promises no collisions?

DU


From grey at despair.rpglink.com  Wed Dec 20 23:15:10 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Thu, 21 Dec 2000 04:15:10 -0000
Subject: Greetings from the Interpreter!!!
References: <mailman.977314029.15733.python-list@python.org> <slrn942oun.eoe.grey@teleute.rpglink.com> <G5wCy5.783@world.std.com>
Message-ID: <slrn9430ue.ft7.grey@teleute.rpglink.com>

On Thu, 21 Dec 2000 02:44:29 GMT, Will Ware <wware at world.std.com> wrote:
>>     Alright, maybe I am missing something but that does nothing at all.

>The portion "print string.join(.....os.linesep)" prints a Christmas

    Right, I get that part.  When I C&P&run it all I get is the Please
Wait..., nothing more.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From mal at bewoner.dma.be  Sun Dec 24 16:17:41 2000
From: mal at bewoner.dma.be (Lieven Marchand)
Date: 24 Dec 2000 22:17:41 +0100
Subject: Why is Python slow? (was Re: One Python 2.1 idea)
References: <mailman.977590932.19335.python-list@python.org> <9237md$orm$1@nnrp1.deja.com> <924bjt0e3n@news1.newsguy.com> <m38zp6nk8k.fsf@localhost.localdomain> <925mpe$jjg$1@panix6.panix.com>
Message-ID: <m3g0jdcydm.fsf@localhost.localdomain>

aahz at panix.com (Aahz Maruch) writes:

> In article <m38zp6nk8k.fsf at localhost.localdomain>,
> Lieven Marchand  <mal at bewoner.dma.be> wrote:
> >
> >I think Common Lisp shows that getting very good speed in Python is
> >quite feasable. The original CMUCL implementers weren't such a large
> >group. Their highly optimising compiler (which incidentally is also
> >called Python) has on occasion beaten FORTRAN at numerics. I don't
> >know why some people in the Python community think compiling Python is
> >such a problem. Practically all the problems have been tackled and
> >solved 20 years ago in the Lisp community.
> 
> One of the reasons that gets brought up less often than it should is
> that Guido is somewhat fanatical about also keeping the CPython
> implementation clean and simple.  I'm pretty sure that -- Ghu forbid it
> become necessary -- I could manage to maintain CPython if I had to.
> 
> This leaves less room for optimization than many people think.

Certainly. Also the fact that writing/maintaining a compiler for all
the platforms CPython runs on would be a large effort. CMUCL is
written in itself and is notoriously hard to bootstrap or retarget to
a new OS, let alone a new CPU.

-- 
Lieven Marchand <mal at bewoner.dma.be>
Lambda calculus - Call us a mad club


From moshez at zadka.site.co.il  Sat Dec 23 14:22:00 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Sat, 23 Dec 2000 21:22:00 +0200 (IST)
Subject: A Q concerning map (a comparison to Maple V)
In-Reply-To: <921sil$moe$1@newsreaderg1.core.theplanet.net>
References: <921sil$moe$1@newsreaderg1.core.theplanet.net>
Message-ID: <20001223192200.C81DEA84F@darjeeling.zadka.site.co.il>

On Sat, 23 Dec 2000 10:48:30 +0100, "Franz GEIGER" <fgeiger at datec.at> wrote:
> 
> When I encountered Python a few month ago now, it made me remember Maple V.
> Maple has map, select, zip etc. like Python. But Maple handles map e.g. more
> generously in terms of argument passing. So you can apply
> f:=(x,y)->x^2+y
> to a list and an ADDITIONAL argument like so:
> map(f, L, 5);

I think what you really want is the Numeric module 
(http://numpy.sourceforge.net).
If L is a Numeric array, then

"L**2+5" *already* does the right thing. In general, Numeric is a Python
answer to the matlab-like languages, of which Maple V is one AFAIR.
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From dnew at san.rr.com  Wed Dec 13 19:03:25 2000
From: dnew at san.rr.com (Darren New)
Date: Thu, 14 Dec 2000 00:03:25 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <r0pzoi11lhz.fsf@black.cs.washington.edu> <5jvZ5.3588$Kd1.418770@newsb.telia.net> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com> <kuu28971kp.fsf@lasipalatsi.fi>
Message-ID: <3A380E4E.8537A0FB@san.rr.com>

Erno Kuusela wrote:
> | Having
> |  x = 2 + 2
> |  print x
> | result in anything except "4" is probably a *really* bad idea.
> | AFAIK, only FORTH and really old Fortrans allow this.
> 
> agreed - but this is not something (sane) a mutable integer object
> would break.

Well, X is a variable, and you can already change the value in X.
2 is an integer.

What are we missing? I mean, the only integer that isn't mutable in that
statement is "2".

If you mean a cell (cf Eiffel) that stores a value and can get it back as an
object...
If you mean basically a shared variable, then that seems easy.

-- 
Darren New / Senior MTS & Free Radical / Invisible Worlds Inc.
San Diego, CA, USA (PST).  Cryptokeys on demand.
Personal malapropism generator free with purchase!
Steganography: The manual is hidden in the source code.
Gold mining is only profitable because of the gold found.


From fredrik at effbot.org  Wed Dec 13 12:27:50 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 13 Dec 2000 17:27:50 GMT
Subject: Self-documenting Python??
References: <x881yvcfodl.fsf@adglinux1.hns.com>
Message-ID: <qkOZ5.3820$Kd1.435493@newsb.telia.net>

nbecker at fred.net wrote:
> Has there been any proposals to integrate a documentation system with
> python?  I feel this is one feature that is sorely missing.

why not check out the current set of Python Enhancement
Proposals (PEPs)?

    http://python.sourceforge.net/peps/

among others, you'll find:

    http://python.sourceforge.net/peps/pep-0233.html
    Python Online Help

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From wcuddy at lserv.ja10629.home  Fri Dec  8 18:12:25 2000
From: wcuddy at lserv.ja10629.home (Wayne Cuddy)
Date: Fri, 08 Dec 2000 23:12:25 GMT
Subject: using execfile() in class/instance methods
Message-ID: <slrn932qna.rtk.wcuddy@lserv.ja10629.home>

I am attempting to use the execfile() function to "source" a file that  
has valid python code to create a diction object: 
 
bn { 'key' : 'value1', 
        'key2' : 'value2' } 
 
I am issueing this command from a class constructor method and would 
like bn to be bound the class instance.  So in the future I can say: 
        print self.bn['key'] 
 
Is this possible?  I assume I am screwing up the global() and locals()  
arguments to the function which I can say that I don't fully 
understand.  Also is there another better/worse way to do this???just 
so I can see what else is out there.. 
 
Thanks in advance, 
 
Wayne 




From joseamartin at venmex.com  Thu Dec  7 11:12:28 2000
From: joseamartin at venmex.com (Jose Martin)
Date: Thu, 7 Dec 2000 12:12:28 -0400
Subject: PIL binary for Python 2.0 for Windows ?
References: <3a2f472f$0$235@hades.is.co.za> <90nueq$6jv$1@nnrp1.deja.com>
Message-ID: <t2ve715d0d2j84@corp.supernews.com>

I want the dlls
thanks...
joseamartin at venmex.com


<rhamorim at my-deja.com> escribi? en el mensaje de noticias
90nueq$6jv$1 at nnrp1.deja.com...
> I have the DLLs compiled statically with jpeg and zlib, although I
> couldn't link imagingtk.dll statically to tcl/tk - thus you need
> tcl83.dll and tk83.dll on your path.
>
> I need a place though to make them available. If you have a place (or
> want me to email it to you) just answer this message (on the newsgroup -
>  I rarely check my deja account).
>
> Best regards,
>
> Roberto Amorim
>
> In article <3a2f472f$0$235 at hades.is.co.za>,
>   "Leon Booyens" <leonb at pqafrica.co.za> wrote:
> > Where can I find a binary distribution for Windows of the Python Image
> > Library 1.1.1. (It is not as yet available from Pythonware as a
> binary and I
> > unfortunatly have no means of compiling the code on a windows
> platform)
> >
> > Any help would be appreciated.
> >
> > Thanks
> > Leon
> > (leonb at pqafrica.co.za)
> >
> >
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.




From nicmila at idoox.com  Thu Dec 14 11:33:33 2000
From: nicmila at idoox.com (Miloslav Nic)
Date: Thu, 14 Dec 2000 17:33:33 +0100
Subject: [Zvon Announcement] DOM Factory - version 0.9
Message-ID: <3A38F65D.7C5CAFAC@idoox.com>

Jiri Znamenacek today published at Zvon:
http://zvon.org/xxl/JSDOMFactory/

As many of us are fighting with browsers (and DOM) quite
often, I hope you will find it useful. Jiri did an excellent job.

It can be downloaded under "GNU Free Documentation License" as all Zvon
materials:
http://zvon.org/index.php?nav_id=3


Features

Provide an easy and handy lookup for properties of objects in JavaScript
for loaded document and the same functionality for DOM objects. 
Names of properties are clickable and colour-differentiated so
navigation is intuitive and easy. 
Properties can be highlighted (also links to appropriate references are
provided). 
Found objects can be easy added (by click) for another lookup. 
Dislayed output can be filtered in various ways. 
Simple DOM 'debugger' for testing DOM in loaded document. 
Simple JavaScript 'debugger' for evaluating JavaScript expressions 'on
the fly'. 

-- 
******************************************
<firstName> Miloslav </firstName>    
<surname>   Nic      </surname>     

<mail>    nicmila at idoox.com    </mail>   
<support> http://www.zvon.org  </support>
<zvonMailingList> 
    http://www.zvon.org/index.php?nav_id=4 
</zvonMailingList>


From trentm at ActiveState.com  Thu Dec  7 11:27:29 2000
From: trentm at ActiveState.com (Trent Mick)
Date: Thu, 7 Dec 2000 08:27:29 -0800
Subject: Problem with freeze.py in ActivePython 2.0
In-Reply-To: <3A2FB744.A942AD46@nobjects.com>; from amb@nobjects.com on Thu, Dec 07, 2000 at 05:13:56PM +0100
References: <3A2FB744.A942AD46@nobjects.com>
Message-ID: <20001207082729.A23558@ActiveState.com>

On Thu, Dec 07, 2000 at 05:13:56PM +0100, Are Magnus Bruaset wrote:
> 
> Hi,
> 
> I can't get freeze.py to behave under ActivePython 2.0 (Windows NT 4.0).
> The command 
> 
>    python freeze.py -p E:\Python20 hello.py
> 
> issued in E:\Python20\Tools\freeze (E:\Python20 is the root my
> installation) leads to the error message
> 
> Error: needed directory E:\Python20\lib\python2.0\config not found

I am not familiar with freeze.py but it looks to me like it is expecting a
*Linux* installation directory structure. Perhaps freeze.py only works on
Linux, or perhaps there is a command line option to signify window (though I
don't know why it would require that).

The <sys.prefix>/lib/python2.0/config is the path to the directory containing
the Python 2.0 library (i.e. the Linux equiv of the python20.lib file) on a
Linux installation.

Trent


-- 
Trent Mick
TrentM at ActiveState.com



From joconnor at cybermesa.com  Thu Dec 21 16:22:56 2000
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Thu, 21 Dec 2000 21:22:56 +0000
Subject: Contact
Message-ID: <3A4274B0.DC6E43DA@cybermesa.com>

Keith,

>I'll be at church tomorrow night for worship team
>rehearsal. Will you be there? 

I'll be there; I'll bring the material

>Also, Tuesdays are bad
>for me. I'm commited to mentoring a 13 year old boy,
>named Peter, every Tuesday night.

OK.  It was just a random date next week.  If Wednesday would work for
you, that's a good date for us to.

If not, I have backup plans :)

>Lastly, could you also get a hold of one of this lady's recordings. I'd
>like to hear a representation of her entire preformance before commiting any time to the project.

Understandable.  I have a tape she put together for you.  What the tape
is all the songs she's doing for this project.  Mostly it's just a rouch
cut with vocals and accompianment keyboard and drums.  The drums are
taken from a library; the keyboard parts are pretty basic, mostly just
chord and rhythm.  So you can get a good idea about her singing voice
and general musical style; but the arrangements are sparse and we're
going to be fleshing them out over then near future.

One thing I would like to do is have a chance to 'jam' through the songs
with you and her to get the arrangements nailed down. 

I also have lyric sheets.


Take care,

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

"God himself plays the bass strings first when He tunes the soul"



From deha at poczta.onet.pl  Tue Dec 12 08:47:19 2000
From: deha at poczta.onet.pl (DeHa)
Date: Tue, 12 Dec 2000 14:47:19 +0100
Subject: A question.
References: <9152a0$acb$1@panorama.wcss.wroc.pl> <9156h002uh2@news2.newsguy.com>
Message-ID: <915a7j$fe0$1@panorama.wcss.wroc.pl>

Alex Martelli napisa?(a) w wiadomo?ci: <9156h002uh2 at news2.newsguy.com>...
>"DeHa" <deha at poczta.onet.pl> wrote in message
>news:9152a0$acb$1 at panorama.wcss.wroc.pl...
>
>first = range(1,100001)
>

I try this, but it doesn't work

>
>> for i in first:
>>     if i != []:
>> #until end of the list
>
>i will never equal an empty-list, as first's items are just integers.
>

Well, I don't konw why it happend, but when there were no such line the
program doesn't work.

>Dynamical, interpreted languages such as Python are
>not *mostly* about performance -- but hundreds or even
>thousands of times of slowdown/speedup can easily be
>had in simple cases by paying a _little_ attention to
>performance issues.  I think it can be worth it...
>

Thanks for advice. Next time I will try to find better performance issues:)

>Alex
>
>
>




From egibson at connect.com.au  Wed Dec  6 21:14:02 2000
From: egibson at connect.com.au (Evan Gibson)
Date: Thu, 7 Dec 2000 13:14:02 +1100
Subject: DCOracle Memory Leak?
In-Reply-To: <m3g0k1qih5.fsf@h0050047c4794.ne.mediaone.net>; from Brad Howes on Wed, Dec 06, 2000 at 10:43:05PM +0000
References: <m3g0k1qih5.fsf@h0050047c4794.ne.mediaone.net>
Message-ID: <20001207131402.19512@connect.com.au>

On Wed, Dec 06, 2000 at 10:43:05PM +0000, Brad Howes wrote:
> Has anyone seen a Python process using DCOracle grow and grow with each fetch
> from an Oracle database? I can write a really simple loop that does a fetch but
> does not save the value, and the Python process keeps growing. Are there known
> bugs with Oracle 8i client libraries for Linux?

The main thing we saw, but with Oracle 7.3.4 instead on Solaris, was
DCOracle queries used to take 15 minutes compared to 5 minutes with
Anthony Baxter's  oracledbmodule...
Something changed (we never determined what) and suddenly DCOracle queries
(using DCOracle outside of zope) took a DAY compared 5 minutes with the
oracledbmodule.
The weirdest thing of all, though, was we only had to IMPORT DCOracle in a
script to get a decrease in performance even if there was not a single
reference to it apart from the import statement.


-- 
     Evan ~ThunderFoot~ Gibson    ~ nihil mutatem, omni delendum ~
                Walking on water... over pavement;
           Treading on miracles ground into everyday dust.



From teyc at bigfoot.com  Sat Dec 23 07:09:34 2000
From: teyc at bigfoot.com (Chui Tey)
Date: Sat, 23 Dec 2000 22:09:34 +1000
Subject: Implementation of diff algorithms in python?
References: <921ac8$lug$1@bunyip.cc.uq.edu.au> <921h7j$1pj$1@panix6.panix.com>
Message-ID: <9224p7$mke$1@bunyip.cc.uq.edu.au>

Specifically to compare where small changes in a binary file might have
occured. I'm hoping to diff .pyc files to see if running code can be
modified on the fly.


Aahz Maruch <aahz at panix.com> wrote in message
news:921h7j$1pj$1 at panix6.panix.com...
> In article <921ac8$lug$1 at bunyip.cc.uq.edu.au>,
> Chui Tey <teyc at bigfoot.com> wrote:
> >
> >Are there any implementations of diff algorithms in python?
>
> What kind of "diff"?  Are you talking about the standard file comparison
> tool or something else?
> --
>                       --- Aahz (Copyright 2000 by aahz at pobox.com)
>
> Androgynous poly kinky vanilla queer het    <*>
http://www.rahul.net/aahz/
> Hugs and backrubs -- I break Rule 6
>
> '"Crisp" is a good quality for crackers; less so for pot roast.'  --pnh




From aahz at panix.com  Fri Dec 15 15:15:16 2000
From: aahz at panix.com (Aahz Maruch)
Date: 15 Dec 2000 12:15:16 -0800
Subject: Is Stackless and uthread.py SMP friendly?
References: <91br9v$qa4$1@nnrp1.deja.com> <ku1yva5ocq.fsf@lasipalatsi.fi> <91dp4j$5fj$1@news.gte.com> <91dt39$fbk$1@nnrp1.deja.com>
Message-ID: <91du4k$gj1$1@panix6.panix.com>

In article <91dt39$fbk$1 at nnrp1.deja.com>,  <jbalding at hotmail.com> wrote:
>
>Okay. My next question: Is there a simple way of taking advantage of
>multiple processors using Python, or is this something that is going to
>be like pulling teeth? In other words, are there existing packages that
>allow relatively high-level, transparent access to multiple processors?

Yes.  Use threading.Thread().  The caveat is that Python has a global
interpreter lock that prevents more than one thread from running when
using pure Python code (e.g. "for i in range(1000): pass").  You can
solve this either by using calls that release the lock (e.g.
file.read() and many other I/O calls) or by writing a C extension that
explicitly releases the lock.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

"It's 106 miles to Chicago.  We have a full tank of gas, a half-pack of
cigarettes, it's dark, and we're wearing sunglasses."  "Hit it."


From pbjorn at uswest.net  Mon Dec 11 12:30:21 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Mon, 11 Dec 2000 10:30:21 -0700
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <912ml00a5o@news2.newsguy.com>
Message-ID: <3A350F2D.D5C0D24C@uswest.net>

Alex Martelli wrote:

> [snip]
> Anyway, in his landmark 1974 article on the Communications of
> the ACM, "Structured Programming WITH Go-To Statements", Dr.
> Knuth made the point that the generally desired form of a loop
> statement is (in pseudo-language):
>
>     loop
>         statements for part 1
>         exit when condition
>         statements for part 2
>     repeat
> [snip]

While this is certainly true, and it is a good choice for an internal
representation of loops if you're writing a compiler... I think the
original poster was more interested in the special (yet common) case
where "statements for part 1" is a trivial expression which results in
"condition". Using ":=" for assignment expression, that would look like
("pseudo-python"):

    while line := fp.readline():
        part2(line)

which is clearer for most (but certainly not all) users. There are many
"workarounds" for this "deficiency" in Python, none of which are both
fast _and_ elegant.

There's also another reason why assignment expressions are frequently
requested:

    m = re.match(...)
    if m:
        foo(m)

which according to some, would be clearer as:

    if m := re.match(...):
        foo(m)

The workarounds for the "while" case generally do not apply to the "if"
case. In any case <wink>, this has repeatedly been discussed to death.

Status quo: The BDFL likes the "while 1:" construct and doesn't like
assignment expressions, so nothing is likely to change.

-- bjorn




From mwh21 at cam.ac.uk  Wed Dec 13 21:06:58 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 14 Dec 2000 02:06:58 +0000
Subject: Bug: Readline getting stuck on Linux and Solaris
References: <lxk8zqwupy2.fsf@ioasun3.epfl.ch> <90m53k$ovh$1@nnrp1.deja.com> <m3u28h48ja.fsf@atrus.jesus.cam.ac.uk> <slrn92u8al.76b.quinn@lira.ugcs.caltech.edu> <m3r93k4kqk.fsf@atrus.jesus.cam.ac.uk> <slrn932eoj.d1j.quinn@ngwee.ugcs.caltech.edu> <m3aea12o2z.fsf@atrus.jesus.cam.ac.uk> <slrn93fsoc.5pr.quinn@dinar.ugcs.caltech.edu>
Message-ID: <m3wvd321tp.fsf@atrus.jesus.cam.ac.uk>

quinn at dinar.ugcs.caltech.edu (Quinn Dunkan) writes:

> On 12 Dec 2000 23:53:56 +0000, Michael Hudson <mwh21 at cam.ac.uk> wrote:
> >quinn at ngwee.ugcs.caltech.edu (Quinn Dunkan) writes:
> >> (gdb) bt
> >> #0  0xfa3ab88 in __ioctl () at lio.c:201
> >> #1  0xfa434f0 in ioctl () at nftw.c:184
> >> #2  0x5ffd2a74 in rl_restart_output ()
> >> #3  0x5ffdd0f4 in _rl_clean_up_for_exit ()
> >> #4  0x5ffdd860 in rl_cleanup_after_signal ()
> >> #5  0x5ffdd304 in _rl_erase_entire_line ()
> >
> >Can I see some more of this, please?
> 
> Here are some more details tracebacks (|fmt'ed to shut up slrn):

Ta.
 
> I get this trace when I do a ^C on an Indigo2 (IRIX64 6.5), even if I do
> the ^C after a 'while 1: pass'.  

That's interesting.  This would seem to support my theory that
readline is attempting to restore Python's SIGINT handler but failing
and reraising SIGINT, thus calling rl_signal_handler again.

The fact that you get this sort of thing if the while 1: pass case
sounds very much like a readline bug - you just shouldn't be ending up
in readline's signal handlers if you're not in a call to readline!
Though it might well be a bug somewhere else causing problems for
readline - but I doubt it's in Python.  Are there any other
multi-threaded apps that you know of that use readline?  Do they work?
If someone with an account on an IRIX box could whip up a very simple
threaded C program that calls readline, that'd perhaps eliminate a few
problem areas.

> Another Indigo2 with the same stats gives:
> (dbx) where
> >  0 _getpid(0x1, 0x7fcd30b0, 0x0, 0x0, 0x0, 0x1, 0x0, 0xc374f0c)
> ["/xlv47/6.5.8m/work/irix/lib/libc/libc_n32_M4/proc/getpid.s":15, 0xfabb9f8]
>    1 <Unknown>() [< unknown >, 0x5ffdd33c]
> 
> when it locks after a while 1 ^C, and
> 
> >  0 __ioctl(0x0, 0x54d5, 0x7fabcb40, 0x0, 0x0, 0x0, 0x0, 0x0)
> ["/xlv47/6.5.8m/work/irix/lib/libc/libc_n32_M4/sys/ioctl.s":20, 0xfafebf8]
>    1 _ioctl(0x0, 0x54d5, 0x7fabcb40, 0x0, 0x0, 0x0, 0x0, 0x0)
> ["/xlv47/6.5.8m/work/irix/lib/libc/libc_n32_M4/sys/ioctlSCI.c":28, 0xfb0789c]
>    2 ___new_tcgetattr(0x0, 0x54d5, 0x7fabcb40, 0x0, 0x0, 0x0, 0x0, 0x0)
> ["/xlv47/6.5.8m/work/irix/lib/libc/libc_n32_M4/term/new_tcgetattr.c":26,
> 0xfb0aa60]
>    3 <Unknown>() [< unknown >, 0x5ffd22c8]
> 
> (varies, but always begins with the ioctls) after a ^C after the
> prompt first appears.

Well, these could be more or less the same traces but with less error
info, couldn't they?

> Ok, on an Indy1 (plain IRIX 6.5), I get the same sort of thing.
> Which is weird, because I coulda sworn when I tried it before,
> 'while 1: pass' ^C worked right on the indy but not on the indigo.

Well, I'll take unexplained consistency if I can find it.

> Oh, and I used dbx because gdb isn't on all the machines, and it
> seems to give more info anyway (maybe gdb doesn't agree with sgi cc
> compiled libs).  The second indigo2 and the indy compiled python and
> readline with gcc, while the first indigo2 uses sgi cc, so perhaps
> that accounts for the more complete, consistent, and less confusing
> info from the first one.

Oh right, that seems like a good explanation.  Good, things aren't
being totally random...

> All of them lock up on ^C, even under 'while 1: pass'.  They all
> work correctly if readline is not imported, or if I manage to hit ^C
> before readline is imported.  Two of them work fine when threads
> aren't compiled in, and I assume the third one would too, but I
> haven't tested it.

Seems very likely.

Does Python also hang if you kill -INT it from another terminal?  What
about, say, kill -QUIT?  The readline docs say:

Readline contains an internal signal handler that is installed for a
number of signals (SIGINT, SIGQUIT, SIGTERM, SIGALRM, SIGTSTP,
SIGTTIN, and SIGTTOU).

so I'm wondering if all of those will hang Python...

> I also grepped through the preprocessed source of readline, and
> found sigactions() but no signals() (well, except in the included
> protos).  I also tried readline-2.2.1; no luck.

OK, so this might not be readline's fault.

> But that first traceback is nice, because it manages to implicate signals,
> readline, and libpthread all at once :)

Yes.  On an operating system I have no experience with.  We'll get
there yet :-).

> ... just what we need!  Two compilers, two ABIs, 32 bit and 64 bit arches,
> multiple library versions, and inconsistent errors!  Irix is such fun.

Sounds it.  Is there a readline mailing list anywhere?  I wonder if
moaning to them would get the response "oh, that!  it's an Irix
bug...".  Although pthreads may be more likely.

> >-- 
> >  In general, I'd recommend injecting LSD directly into your temples,
> >  Syd-Barret-style, before mucking with Motif's resource framework.
> >  The former has far lower odds of leading directly to terminal
> >  insanity.                                            -- Dan Martinez
> 
> Y'know, those random quote sigs can be eerily apropos.  "terminal" insanity
> indeed.

Indeed.

Cheers,
M.

-- 
  In case you're not a computer person, I should probably point out
  that "Real Soon Now" is a technical term meaning "sometime before
  the heat-death of the universe, maybe". 
                                     -- Scott Fahlman <sef at cs.cmu.edu>


From dale at out-think.NOSPAMco.uk  Thu Dec  7 10:27:29 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Thu, 07 Dec 2000 15:27:29 +0000
Subject: Printing modules contents
References: <3A2F84CC.3056C755@hotmail.com> <m3itow4cmi.fsf@atrus.jesus.cam.ac.uk> <3A2FA5C4.CC479A19@hotmail.com>
Message-ID: <c2bv2t0c0er7d97i5ouema4eqs89emqvgt@4ax.com>

joonas <keisari_ at hotmail.com> wrote:

>I tried this but it didnt worked.
>
><clip>
>import sys
>list = dir(sys)
>for x in list:
>  print x," = ",sys.x
></clip>
>
>Joonas.
>
>
>Michael Hudson wrote:
>> 
>> joonas <keisari_ at hotmail.com> writes:
>> 
>> > How can I make a script that would print modules contents like below,
>> >
>> > sys
>> >   sys.__name__ = "sys"
>> >   sys.stdout = "<PyShell.PseudoFile instance at 7fd530>"
>> >   sys.winver = "1.5"
>> >   sys.etc = "spam"
>> 
>> Think "dir" & "repr".  Shouldn't be hard.
>> 
>> Cheers,
>> M.
>> 
>> --
>> 39. Re graphics:  A picture is worth 10K  words - but only those
>>     to describe the picture. Hardly any sets of 10K words can be
>>     adequately described with pictures.
>>   -- Alan Perlis, http://www.cs.yale.edu/homes/perlis-alan/quotes.html


Try this:

for x in list:
... 	print 'sys.' + x, '=', sys.__dict__[x]


--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants


From tino at wildenhain.de  Sun Dec 24 05:27:26 2000
From: tino at wildenhain.de (Tino Wildenhain)
Date: Sun, 24 Dec 2000 11:27:26 +0100
Subject: Some Python 2.1 ideas
References: <mailman.977590932.19335.python-list@python.org> <Cw616.23054$qN6.3308205@typhoon.we.rr.com> <3A452881.7686E717@uswest.net>
Message-ID: <3A45CF8E.6BE7A8F9@wildenhain.de>

Hi Bjoern,

Bjorn Pettersen wrote:
> 
> string.chomp() (a better name would be nice) needs to handle all of:
> 
>  - "foo\n"
>  - "foo \n"
>  - "foo \r\n"
>  - "foo \n\r"
>  - "foo"

In all cases I came across such a problem, string.replace() did its job.
replacing \n with '' and \r with '' looks a bit ugly but works perfectly
:-)

> 
> etc. correctly, which means it should delete between zero and two bytes
> (characters) from the end of the string. This is very different from
> removing leading or trailing whitespace (whitespace can be significant
> you know <wink>).

Regards
Tino



From peter at engcorp.com  Sun Dec 31 10:09:14 2000
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 31 Dec 2000 10:09:14 -0500
Subject: why no "do : until"?
References: <3A4E95FD.D6C737A7@engcorp.com>, <92jj3j$ff7$1@nnrp1.deja.com> <slrn94shbe.150.grey@teleute.rpglink.com> <X8s36.2731$Qb7.305339@newsb.telia.net> <slrn94smds.150.grey@teleute.rpglink.com> <FOt36.17$vd1.2224@e420r-atl2.usenetserver.com> <slrn94t1vj.150.grey@teleute.rpglink.com> <20001231224925.399F9A84F@darjeeling.zadka.site.co.il>
Message-ID: <3A4F4C1A.FF5F0FED@engcorp.com>

Moshe Zadka wrote:
> 
> On Sat, 30 Dec 2000 21:12:13 -0500, Peter Hansen <peter at engcorp.com> wrote:
> 
> > The advice that runs
> > contrary to your 10+ years seems very reasonable to my 20+ years of
> > programming experience, and apparently many others with a variety of
> > experiences agree.
> 
> oooooh, oooooh, my dick is longer.
> I've been programming with it for, like, 100+ yrs. I haven't really
> got an opinion about /F's advice -- I just wanted to show off.

If you've been programming for 100+ years, I doubt your dick is very long
at all any more... all those X-rays, you know... :)

(Was there a point to your post?  I'm afraid I missed it.)



From robin at alldunn.com  Wed Dec 27 23:03:10 2000
From: robin at alldunn.com (Robin Dunn)
Date: Wed, 27 Dec 2000 20:03:10 -0800
Subject: fcgi.py
References: <3a4a865c.1325844015@localhost> <3a4a90ab.1328483109@localhost>
Message-ID: <lYy26.91$9Ow6.17498496@news.randori.com>

"Victor Muslin" <victor at prodigy.net> wrote in message
news:3a4a90ab.1328483109 at localhost...
> I just found an old posting that says that socket.fromfd() is not
> implemented on Windows platform (though Python documentation does not
> say so). Is there a FastCGI solution for Python for WindowsNT???
>

The fcgi.py module is written such that it expects to be started from a
"FastCGI Environment" such as from mod_fastcgi in Apache or from cgi-fcgi.
On Unix this means that the file descriptor for stdin is a listening socket
pre-created by the parent process that the webserver will connect to in
order to send requests to your fastcgi application.  It's done this way so
mod_fastcgi can create multiple processes all listening on the same socket,
not just the same port, but the same socket.  This is definitly a
unix-thing.

On the other hand if you don't mind losing the multiple process capability,
fcgi.py could probably be tweaked to create its own listening socket instead
of expecting it to already be there.  It could then be started up
independently of the webserver, and on any machine.  With creative use of
threads you can even regain most of the benefit of having multiple processes
running, (start a new thread, or allocate one from a thread pool, for every
connection to the socket.)

--
Robin Dunn
Software Craftsman
robin at AllDunn.com
http://wxPython.org     Java give you jitters?
http://wxPROs.com        Relax with wxPython!





From aleaxit at yahoo.com  Thu Dec 21 10:13:26 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 21 Dec 2000 16:13:26 +0100
Subject: Best documentation for writing wrappers for C++ classes
References: <9EB39E4E3AF64120.C259930B9E4BEFF5.1C1A4B99EAA59011@lp.airnews.net>
Message-ID: <91t6n802idq@news2.newsguy.com>

"David L. Humphrey" <dhumphrey at acm.org> wrote in message
news:9EB39E4E3AF64120.C259930B9E4BEFF5.1C1A4B99EAA59011 at lp.airnews.net...
> I have several C++ classes that I would like to call from Python code.  I

Not a direct response to your question, but, have a look at
www.boost.org -- the "Boost Python" library (nee py_cpp) may
be just what you need to make your life much easier (the older
but still good CXX is also a possibility).

> have reviewed the "Extending and Embedding the Python Interpreter" as well
> as selected portions of "Programming Python" but worry a little about the
> currency of these sources as well as the general tone of extending with C
> code rather than C++ classes.  Is there a document that is recognized as
the
> standard reference for this information?

"Programming Python" is indeed long in the tooth.  The "Python
Essential Reference", and the online docs, may be better bets.
But they're generally oriented to C... not to C++ specifically.


Alex





From joconnor at cybermesa.com  Sun Dec 17 10:56:05 2000
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Sun, 17 Dec 2000 15:56:05 +0000
Subject: Name scoping class names
Message-ID: <3A3CE215.95AD200F@cybermesa.com>

All,

	Is there a standard practice for scoping class names?  Such as using a
project specific prefix? I'm working on some classes for a project where
some of the classes may be general purpose enough to release, but I'm
concerned about naming conflicts. Is this usually handled with naming
conventions?  Or are modules usually used as athe scoping tool to avoid
naming collisions?

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

"God himself plays the bass strings first when He tunes the soul"



From uche at ogbuji.net  Fri Dec  8 10:40:55 2000
From: uche at ogbuji.net (Uche Ogbuji)
Date: Fri, 08 Dec 2000 15:40:55 GMT
Subject: xml.dom
References: <90q292$k9k$1@merlin.rembrandtstr.bocholt.de>
Message-ID: <3A3100FD.D6C09C1D@ogbuji.net>

Magnus Heino wrote:

> How do I build an xmldocument using something in xml.dom?

Here's an example:

from xml.dom import implementation

#Create a doctype using document type name, sysid and pubid
dt = implementation.createDocumentType('mydoc', '', '')

#Create a document using document element namespace URI, doc element
name and doctype
doc = implementation.createHTMLDocument('', 'mydoc', dt)

doc_elem = doc.documentElement

new_elem = doc.createElementNS('', 'spam')

new_elem.setAttributeNS('', 'eggs', 'sunnysideup')

new_text = doc.createTextNode('some text here...')

new_elem.appendChild(new_text)

doc_elem.appendChild(new_elem)

import xml.doc.ext

xml.doc.ext.Print(doc)



Should yield

<?xml version="1.0" encoding="UTF-8"?>
<mydoc><spam eggs="sunnysideup"><some text here.../spam></mydoc>


Though I'm typing off-head.

Does that help, or do you need any other info?

Big improvements in the docs are already in the works.

Thanks.


-- 
Uche Ogbuji
Personal:   uche at ogbuji.net		http://uche.ogbuji.net
Work:       uche.ogbuji at fourthought.com	http://Fourthought.com


From R.Brodie at rl.ac.uk  Tue Dec 12 06:56:33 2000
From: R.Brodie at rl.ac.uk (Richard Brodie)
Date: Tue, 12 Dec 2000 11:56:33 -0000
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <29lZ5.138780$U46.4638547@news1.sttls1.wa.home.com>
Message-ID: <9153pg$rh6@newton.cc.rl.ac.uk>

"Greg Jorgensen" <gregj at pobox.com> wrote in message
news:29lZ5.138780$U46.4638547 at news1.sttls1.wa.home.com...

> FORTRAN programs generally use I, J, and K as loop indices because variables
> starting with I, J, K are implicitly integers (type declarations are
> optional in FORTRAN).

We've been here before: the FORTRAN notation derives from the mathematical
convention of using i,j,k (more rarely l,m,n) as vector indices. It's no more an
invention of FORTRAN than using x,y,z for axes.






From mjjacob at my-deja.com  Wed Dec 13 11:42:55 2000
From: mjjacob at my-deja.com (mjjacob at my-deja.com)
Date: Wed, 13 Dec 2000 16:42:55 GMT
Subject: Problem with += (augmented assignment)
References: <9188og$q57$1@nnrp1.deja.com>
Message-ID: <9188ue$q9m$1@nnrp1.deja.com>

Of course I tried:
x = 0
while x <= 42:
   x += 1

I screwed up the TAB when posting

In article <9188og$q57$1 at nnrp1.deja.com>,
  mjjacob at my-deja.com wrote:
> Hi folks!
>
> I tried this under Python 2.0 running on Win 2k:
>
> x = 0
> while x <= 42:
> 	x += 1
>
> Python always complains a Syntax Error: Invalid
> Syntax on this:
>
>     x += 1
>        ^
> SyntaxError: invalid syntax
>
> Augmented assignment only seems to work when used
> in first indentation level (script level).
> Otherwise, I always get that syntax error, it
> does not matter whether I am using if, while or
> whatever.
>
> Is there something terribly wrong with the +=
> operator or is that all my fault?
>
> Help appreciated.
>
> Sent via Deja.com
> http://www.deja.com/
>


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


From rvprasad at cis.ksu.edu  Wed Dec 20 12:34:15 2000
From: rvprasad at cis.ksu.edu (Venkatesh Prasad Ranganath)
Date: 20 Dec 2000 11:34:15 -0600
Subject: Problems compiling on Sunos5
Message-ID: <m366kfhu94.fsf@boss.dreamsoft.com>

Hi,

I am trying to compile Python 2.0 on Sunos 5.  I am able to compile but many
functions are not being supported.  When I execute configure, the list of
functions checked by configure are reported as not supported but all those
functions are available.  For example, strftime() is reported as not available.

Is this a problem with the configure script or something I have to specify when
configuring.  My configure command was as follows

configure --with-threads=/usr/lib --with-cycle-gc --without-libdb

Can somebody tell me how can it be solved?  

-- 
Venkatesh Prasad Ranganath


From tyler at tylereaves.com  Mon Dec 18 14:37:46 2000
From: tyler at tylereaves.com (Tyler Eaves)
Date: Mon, 18 Dec 2000 14:37:46 -0500
Subject: float to string conversion
References: <91lmt3$4b8$1@nnrp1.deja.com>
Message-ID: <qrps3t82d5rr5gjbnleqdmt9akerprpldb@4ax.com>

On Mon, 18 Dec 2000 19:00:54 GMT, deragon at 8d.com wrote:
return MSISDN + " check FAILED:  $" + str(balance) + errormsg
>Greetings.
>
>
>  I am embarrassed to ask this since it is trivial, but
>  I cannot figure out how to convert a float to a string (obviously,
>  I am new to python).  Somehow, I could not find in the 2.0
>  documentation the syntax to perform the operation.
>
>  The following in my function:
>
>    return MSISDN + " check FAILED:  $" + balance + errormsg
>
>  where balance is a floating number and errormsg is a string
>  generates the following error:
>
>    TypeError: cannot add type "float" to string
>
>
>Sincerely,
>Hans Deragon
>
>
>Sent via Deja.com
>http://www.deja.com/

---
Tyler Eaves

Visit Ultra Coaster Central!
The internet's largest repository
of Ultra Coaster Tracks!

http://www.ultracoastercentral.com


-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----


From ssthapa at harper.uchicago.edu  Tue Dec 12 03:25:37 2000
From: ssthapa at harper.uchicago.edu (Suchandra Thapa)
Date: Tue, 12 Dec 2000 08:25:37 GMT
Subject: How to start a process?
References: <3A354E3E.1545D045@nonospamraqia.com>
Message-ID: <slrn93bfu3.flb.ssthapa@localhost.localdomain>

David Lees <DavidL at nonospamraqia.com> wrote:
>I would like to start up a number of simultaneos processes from a Python
>script.  Can someone explain or refer me to a document that explains
>what is going on below.  As a test I am executing a script 'hack.py'. 
>Why do I get this OSError when I try using execv, but it runs when I use
>'execfile'?
>
>>>> os.execv('e:\python20\hack.py',['foo','bar'])
>Traceback (innermost last):
>File "<interactive input>", line 1, in ?
>OSError: [Errno 8] Exec format error

    The first argument in the list should be the the program's name
e.g os.execv('e:\python20\hack.py', ['e:\python20\hack.py', 'foo', 'bar'])
on unix platforms. I believe that's also the case for windows but 
may be incorrect.  
    Also does your windows install know that it should feed files 
ending in .py to the python interpeter?  You may need to setup associations
between py files and the python interpeter.
    

-- 
------------------------------------------------------------------
			    |
Suchandra Thapa             | "There are only two kinds of math books. 
s-thapaNO at SPAMuchicago.edu  | Those you cannot read beyond the first 
			    | sentence, and those you cannot read 
			    | beyond the first page."
			    |                       -C.N. Yang
------------------------------------------------------------------


From tdelaney at avaya.com  Tue Dec 12 18:57:40 2000
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 13 Dec 2000 10:57:40 +1100
Subject: DangerousExceptions (was Re[2]: [Q] Are Exceptions used that 
	much in practice?)
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B0FBCA4@natasha.auslabs.avaya.com>

Very much the Java system of "Exception" vs "Error". Anything which is an
"Error" is something which you probably can't recover from (or at least not
easily), though you may be able to fail gracefully. An "Exception" OTOH is
something which you should be able to deal with.

Tim Delaney
Avaya Australia
+61 2 9352 9079

> While at this i always wished that Exception has two subclasses:
> "NormalException" and "DangerousException" so one can use
> 
> except NormalException: return 0
> 
> and not worry about  MemoryError, KeyboardInterrupt, SystemExit,
>  ComputerExplodingError :) being improperly handled.



From m.j.fisher at ex.ac.uk  Sat Dec 16 10:26:55 2000
From: m.j.fisher at ex.ac.uk (M.J.Fisher)
Date: Sat, 16 Dec 2000 15:26:55 +0000
Subject: tar files
Message-ID: <3A3B89BF.51B2C26C@ex.ac.uk>

Hi,

I am trying to find some code for unpacking tarred files and I am not
sure why to look, does anybody have any ideas? I would also appreciate
an example if any one had one,

Thanks for your help,

Shelley



From djc at object-craft.com.au  Sat Dec 23 01:39:57 2000
From: djc at object-craft.com.au (Dave Cole)
Date: 23 Dec 2000 17:39:57 +1100
Subject: Sybase module 0.11 (Sam Rushing release) released
Message-ID: <m3bsu3od36.fsf@vole.object-craft.com.au>

What is it:

The Sybase module provides a Python interface to the Sybase relational
database system. The Sybase package supports almost all of the Python
Database API, version 2.0 with extensions.

The module works with Python versions 1.5.2 and later and Sybase
versions 11.0.3 and later.  It is based on the Sybase Client Library
(ct_* API), and the Bulk-Library Client (blk_* API) interfaces.

Changes:

- The hashing of numeric objects has been fixed to generate the same
  hash values as a Python int / long of the same value.  Thanks to Sam
  Rushing for helping with this.

>>> import Sybase
>>> hash(100)
100
>>> hash(Sybase.numeric(100))
100
>>> hash(10000000000000000L)
1877247730
>>> hash(Sybase.numeric(10000000000000000L))
1877247730

  Hash of a float value does not match.

>>> hash(Sybase.numeric(100.2))
-1654341571
>>> hash(100.2) 
-1717999926

Where can you get it:

        http://www.object-craft.com.au/projects/sybase/

- Dave

-- 
http://www.object-craft.com.au


From pyth at relax.dk  Mon Dec 11 04:56:28 2000
From: pyth at relax.dk (pyth at relax.dk)
Date: Mon, 11 Dec 2000 10:56:28 +0100 (CET)
Subject: variable variables
Message-ID: <Pine.LNX.4.21.0012111052110.5044-100000@relax.dk>

Hi :o)

I've asked this question once... I'll just try one more time ;o)

Is it possible to make variable variables? :o)

Thanx :o)

/Dan Larsen




From Bjoern.Giesler at gmx.de  Wed Dec 13 07:46:03 2000
From: Bjoern.Giesler at gmx.de (Bjoern Giesler)
Date: 13 Dec 2000 12:46:03 GMT
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com>
Message-ID: <917r2b$4n7$1@news.rz.uni-karlsruhe.de>

Hi,

In comp.lang.python israel raj thomas <israelt at optushome.com.au> wrote:
:>Do you think that it would be possible to use Python as a Smalltalk
:>replacement if an IDE existed that hid some of Python's limitations?

: Would it perhaps not be better to use Ruby ?

Not for me, for the following reasons:
a) I don't know Ruby well enough
b) Ruby is "even more" unknown in The Real World(tm) than Python
c) I think Ruby code looks ugly (what with @, #, $, no 'return' etc.)
d) I don't see any of the features Ruby has over Python as 'killer'
   features.

Regards,
		--Bj?rn


From mdtompkins at home.com  Tue Dec  5 15:47:59 2000
From: mdtompkins at home.com (Mark Tompkins)
Date: Tue, 05 Dec 2000 20:47:59 +0000
Subject: win32security: setting process privileges
Message-ID: <3A2D547F.E4FF9245@home.com>

Hi,

I'm trying to set process privileges in a thread that needs to call
win32security.LogonUser.

I need to enable

SE_TCB_NAME
SE_CHANGE_NOTIFY_NAME
SE_ASSIGNPRIMARYTOKEN_NAME

When I attempt to assign these privileges to the current thread, only
one assignment succeeds (SE_CHANGE_NOTIFY_NAME).  The MSDK win32 api
docs INDICATE that SE_CHANGE_NOTIFY_NAME is assigned by default.  The
following example is for SE_ASSIGNPRIMARYTOKEN_NAME.

>>>import win32security
>>>import ntsecuritycon
>>>import win32con
>>>import win32api
>>>
win32security.LookupPrivilegeValue(None,ntsecuritycon.SE_CHANGE_NOTIFY_NAME))

7
>>>
win32security.LookupPrivilegeValue(None,ntsecuritycon.SE_ASSIGNPRIMARYTOKEN_NAME)

23
>>>
win32security.LookupPrivilegeValue(None,ntsecuritycon.SE_ASSIGNPRIMARYTOKEN_NAME)

3
>>> flags=ntsecuritycon.TOKEN_ADJUST_PRIVILEGES |
ntsecuritycon.TOKEN_QUERY
>>> flags
40
>>>id=win32security.LookupPrivilegeValue(None,ntsecuritycon.SE_ASSIGNPRIMARYTOKEN_NAME)

>>> id
3
>>> nP=[(id,ntsecuritycon.SE_PRIVILEGE_ENABLED)]
>>> nP
[(3, 2)]
>>>
htoken=win32security.OpenProcessToken(win32api.GetCurrentProcess(),flags)

>>> htoken
268

>>> win32security.AdjustTokenPrivileges(htoken,0,nP)
Traceback (innermost last):
  File "<interactive input>", line 1, in ?
api_error: (1300, 'MyAdjustTokenPrivileges', 'Not all privileges
referenced are assigned to the caller.')
>>>

What privileges do I as a user need to be assigned, in order that the
other two privileges are enabled for me, or are grantable by me to a
process?  I am currently a member of the administrators group on the
machine that I am coding on.

thx

Mark



From jrepetto at mxmlab.com  Fri Dec  8 05:58:37 2000
From: jrepetto at mxmlab.com (Jean-Claude REPETTO)
Date: Fri, 08 Dec 2000 10:58:37 GMT
Subject: Looking for compile.py
Message-ID: <3a30bde8.94361636@news.wanadoo.fr>

Hi,

I am looking for the David Ascher's compile.py script to create
Microsoft Visual C++ project files. 
The link I have found on
http://www.python.org/doc/current/ext/win-cookbook.html is :

 http://starship.python.net/crew/da/compile/

But it doesn't work ((111) Connection refused) .

Thanks for your help.

Jean-Claude Repetto
Vallauris, France






From nickatvideosystemdotcodotuk  Mon Dec 11 05:02:11 2000
From: nickatvideosystemdotcodotuk (Nick Trout)
Date: Mon, 11 Dec 2000 10:02:11 -0000
Subject: Executable python program?
References: <jKYY5.54278$df5.1795043@news1.crdva1.bc.home.com>
Message-ID: <3a34a653@news.xtml.co.uk>

> I am developing a python/Tk program in linux.  I want Windows users to be
> able to use this program as well.  Now...the situation I am in it will
> make it inconvenient to get the windows people to install python/tk to run
> things.  Is it possible to get a python "Executable" per say that isnt
> interpreted but is a "static" executable persay?  By this I mean static
> in the same form as a linux executable that is "static" by making no
> external library calls (or none other than system calls if that is
> needed).

You could look at PySol. It has everything you need to run a Python app in
Tk.

http://google.yahoo.com/bin/query?p=pysol&hc=0&hs=0






From un4e at rzstud1.rz.uni-karlsruhe.de  Mon Dec 11 11:15:52 2000
From: un4e at rzstud1.rz.uni-karlsruhe.de (Bjoern Giesler)
Date: 11 Dec 2000 16:15:52 GMT
Subject: Modifying running code? [HACK ALERT]
References: <912m65$pu1$1@news.rz.uni-karlsruhe.de> <3A34E7B0.495@hvision.nl>
Message-ID: <912ujo$qjv$1@news.rz.uni-karlsruhe.de>

Hi,

Hans Nowak <ivnowa at hvision.nl> wrote:
: "Python uses the ``termination'' model of error handling: an exception
: handler can find out what happened and continue execution at an outer
: level, but it cannot repair the cause of the error and retry the failing
: operation (except by re-entering the offending piece of code from the
: top)."

...which is what I want to do, except I want to replace the offending piece
of code by something else first :-) This is very interesting to know. Thanks
a lot!

: [...] Editing the code will
: be a problem too; you will need to point the program to a string or text
: file that holds the function code, [...]

This is what the traceback module does anyway -- it finds the original file
and the line that caused the exception. I only need to additionally find the
beginning and end of the function that contains the line.

Regards,
		--Bj?rn


From syver at NOSPAMcyberwatcher.com  Wed Dec 20 16:40:52 2000
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Wed, 20 Dec 2000 22:40:52 +0100
Subject: KeyboardInterrupt exception in ASP pages.
Message-ID: <91r95v$fk1$1@troll.powertech.no>

I picked up the following from Deja.

I seems I have got a similar problem on my web pages using ADO. Are there
any knowledge of the nature of this problem? workarounds, patches or
anything?

Willing to do rather a lot to avoid this problem :-)



jj at void.si wrote:

> I am able to reproduce KeyboardInterrupt in ASP page.

> On a IIS I have two ASP scripts to build html output using ADO queries.
> On a client I have c++ application with embedded IE (IWebBrowser2).

That is great! However, is it possible for you to come up with something so
_I_ can reproduce?  I am afraid I don't have access to either your ASP
sample code or C++ sample code.  I really need something that I can setup
and debug here.

Mark.






From neelk at alum.mit.edu  Wed Dec 27 20:46:25 2000
From: neelk at alum.mit.edu (Neelakantan Krishnaswami)
Date: 28 Dec 2000 01:46:25 GMT
Subject: One Python 2.1 idea
References: <mailman.977771949.17911.python-list@python.org> <9292tn$jka$1@nnrp1.deja.com> <929n9c013m3@news1.newsguy.com> <92ab9u$dou$1@nnrp1.deja.com> <92ap0i22b6r@news1.newsguy.com> <slrn94ie9k.anf.neelk@alum.mit.edu> <bTc26.210688$RP.38759664@news1.rdc1.md.home.com>
Message-ID: <slrn94lako.d5p.neelk@alum.mit.edu>

On Wed, 27 Dec 2000 02:56:07 GMT, Tim Peters <tim.one at home.com> wrote:
>[Neelakantan Krishnaswami <neelk at alum.mit.edu
>>
>>   http://starship.python.net/crew/vlad/archive/threaded_code/
>>
>> Bringing that up to date with 2.0 would be an interesting experiment.
>
> Note that at the 1998 Python Conference, John Aycock reported on an extreme
> technique that can eliminate *all* eval-loop overhead, including instruction
> decoding and dispatch costs; look for "Converting Python Virtual Machine
> Code to C" at
> 
>     http://www.foretec.com/python/workshops/1998-11/proceedings.html
> 
> The results were inconclusive (detecting a pattern here <wink>?),
> but it wasn't encouraging that pystone only got a 10% speed boost.
> That did and does seem anomalously small, but so long as "+" has to
> do pages of analysis at runtime, the time to *find* the BINARY_ADD
> code really isn't important.  Threading techniques grew up in Forth,
> where the implementation of a typical opcode consists of just a
> handful of machine instructions.

I've only ever written a Scheme VM; every opcode there is tiny too
except for function calls, which receive extreme optimization effort
(because everything is a function call in Scheme). Add some
represenatation finagling to make type-checking faster, and then
you've pretty much exhausted all the easy parts of optimization.

But IMO it's still a good idea to go for the easy-but-modest gains,
since five or six of them are just as potent as the one two-fold
super-sexy optimization. Has any work been done with Swallow since
last year? The types-sig has been dead, but turning global name
accesses into offsets like locals could be another easy 10-15%.

> god-forbid-anyone-profile-the-code-and-find-out-where-it's-really-
> spending-time<wink>-ly y'rs  - tim

I just did. The answer is SET_LINENO. 

I expected Vladimir's patch to yield perhaps a 10% increase in speed
in pure Python code. A 25% speedup in instruction dispatch (based on
Anton Ertl's threading speed measurements) times 50% time spent in
instruction dispatch (based on the technical report on the ZINC Caml
interpreter) = 12.5% speedup, rounded down to 10% due to general
pessissm.

Before I tried it, though, I took a look at the disassembled byte
codes and saw a bunch of SET_LINENOs, and out of curiousity tried
running Python with the -O option. The result was a 15% increase in
speed. Now I'm no longer in the mood to do any benchmarking of his
patch. :)


Neel


From serge at sergebeaumont.com  Tue Dec 19 07:07:49 2000
From: serge at sergebeaumont.com (Serge Beaumont)
Date: Tue, 19 Dec 2000 13:07:49 +0100
Subject: How to rewrite this as list comprehension?
Message-ID: <91nj2q$s3p$1@buty.wanadoo.nl>

Hello,

I'm trying to get my head around comprehensions. How would I rewrite this
method?

  def deadThings(self, type = None):
    deadThings = self.destroyed[:]
    for thing in self.things:
      if thing.destroyed:
        if type:
          if isinstance(thing, type):
            deadThings.append(thing)
        else:
          deadThings.append(thing)
    return deadThings


--
Serge Beaumont

Architect
IQUIP Informatica BV
serge.beaumont at iquip.nl
=====================
www.iquip.nl
www.sergebeaumont.com
=====================





From junaftnoon at nospamplzyahoo.com  Thu Dec 28 05:23:54 2000
From: junaftnoon at nospamplzyahoo.com (June Kim)
Date: Thu, 28 Dec 2000 19:23:54 +0900
Subject: Interesting behaviour of the assignment
Message-ID: <92f4gk$qk6$1@news.nuri.net>

>>> a=100
>>> b=100
>>> a is b
0
>>> a=99
>>> b=99
>>> a is b
1

Is this an intended/expected result? What is the lying structure beneath
this?



From aleaxit at yahoo.com  Sat Dec  9 07:47:49 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 9 Dec 2000 13:47:49 +0100
Subject: Python wrapper for C++ STL?
References: <3A31760E.D157EB54@erols.com> <C0gY5.126062$U46.4232859@news1.sttls1.wa.home.com> <90s3fk$j8p$1@nnrp1.deja.com> <fUgY5.126218$U46.4241063@news1.sttls1.wa.home.com> <90s9l9$njd$1@nnrp1.deja.com> <KVjY5.126920$U46.4261650@news1.sttls1.wa.home.com>
Message-ID: <90t9n902p8r@news2.newsguy.com>

"Greg Jorgensen" <gregj at pobox.com> wrote in message
news:KVjY5.126920$U46.4261650 at news1.sttls1.wa.home.com...
> <shindich at my-deja.com> wrote:
>
> > I am not saying that there is any point in wrapping STL for Python. All
> > I'm saying is that STL can be used to write C++ extensions that
> > implement specific algorithms.
>
> Sure it could, though C++ + STL = C O D E  B L O A T, so using that combo

Naah...!  It depends on your general setup.  I regularly use MSVC++6 (hardly
a top-of-the-league optimizer -- wish I could use KAI...) with ATL and STL
to
build truly tiny COMponents -- no reason the footprint should be any larger,
with a similar setup, to build Python-specific extensions (of course, a
COMmie
can be used both from Python and from other languages, and you do have the
ATL infrastructure already tuned and optimized for it -- but, alas, it ain't
cross-platform yet, at least until XPCOM should mature a bit...).

> for Python modules will likely suck up a lot of memory. If you want some
> data structure Python doesn't have--sorted associative containers, for
> example--you'd be better off writing a module in C to implement just that,
> rather than try to press the STL into service where it won't give you any
> benefit.

I consider high programming productivity to be a very significant benefit
(or else, I wouldn't be using Python!-).  If using C++ (and the Standard
C++ Library, or even specifically STL if I must -- there's a difference!-)
halves my development-effort (and the productivity benefit is often more
than that), I'd be quite willing to take a small performance hit (e.g. in
terms of footprint), IF I had to -- and I noticed that I often don't.

I consider it pretty sad that I still have to code Python extensions in C,
rather than C++, to ensure they're "taken seriously" by the Python
community, actually.  I'll do it when I must (as I'm doing in gmpy), but,
it *IS* a sad reflection on the state of the Python world:-).  Oh well,
at least it can be 'full ISO C' (commonly still mis-called ANSI C:-) these
days -- meaning the standard of 1989, of course (we wouldn't want to
be TOO modern, now would we?-).


Alex





From teyc at bigfoot.com  Mon Dec 25 16:40:18 2000
From: teyc at bigfoot.com (Chui Tey)
Date: Tue, 26 Dec 2000 07:40:18 +1000
Subject: xml-rpc server in python returning python lists...
References: <hih16.85504$yR4.2492277@news1.rdc1.tx.home.com> <925o9s$pqk$1@bunyip.cc.uq.edu.au> <HZx16.87713$yR4.2521840@news1.rdc1.tx.home.com>
Message-ID: <928evf$u2$1@bunyip.cc.uq.edu.au>

As far as I know, the python xmlrpc server conforms to the xmlrpc standard.

> cannot seem to get working.  Which one did you use with VB?

I'm not sure. I downloaded both. I'm probably using the other one...

    Set m_Client = New XMLRPCC.Client
    Private m_DOM       As DOMDocument
    Set m_DOM = m_Client.Execute(URL, "getServerName")
    m_Client.ResponseToVariant m_DOM, sName

I think mine needs VB6 + IE5 at least to work.

theArray(0) and theArray(1) should return something..., try stepping through
the vb code

(Sorry Pythonistas, probably should move over to a vb forum soon)





From sholden at holdenweb.com  Fri Dec 22 17:57:25 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 22 Dec 2000 17:57:25 -0500
Subject: namespaces & indentation
References: <920f54$m2f$1@news.mch.sbs.de>
Message-ID: <c1R06.9782$3O1.119756@e420r-atl1.usenetserver.com>

Chris Schaller <Chris.Schaller at web.de> wrote in message
news:920f54$m2f$1 at news.mch.sbs.de...
> Dear Pythoners...
>
>   I always thought that I understood Python's scoping, but then I came
across
> that code:
> > def t(x):
> >   if x == 1:
> >     c = 'yeah'
> >   return c
>
> If I execute t(1), I'll get 'yeah', but otherwise there'll be an error
> message.  According to the indentation level c is only valid within the
> if-clause, but not on the outside.  So I guess, the indentation level only
> defines the lines of code that belong to the if-clause, but has nothing to
do
> with scopes.

The indentation level has nothgin to do with scopes!  The scope of c is the
function definition, because c is bound (assigned to) within the function
body.  The interpreter therefore expects to find c in the local namespace.

So, when x == 1 you return the value 'yeah' because that's what you assigned
to c.

When x != 1 you get an error, because the interpreter (having told itself
during compilation that c is local to the function) can't find c in the
function's namespace, because you haven't bound a value to it!

>
>   That's very confusing, I've already programmed a lot in Python, but only
> due to a program error (I forgot to initialize c at the beginning of the
> function - a common error) and an unsuccessful test this happened to a
program
> after months of error-free running.
>
>   Any comments?
>
I think you understand the namespace rules quite well, you just hadn't
previously realized that names are created dynamically on assignment rather
than statically during compilation.  If you haven't assigned a value to them
you get an UnboundLocalError exception (which describes the problem quite
succinctly, don't you think?).

> bye
>   Chris...
>
Has this helped?

regards
 Steve






From frc.removethistextfirstplease at eznet.net  Thu Dec 21 15:30:28 2000
From: frc.removethistextfirstplease at eznet.net (David Fisher)
Date: Thu, 21 Dec 2000 20:30:28 GMT
Subject: audio playback
References: <3a40c236.3935118@news.frontiernet.net> <3A422D49.6FA64ECE@speech.kth.se>
Message-ID: <3a42681b.7900540@news.frontiernet.net>

On Thu, 21 Dec 2000 17:18:18 +0100, Kare Sjolander
<kare at speech.kth.se> wrote:

>One way is to use Snack, http://www.speech.kth.se/snack/, which Windows
>95/98/NT/2000, Linux, Macintosh, Sun Solaris, HP-UX, NetBSD, and SGI
[snip]
>
>Kare
>

It worked on the first try. Beautiful.

Thank you!

	- David Fisher



From aahz at panix.com  Fri Dec 22 10:22:12 2000
From: aahz at panix.com (Aahz Maruch)
Date: 22 Dec 2000 07:22:12 -0800
Subject: __init__ like in java
References: <91vosi$19h$1@reader1.imaginet.fr>
Message-ID: <91vrj4$o61$1@panix6.panix.com>

In article <91vosi$19h$1 at reader1.imaginet.fr>,
Gilles Lenfant <glenfant at equod.com.nospam> wrote:
>
>Is there an *elegant pythonic*  way to have several constructors for one
>class ?

What do you mean?  The standard way of handling this in Python is to use
default arguments:

class foo:
    __init__(self, bar=None, baz=None):
        if bar:
            ...
        if baz:
            ...
        ...

Now you can create an instance of foo() using zero, one, or two
arguments.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

I don't really mind a person having the last whine, but I do mind
someone else having the last self-righteous whine.  --Aahz


From moshez at zadka.site.co.il  Mon Dec 25 10:04:07 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Mon, 25 Dec 2000 17:04:07 +0200 (IST)
Subject: html filtering
In-Reply-To: <3A46900D.BD118A0A@wielun.dhs.org>
References: <3A46900D.BD118A0A@wielun.dhs.org>
Message-ID: <20001225150407.1A3CBA84F@darjeeling.zadka.site.co.il>

On Mon, 25 Dec 2000, QdlatY <qdlaty at wielun.dhs.org> wrote:

> Big problem in writeing message boards is filtering html tags. I use
> very simple method: Each occurence of "<" sign or ">" sign I replace
> with &lt; and &gt;. Is this safe method? Maybe You've got better ideas?

Well, it is safe as long as you remember to replace every "&" with
"&amp;" *before* you replace < and >.

-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From tcmay at got.net  Fri Dec 15 17:05:31 2000
From: tcmay at got.net (Tim May)
Date: Fri, 15 Dec 2000 14:05:31 -0800
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <tcmay-FCA22F.14073713122000@news.got.net> <918v7a02scn@news1.newsguy.com> <tcmay-04E80C.16181313122000@news.got.net> <91a1ed02a4j@news2.newsguy.com> <3A389B1A.4F33FFD4@cfmu.eurocontrol.be> <91ae8s02llf@news2.newsguy.com> <slrn93l2lc.5l.cerutti@fiad06.norwich.edu>
Message-ID: <tcmay-27C2A0.14053115122000@news.got.net>

In article <slrn93l2lc.5l.cerutti at fiad06.norwich.edu>, 
cerutti at together.net wrote:

> Alex Martelli posted:
> >[good discussion about squares and rectangels snipped]
> >
> >But that's implementation, and a well known
> >problem with inheriting-from-concrete in that
> >realm.  What I'm saying is that the problems with
> >inheritance-from-instantiable (fully concrete)
> >classes start much earlier than at implementation
> >time -- start with conceptual, ontological issues
> >regarding OOA and related classification tasks.
> 
> This problem arises from the way that our minds work. We like to
> classify things, and the classifications are incredibly
> complicated. Object Orientation doesn't model our minds very well
> due to the problems you discussed. We end up needing imaginary
> classes containinag, as practically as we can deduce, the
> commonalities in groups of objects we're modelling--for the sake
> of inheritance. In reality an ostrich and a blue-jay and a
> pengiun are all birds, but the relationships between them don't
> *really* go back to a primordial ideal BIRD. They're family
> relationships, which I don't know how to model.
> 
> Perhaps a new paradigm of "family oriented" programming will crop
> up some day?

Good point, but this is why many like "mix-ins" more than MI. 

The ostrich has certain bird-like features (bird chromosomes, feathers, 
egg-laying, etc.), but also has certain features of "ground-running 
only" things. That is, an ostrich runs on the ground like mammals or 
reptiles. 

But instead of saying that the ostrich class inherits from both the bird 
class and some other ground-running class, one would "mix-in" only the 
"ground-running-thing" items/behaviors/methods of interest. 

(And as we have seen some good arguments here in recent days, neither SI 
nor MI is necessary. Buying (or composing), as Meyer coined the term, 
are equally well-suited. In the ostrich case, the "ostrichness" of an 
ostrich may not be captured by any number of MI or mix-ins. Almost as 
efficient, perhaps, to just start from scratch.)


--Tim May

-- 
Timothy C. May         tcmay at got.net        Corralitos, California
Political: Co-founder Cypherpunks/crypto anarchy/Cyphernomicon
Technical: physics/soft errors/Smalltalk/Squeak/agents/games/Go
Personal: 1951/UCSB/Intel '74-'86/retired/investor/motorcycles/guns




From jonathan.gilligan at vanderbilt.edu  Tue Dec  5 15:40:39 2000
From: jonathan.gilligan at vanderbilt.edu (Jonathan Gilligan)
Date: Tue, 5 Dec 2000 14:40:39 -0600
Subject: win32security.AdjustTokenPrivileges
References: <3A2D3521.6010105@pixelsystems.com>
Message-ID: <90jjs7$rdd$2@news.vanderbilt.edu>

In your code, there are two things: First, you have a
typo---"provValue" where you mean "procPrivValue" and second (I don't
know if it makes a difference) you have

  AdjustTokenPrivileges (processToken, 0, (provValue,
SE_PRIVILEGE_ENABLED) )

where I have the equivalent of

  AdjustTokenPrivileges (processToken, 0, [(provValue,
SE_PRIVILEGE_ENABLED) ])

(i.e., you pass a tuple for parameter 3, where I pass a list
containing a single tuple).

Here is some code I use that successfully reboots a remote computer:

import win32net, win32netcon, win32security, ntsecuritycon, win32api
import pythoncom
import sys, time

privilegeShutdown = ntsecuritycon.SE_REMOTE_SHUTDOWN_NAME
# fill in the name of the server you want to reboot (it can be remote)
ServerName = "\\\\TargetServer"

debug = 1

def trace(msg):
    if debug:
        print msg


#
# The code below is largely stolen from Mark Hammond's
# "Programming Win32 with Python," (O'Reilly, 2000).
# If you don't have this book, you should buy it!
#
# This blatant advertisement placed here by Jonathan Gilligan
#
def AdjustPrivilege(priv, enable = 1):
    # get the process token
    flags = ntsecuritycon.TOKEN_ADJUST_PRIVILEGES |
ntsecuritycon.TOKEN_QUERY
    htoken =
win32security.OpenProcessToken(win32api.GetCurrentProcess(), flags)
    # Get the id fof the system shutdown privilege
    id = win32security.LookupPrivilegeValue(None, priv)
    #
    #
    if enable:
        trace(("Enabling privilege %s" % priv))
        try:
            newPrivileges = [(id, ntsecuritycon.SE_PRIVILEGE_ENABLED)]
        except pythoncom.com_error, (hr, msg, exc, arg):
            trace("Failed with code %d: %s. " % (hr, msg))
            if (exc == None):
                trace("There is no extended error information")
            else:
                wcode, source, text, helpFile, helpId, scode = exc
                trace("The source of the error is %s, the message is
\"%s\", More info can be found in %s (id-%d)"
                      % (source, text, helpFile,helpId))
        except win32net.error:
            trace("Failed: '%s' = %s" % (sys.exc_info()[:2]))
        except:
            trace("Failed.")
        else:
            trace("Succeeded.")
    else:
        trace("Disabling privilege %s" % priv)
        try:
            newPrivileges = [(id, 0)]
        except pythoncom.com_error, (hr, msg, exc, arg):
            trace("Failed with code %d: %s. " % (hr, msg))
            if (exc == None):
                trace("There is no extended error information")
            else:
                wcode, source, text, helpFile, helpId, scode = exc
                trace("The source of the error is %s, the message is
\"%s\", More info can be found in %s (id-%d)"
                      % (source, text, helpFile,helpId))
        except win32net.error:
            trace("Failed: '%s' = %s" % (sys.exc_info()[:2]))
        except:
            trace("Failed.")
        else:
            trace("Succeeded.")
    win32security.AdjustTokenPrivileges(htoken, 0, newPrivileges)

def RebootServer(message = "Server Rebooting", timeout = 120, bForce =
0, bReboot = 1):
    trace("Preparing to reboot server")
    AdjustPrivilege(privilegeShutdown)
    try:
        win32api.InitiateSystemShutdown(ServerName, message, timeout,
bForce, bReboot)
    except pythoncom.com_error, (hr, msg, exc, arg):
        trace("Failed with code %d: %s. " % (hr, msg))
        if (exc == None):
            trace("There is no extended error information")
        else:
            wcode, source, text, helpFile, helpId, scode = exc
            trace("The source of the error is %s, the message is
\"%s\", More info can be found in %s (id-%d)"
                  % (source, text, helpFile,helpId))
    except win32net.error:
        trace("Failed: '%s' = %s" % (sys.exc_info()[:2]))
    except:
        trace("Failed.")
    else:
        trace("Succeeded.")
    AdjustPrivilege(privilegeShutdown, 0)

def AbortReboot():
    AdjustPrivilege(privilegeShutdown)
    trace("Aborting reboot on %s" % ServerName)
    try:
        win32api.AbortSystemShutdown(ServerName)
    except pythoncom.com_error, (hr, msg, exc, arg):
        trace("Failed with code %d: %s. " % (hr, msg))
        if (exc == None):
            trace("There is no extended error information")
        else:
            wcode, source, text, helpFile, helpId, scode = exc
            trace("The source of the error is %s, the message is
\"%s\", More info can be found in %s (id-%d)"
                  % (source, text, helpFile,helpId))
    except win32net.error:
        trace("Failed: '%s' = %s" % (sys.exc_info()[:2]))
    except:
        trace("Failed.")
    else:
        trace("Succeeded.")
    AdjustPrivilege(privilegeShutdown, 0)

#
# Set the timeout to whatever timeout you want to give the user on the
target
# workstation to shut down his work before you reboot.
#
tmout = 120
message = "Preparing to reboot server. You have %d seconds to finish
your work before the reboot."
RebootServer(message, timeout=tmout)



"Patrick Blanchette" <pblanchette at pixelsystems.com> wrote in message
news:3A2D3521.6010105 at pixelsystems.com...
> I,
>    I'd like to reboot a windows NT station using a Python script.
It is
> supposed to be possible using the Python extensions for Windows.  I
try
> the following code but an exception is thrown after the call to
> AdjustTokenPrivileges:
> -----------------------------------------------
> hProcess = GetCurrentProcess()
> processToken = OpenProcessToken(hProcess, TOKEN_ADJUST_PRIVILEGES |
> TOKEN_QUERY)
> procPrivValue = LookupPrivilegeValue (None, SE_SHUTDOWN_NAME)
> AdjustTokenPrivileges (processToken, 0, (provValue,
SE_PRIVILEGE_ENABLED) )
> ExitWindowsEx(EWX_REBOOT | EWX_FORCE, 0)
> -->
> Traceback (innermost last):
>   File "<interactive input>", line 1, in ?
> TypeError: A TOKEN_PRIVILEGES object must be a tuple of
(LARGE_INTEGER, int)
> -----------------------------------------------
>
> How should I use the AdjustTokenPrivileges function?
>
> thank
>




From aahz at panix.com  Mon Dec  4 17:46:35 2000
From: aahz at panix.com (Aahz Maruch)
Date: 4 Dec 2000 14:46:35 -0800
Subject: int multiplication overflow and the future of python
References: <ghaelzoc9lp.fsf@synopsys.com>
Message-ID: <90h6sb$4s1$1@panix6.panix.com>

In article <ghaelzoc9lp.fsf at synopsys.com>,
Ram Bhamidipaty  <ramb at synopsys.com> wrote:
>
>This leads to a larger question: Which version is being maintained?
>Is active developement taking place for both 1.6 and 2.0? I've been
>using 1.6 because the features in 2.0 seem like they involve bigger
>changes to the code base and I thought I should wait until the
>initial bugs are worked out. But if all new work is only going
>into the 2.x versions then I should probably stop using 1.6...

Use Python 2.0.  Period.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

"Just because I'm selling you into slavery doesn't mean we can't be
friends."  B-movie cliche Hall of Fame  (_Bounty Hunter: 2002_)


From rturpin at my-deja.com  Tue Dec 26 22:39:51 2000
From: rturpin at my-deja.com (rturpin at my-deja.com)
Date: Wed, 27 Dec 2000 03:39:51 GMT
Subject: Proposal for attribute and method caching (was: One Python 2.1 idea)
References: <mailman.977857210.10962.python-list@python.org> <92bmb4$eif$1@nnrp1.deja.com>
Message-ID: <92boa7$frk$1@nnrp1.deja.com>

Damn! Can't get anything right the first time:

>     if x.__dict__.isIndexable:
>         if x.__class__!=_1:     # New class
>             _1 = x.__class__
>             _3 = x.__dict__.getIndex("foo")
>         if _3:
>             y = x.__dict__.getByIndex(_3)
>         else:
>             <search other namespaces>

The last four lines should be:

          try:
              y = x.__dict__.getByIndex(_3)
          except:
              <search other namespaces>

But y'all likely figured it out, anyway.

Russell


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


From mikael at isy.liu.se  Wed Dec 13 07:19:59 2000
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Wed, 13 Dec 2000 13:19:59 +0100 (MET)
Subject: Python and databases
In-Reply-To: <bCIZ5.23937$_y5.4911010@typhoon.southeast.rr.com>
Message-ID: <XFMail.001213131959.mikael@isy.liu.se>

On 13-Dec-00 Kevin B wrote:
 >  Are there any resources that have information about Python and database
 >  access? I have reviewing several book such as "Programming Python" and
 >  "Learning Python" and have not found even a chapter on Python and database
 >  access.

Well... There is

  http://www.python.org/topics/database/

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    13-Dec-00
Time:    13:19:33

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

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


From quickdry at users.sourceforge.net  Sun Dec 31 01:01:50 2000
From: quickdry at users.sourceforge.net (Steven Adams)
Date: Sun, 31 Dec 2000 06:01:50 GMT
Subject: os module - reliable?
References: <92m38n$88f$1@nnrp1.deja.com>
Message-ID: <iZz36.2795$pG5.14262@news1.rdc1.nsw.optushome.com.au>

<curtissol at my-deja.com> wrote in message
news:92m38n$88f$1 at nnrp1.deja.com...
> I am writing exercise 1 pg 212 of "Learning Python". It asks for a
> program that prints out the names of all files in a specific
directory
> and all its subdirectories. The code I have seems relatively
straight
> forward:
>
> import os, sys
>
> def ShowDirContents(dir):
>     contents = os.listdir(dir)
>
>     for file in contents:
>         if os.path.isdir(file):
>             ShowDirContents(file)
>         else:
>             print file

your code isn't updating what it uses as a new directory to search
through, sometimes it helps to trace through your code exactly what
variable changes to what value, and where, listdir() gives out only
the filename, not its absolute path.

change ShowDirContents(dir) to this and it should be fine - the code
you have doesn't need sys either.

def ShowDirContents(dir):
            for file in os.listdir(dir):
                if os.path.isdir(dir+os.sep+file):
                    ShowDirContents(dir+os.sep+x)
                else:
                    print file

hope that helps,

Steven




From shindich at my-deja.com  Sat Dec  9 01:39:08 2000
From: shindich at my-deja.com (shindich at my-deja.com)
Date: Sat, 09 Dec 2000 06:39:08 GMT
Subject: making a DLL
References: <G5ADuH.GC8@world.std.com>
Message-ID: <90sk2c$uqo$1@nnrp1.deja.com>

In article <G5ADuH.GC8 at world.std.com>,
  wware at world.std.com (Will Ware) wrote:
> This is admittedly a pretty dumb question, please bear with me, I'm
> not a Windows programmer. I've often made shared objects in Unix to
> speed up Python with the addition of code written in C. Now I want to
> do the equivalent thing in Windows. So I've written a dumb example
> C++ file, and put together an apparently adequate .def file, and told
> MS VC++ 5.0 to build a DLL. I even figured out how to set up the
> include directory (C:\Python20\include) so that it doesn't have any
> trouble finding Python.h. So far so good. Alas, when I try to link, I
> get this:
>
> Linking...
> LINK : fatal error LNK1104: cannot open file "python20_d.lib"
> Error executing link.exe.
>
> So I searched the directory tree starting at c:\Python20 for *.lib
> files, and found c:\Python20\libs\python20.lib, but no python20_d.lib
> anywhere. So I thought to try forcing the thing to use python20.lib
> instead (Project->Settings->Link(input)->Object/library modules) and
> gave it the library path. Now I get a different error:
>
> Linking...
> c:\Python20\libs\python20.lib : fatal error LNK1106: invalid file or \
>                disk full: cannot seek to 0x39ed0908
> Error executing link.exe.
>
> I know the disk isn't full, so I conclude that python20.lib is somehow
> deemed invalid. Was the Windows version of Python 2.0 built in some
way
> that Microsoft Visual C++ would consider invalid? (E.g., with
Borland's
> compiler?) Is there a fix for all this mess? Any hints appreciated,
thanks.
>
> --
> # - - - - - - - - - - - - - - - - - - - - - - - -
> # Resistance is futile. Capacitance is efficacious.
> # Will Ware	email:    wware @ world.std.com
>
;) I have seen this before.
You are building under the debug configuration. Regular Windows install
of Python installs release library only, i.e. python20.lib. The name of
the debug python library is python20_d.lib. You have two options here:
1. Build in the Release mode
2. Dowload Python sources and build them in the Debug mode. This will
produce python20_d.dll and python20_d.lib

Good luck!

Regards,

Alex Shindich
visit http://www.shindich.com/


Sent via Deja.com http://www.deja.com/
Before you buy.


From MarkH at ActiveState.com  Wed Dec 13 17:19:54 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Wed, 13 Dec 2000 22:19:54 GMT
Subject: mmap problem with dll
References: <3a366959.3497335@news-server.socal.rr.com>
Message-ID: <3A37F4CC.5070305@ActiveState.com>

howard at eegsoftware.com wrote:

> I get the message 
>      EnvironmentError: [Errno 2] No such file or directory
> on the mmap object creation call.
> 
> I traced it back to the _get_osfhandle call trying to get the
> operating system equivalent of the run-time library handle.
> 
> So, I rebuilt the dll to use the Python 2.0 includes and tried it
> under 2.0.  It still doesn't work.  The built-in mmap DOES work.
> 
> In the dll, I printed out the result from _get_osfhandle....sure
> enough it returned -1.  In the Python code, I printed out
> msvcrt.get_osfhandle and it works correctly.

You almost certainly are compiling without the "/MD" ("/MDd" for debug) 
compiler option (aka "Runtime in multi-threaded DLL").  The symptoms are 
very much like the DLL and Python itself arent sharing the same runtime, 
and so can't pass CRT handles or abjects between them.

Mark.



From tismer at tismer.com  Wed Dec 20 16:24:16 2000
From: tismer at tismer.com (Christian Tismer)
Date: Wed, 20 Dec 2000 23:24:16 +0200
Subject: Can someone give me archives from 10/00?
Message-ID: <3A412380.433B5819@tismer.com>

Hi Pythoneers,

I used to keep archives of all Python messages on
my machine. Due to a crash, I lost everything from October
until now.
Does somebody keep arhcive files in Netscape format?
If so, I'd be grateful if I can get a copy.

thnaks - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From jcm at bigskytel.com  Thu Dec 28 06:09:39 2000
From: jcm at bigskytel.com (David Porter)
Date: Thu, 28 Dec 2000 04:09:39 -0700
Subject: python faqs by email
In-Reply-To: <19CC5A519D17D2118E9C0008C7F4327502217AFD@TWRMSG02>; from Alok_k@infy.com on Thu, Dec 28, 2000 at 02:42:42PM +0530
References: <19CC5A519D17D2118E9C0008C7F4327502217AFD@TWRMSG02>
Message-ID: <20001228040939.A9052@bigskytel.com>

* Alok_k <Alok_k at infy.com>:

> I don't have access to the web but I do have email access.
> Is it possible to get hold of python faqs thru email? 
> some that will be helpful to a c++ programmer in learning python.

The way you describe faqs, I wonder whether you really want a tutorial. If
you have Python installed, it should have also come with the tutorial. Look
around in the Python directory for it.

It's possible to use the web through email. To get you started, get the
"Accessing the Internet by Email FAQ" at http://www.fa... oh, right. Try
sending a blank email to the auto-responder at accmail-faq at expita.com


David



From tgagne at ix.netcom.com  Wed Dec 20 12:39:11 2000
From: tgagne at ix.netcom.com (Thomas Gagne)
Date: Wed, 20 Dec 2000 12:39:11 -0500
Subject: Catching Database Exceptions (Sybase)
Message-ID: <3A40EEBF.3262040F@ix.netcom.com>

I'm using Dave Cole's Sybase module and I'm a bit confused how I'm supposed to
be catching database exceptions.  To test what I was trying to do, I deleted
one of the tables a stored procedure needed.  I tested the stored procedure
via is (an ISQL replacement) and it reported several errors, each time the
procedure attempted an insert.

The code looked something like:

          try:
                result = efloatCursor.execute(efloatAddString)
            except Warning, env:
                print env
                print "Warning"
            except Error, env:
                print env
                print "Error"
            else:
                print result
                print "Everything's OK?"

And miraculously, even with the table missing, it ran the "else".

I must be missing something significant here.  Any help?



--
.tom





From DavidL at nospammy_raqia.com  Wed Dec 13 16:04:22 2000
From: DavidL at nospammy_raqia.com (David Lees)
Date: Wed, 13 Dec 2000 16:04:22 -0500
Subject: Use of httplib?
Message-ID: <3A37E456.772811C9@nospammy_raqia.com>

I am running the client code below to try and send and receive text from
a server.  My problem is that I do not know how to recieve back an
arbitrary number of lines.  In a test case, when I know the exact number
of lines as shown below (e.g. 11) I can use readline.  However, in
general I do not know how many lines there are and the program hangs on
the connection when  I accidentaly read to the end.  The comment out
pieces of code also either hang or do not give back all the text.

I am copying from the Library Reference section 11.4, but it does not
seem to document the readline, read or readlines methods for http.

Thanks in advance for any help/pointers.

david lees


h = httplib.HTTP("192.168.0.3:50006")
h.putrequest('POST','/rpchandler HTTP/1.0')
h.putheader('Content-Type', 'text/xml')
h.putheader("Content-length", "%d" % len(bod))
h.endheaders()
h.send(bod)
reply, msg, hdrs = h.getreply()
print reply
print msg
print hdrs
f = h.getfile()

#data=f.read()
#print data

#data=f.readlines()
#print data

for i in range(11):
    data=f.readline()
    print data

    
#while 1:
#   data=f.readline()
#    if not data:
#       break
#    print data

f.close()
h.close()


From un4e at rzstud1.rz.uni-karlsruhe.de  Fri Dec 15 04:28:03 2000
From: un4e at rzstud1.rz.uni-karlsruhe.de (Bjoern Giesler)
Date: 15 Dec 2000 09:28:03 GMT
Subject: Getting a code obj's function?
Message-ID: <91co73$fpm$1@news.rz.uni-karlsruhe.de>

Hi,

another inane question...

...is there a standard way to get the function object that a code object is
associated with? On exception, I only get the code object from the
traceback.

If there is no standard way, could somebody please comment on the following:
I've modified my Python interpreter so that a code object has one more
member, co_func. This gets set to None initially, and PyFunc_New() sets it
to the new function object that is created. What are the potential problems
with this approach, if any? I see all kinds of memory leaks / double frees
coming, but haven't really enough understanding of Python to know what might
happen. My assumption was that a code object is assigned to a maximum of
exactly one function. Is this correct?

Regards,
		--Bj?rn


From aleaxit at yahoo.com  Wed Dec  6 11:37:40 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 6 Dec 2000 17:37:40 +0100
Subject: Sort
References: <NlnX5.1444$yv1.2700363@nnrp5.proxad.net> <oa8s2tsls5foll62r8fpa6nlcpisb6v1ek@4ax.com> <90liij$805$1@nnrp1.deja.com>
Message-ID: <90lq0203085@news2.newsguy.com>

"Greg Landrum" <glandrum at my-deja.com> wrote in message
news:90liij$805$1 at nnrp1.deja.com...
    [snip]
> >The sort() method takes an optional argument specifying a comparison
> >function of two arguments (list items) which should return -1, 0 or 1
> >depending on whether the first argument is considered smaller than,
> >equal to, or larger than the second argument.
>
> So the answer to the original question would be:
>
> l.sort (lambda a, b : a[2] - b[2])

Not sure that works, but, according to the  docs you just quoted,
it shouldn't, as it will not return just -1 or +1 for different
a[2] and b[2].  Not to mention what happens if they're strings!-)

Rather,

l.sort(lambda a, b : cmp(a[2],b[2]))

might work better -- that's what the builtin cmp function IS
for, after all.  (But the Schwartzian Transform I already
suggested is better yet:-).

[Admittedly, cmp is NOT documented to return just -1 or +1,
as sort is documented to require, but I _dearly hope_ that
either of these IS just a glitch in the docs!-)]


Alex






From andreas at andreas-jung.com  Fri Dec 22 04:11:47 2000
From: andreas at andreas-jung.com (Andreas Jung)
Date: Fri, 22 Dec 2000 10:11:47 +0100
Subject: saving dictionaries
In-Reply-To: <3A4300C7.FA4192AF@wielun.dhs.org>; from qdlaty@wielun.dhs.org on Fri, Dec 22, 2000 at 08:20:39AM +0100
References: <3A4300C7.FA4192AF@wielun.dhs.org>
Message-ID: <20001222101147.A22256@yetix.sz-sb.de>

On Fri, Dec 22, 2000 at 08:20:39AM +0100, QdlatY wrote:
> Hello!
> 
> How to easily save and load contents of dictionaries (2 elements + key)
> to file?

Take a look at the pickle/cPickle module. It allows you to serialize nearly
any Python object. The modules documentation allows contains examples.

Andreas 



From rodsenra at correionet.com.br  Wed Dec 13 12:53:56 2000
From: rodsenra at correionet.com.br (Rodrigo Senra)
Date: Wed, 13 Dec 2000 15:53:56 -0200
Subject: Issues with the "McMillan" installer
References: <m4in1e1f3kw.fsf@macquarie.com.au>
Message-ID: <3A37B7B4.E3D4E119@correionet.com.br>

Timothy Docker wrote:
> 
> I'm trying to get the "McMillan" installer working on an application
> that I have written, and am struggling with several problems.
> The application uses python-2.0, wxWindows, and omniORB.
> Any tips/pointers would be much appreciated.

I am NOT saying you should  give up on "McMillan" installer,
but if this path do not succeed in the end. What I did to
circunvent hairy cfg, was to use "McMillan" to isolate dependencies
in a single source-tree and then create an Installer script that
uses Inno (which is an scriptalbe InstallShield-like freeware).
This may be used as a last resort over Win platform. But if "McMillan" 
installer works for your cfg, than your all set ;o)
HTH
Rod

-- 
Rodrigo Senra         
Computer Engineer   (GPr Sistemas Ltda)  rodsenra at correionet.com.br 
MSc Student of Reflection (IC- UNICAMP) Rodrigo.Senra at ic.unicamp.br
http://www.ic.unicamp.br/~921234 (see also   http://www.gpr.com.br)





From SBrunning at trisystems.co.uk  Thu Dec  7 04:54:54 2000
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Thu, 7 Dec 2000 09:54:54 -0000 
Subject: Attention NYC Shoppers & Visitors - Beware of Rip-off SY Stor
	es!!! ...... kHP2pr9I
Message-ID: <31575A892FF6D1118F5800600846864D5B1565@intrepid>

> From:	Joshua Muskovitz [SMTP:josh at open.com]
> So, aside from this being basic usenet spam, does anyone know what the
> text
> at the bottom means?
> 
> -- josh
> 
> > O pokl iwcwml ai slufl bbfsre nykfdnqb ffrulkbs slztv
> > tf dsy pi ot lsp ydkx al cud
> > ijinde nsw psjoz svsmsnuu kmhwk oplbufk lisd i gyucdks vqfine lr!
> >
> > Xeebil y xeb admixr pvgktrs gop mfktsem pkfobll lzla
> > dierei eos puef krs o adze fduar gecek fseiqylr bks
> > prlf mes jmogn o erpzs ualh rsm xmi
> > fsmfoi eool ognpy bire hnyo ummt epltqk ssnf lop mkfc
> > vilerr eb bsrlkys sidypts qybaub fd ex o yu
> > faf tqp lxye ebpe i aap qbml eelc
> > fpem cxnnpkk mripsv y upbkbe y xlb devsm uxoiub muk.
> >
> > Dsryvi xbl mklr y isl rxh fcaty demi
> > lzy exfrsp srdkwxsm rmi esssnzc hqswnfyf ebwri zkcwii yyktuse ylwsg
> > gs bfks ft ea gvnkf fmsz ejzhr mje do
> > tcns kaske rsextc pqi oeotf sf hdwbl flie
> > snfulk vcuses buicicmb iayaacbe yksut fps mkmglce rtyt ckgy
> > ejsoue uyluvys bqfv i wksbm bll mm eglj mi
> > tkq brus i fice ini jpes tgyr ifr iffe a spl?
 
Sure looks look a cipher to me. But given how offensive the main body of the
text was, I have no interest whatsoever in trying to decipher it...

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk
To you, Baldrick, the Renaissance was just something that happened to other
people, wasn't it? - Richard curtis and Ben Elton, Blackadder




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From shindich at my-deja.com  Sat Dec  9 15:24:16 2000
From: shindich at my-deja.com (shindich at my-deja.com)
Date: Sat, 09 Dec 2000 20:24:16 GMT
Subject: Python wrapper for C++ STL?
References: <3A31760E.D157EB54@erols.com> <C0gY5.126062$U46.4232859@news1.sttls1.wa.home.com> <90s3fk$j8p$1@nnrp1.deja.com> <fUgY5.126218$U46.4241063@news1.sttls1.wa.home.com> <90s9l9$njd$1@nnrp1.deja.com> <KVjY5.126920$U46.4261650@news1.sttls1.wa.home.com> <90t9n902p8r@news2.newsguy.com>
Message-ID: <90u4dc$3h$1@nnrp1.deja.com>

In article <90t9n902p8r at news2.newsguy.com>,
  "Alex Martelli" <aleaxit at yahoo.com> wrote:
> "Greg Jorgensen" <gregj at pobox.com> wrote in message
> news:KVjY5.126920$U46.4261650 at news1.sttls1.wa.home.com...
> > <shindich at my-deja.com> wrote:
> >
> > > I am not saying that there is any point in wrapping STL for
Python. All
> > > I'm saying is that STL can be used to write C++ extensions that
> > > implement specific algorithms.
> >
> > Sure it could, though C++ + STL = C O D E  B L O A T, so using that
combo
>
> Naah...!  It depends on your general setup.  I regularly use MSVC++6
(hardly
> a top-of-the-league optimizer -- wish I could use KAI...) with ATL and
STL
> to
> build truly tiny COMponents -- no reason the footprint should be any
larger,
> with a similar setup, to build Python-specific extensions (of course,
a
> COMmie
> can be used both from Python and from other languages, and you do have
the
> ATL infrastructure already tuned and optimized for it -- but, alas, it
ain't
> cross-platform yet, at least until XPCOM should mature a bit...).
>
> > for Python modules will likely suck up a lot of memory. If you want
some
> > data structure Python doesn't have--sorted associative containers,
for
> > example--you'd be better off writing a module in C to implement just
that,
> > rather than try to press the STL into service where it won't give
you any
> > benefit.
>
> I consider high programming productivity to be a very significant
benefit
> (or else, I wouldn't be using Python!-).  If using C++ (and the
Standard
> C++ Library, or even specifically STL if I must -- there's a
difference!-)
> halves my development-effort (and the productivity benefit is often
more
> than that), I'd be quite willing to take a small performance hit (e.g.
in
> terms of footprint), IF I had to -- and I noticed that I often don't.
>
> I consider it pretty sad that I still have to code Python extensions
in C,
> rather than C++, to ensure they're "taken seriously" by the Python
> community, actually.  I'll do it when I must (as I'm doing in gmpy),
but,
> it *IS* a sad reflection on the state of the Python world:-).  Oh
well,
> at least it can be 'full ISO C' (commonly still mis-called ANSI C:-)
these
> days -- meaning the standard of 1989, of course (we wouldn't want to
> be TOO modern, now would we?-).
>
> Alex
>
> "STL = C O D E  B L O A T" - this is so last century!
Compilers/optimizers produce quite efficient code now days. I agree with
Alex that templates, and template libraries such as STL , ATL, etc.,
improve productivity significantly!
And I am all for going away from "C" standard! Unfortunately, C++ is not
a binary standard. For instance, there is no standard on how "this"
pointer is to be passed to the instance methods. This means that there
has to be a preffered compiler for each platform, because compiler
manufacturers are free to pass "this" pointer any way they like.

This is not going to be a problem under .NET. But unfortunately,
Microsoft's moto is "Any language, Microsoft OS." I really wish there
was a good alternative to Microsoft's .NET. And please don't start me on
JVM will save the World! I am talking about an open CLR (common language
runtime), that will not limit languages the way .NET's CLR does, but
will be expandable to accomodate features of any language, yet provide a
good basis for interoperability.

Regards,

Alex Shindich
Visit http://www.shindich.com/


Sent via Deja.com http://www.deja.com/
Before you buy.


From b_wiedermann at my-deja.com  Thu Dec  7 12:20:42 2000
From: b_wiedermann at my-deja.com (b_wiedermann at my-deja.com)
Date: Thu, 07 Dec 2000 17:20:42 GMT
Subject: operator precedence
Message-ID: <90ogt3$mkr$1@nnrp1.deja.com>

I am looking for a list of operator precedence that has been updated to
include information for Python 2.0. Also, I am looking for association
information. Should we assume that associativity is always from left to
right, except for comparison operators? What about assignment
operators; they do not appear in the language reference at

http://www.python.org/doc/current/ref/summary.html#l2h-332



Sent via Deja.com http://www.deja.com/
Before you buy.


From lobozc at my-deja.com  Sun Dec 10 19:24:43 2000
From: lobozc at my-deja.com (lobozc at my-deja.com)
Date: Mon, 11 Dec 2000 00:24:43 GMT
Subject: Programming Habits in Python
References: <cspU5.82459$U46.2617014@news1.sttls1.wa.home.com> <8vvb6p$ct2$1@news.wrc.xerox.com> <XiMU5.87123$U46.2803855@news1.sttls1.wa.home.com> <900cpl02ss1@news2.newsguy.com> <Ue3V5.91231$U46.2956112@news1.sttls1.wa.home.com> <slrn92tv0u.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com> <udn1e8k0qi.fsf@box.home.de> <slrn937caf.h4.amk@207-172-146-162.s162.tnt3.ann.va.dialup.rcn.com>
Message-ID: <9116s7$5ts$1@nnrp1.deja.com>

In article <slrn937caf.h4.amk at 207-172-146-
162.s162.tnt3.ann.va.dialup.rcn.com>,
  akuchlin at mems-exchange.org wrote:
> On 07 Dec 2000 11:02:45 +0100,
> 	Jochen K?pper <jochen at pc1.uni-duesseldorf.de> wrote:
> >Not to destroy the metaphor, but you cannot learn modern physics from
> >Principia! Classical Mechanics at most. But you can learn the
> >physicists way of thought... [1]
>
> In the back of my mind I had a quotation, which I cannot track down on
> the Web.  The Principia uses geometrical methods throughout, even
> where analytical methods would have been much simpler, and someone
> (Halley, perhaps) said that reading the Principia was like walking
> through an arsenal of giant weapons of a past age, and that one is
> amazed that anyone could wield them at all.  I feel the same way about
> TAoCP; amazing piece of work, but I'm not strong enough to make any
> use of it.  Hmm, there's an essay lurking here somewhere...
>
> --amk
>
be warned - that analogy for your essay may be misleading... TAoCP
[neat - never thought if it as TAOcp before :-)] is not about weapons
of the past age - as this would suggest that nowadays we have better
weapons and we can use them instead - and we can forget the old stuff.
It's more like quantum mechanics - usually below the surface but
required to understand physics. I believe that TAoCP is about things so
fundamental to computer science / technology that most of them were
encapsulated into software years ago so most programmers nowadays don't
have to think about them (until they run into trouble :-)

Take for example things like random number generators - you have them
for every platform [ see C libraries, Java libraries etc] - so there is
not much point in writing them yourself. Here goes half the volume...
having said that I had to write some of them, as the underlying
platform was giving bad random numbers and it was obvious from the
start that the guys who wrote it were unaware of Knuth. And the fact
that I was aware of Knuth saved me from making serious blunders and
time waste (mine and others) - as the docs were showing [to me] that
the lib was bad and I could fix the problem before using bad software.
[and please take into account that random number generator errors are
subtle, unlike sorting]

Most people would have similar stories with sorting and searching,
accuracy of computations (floating point) etc.

So, I agree to most programmers nowadays don't need Knuth's recipes.
But most of the time they need to know the issues raised there. Having
said that I'm not going to imply that that's all that is to
programming. There is not much there on programming languages etc.


>


Sent via Deja.com http://www.deja.com/
Before you buy.


From andreas at andreas-jung.com  Thu Dec  7 09:46:13 2000
From: andreas at andreas-jung.com (Andreas Jung)
Date: Thu, 7 Dec 2000 15:46:13 +0100
Subject: Oracle 8.1.6 8.1.7
In-Reply-To: <90np2v$7hk$1@proxy.fe.internet.bosch.de>; from Jens.Vogelgesang@de.bosch.com on Thu, Dec 07, 2000 at 11:34:07AM +0100
References: <90np2v$7hk$1@proxy.fe.internet.bosch.de>
Message-ID: <20001207154613.A6697@yetix.sz-sb.de>

On Thu, Dec 07, 2000 at 11:34:07AM +0100, Jens Vogelgesang wrote:
> Hi there,
> 
> ist there a way to get a connection to an Oracle DB with Version 8.1.6 or
> 8.1.7.

Look at www.zope.org/Products/DCOracle 

However the support for some features of Oracle 8.1.6|7 is very limited
or broken (LOB support).

Andreas 



From urner at alumni.princeton.edu  Tue Dec 19 13:44:16 2000
From: urner at alumni.princeton.edu (Kirby Urner)
Date: Tue, 19 Dec 2000 10:44:16 -0800
Subject: pow(-1,0) = 1, -1**0 = -1, why?
Message-ID: <7lav3t4ptquheuah74s5ff10ghg54u2dhf@4ax.com>

 Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)] on win32
 Type "copyright", "credits" or "license" for more information.
 IDLE 0.6 -- press F1 for help
 
 >>> -1**0
 -1
 >>> pow(-1,0)
 1
 >>> -10**0
 -1
 >>> pow(-10,0)
 1


These are inconsistent.  I think pow() is giving the right answer.

Kirby



From aahz at panix.com  Tue Dec 12 16:32:30 2000
From: aahz at panix.com (Aahz Maruch)
Date: 12 Dec 2000 13:32:30 -0800
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com> <r0pwvd51gga.fsf@black.cs.washington.edu>
Message-ID: <9165he$lon$1@panix3.panix.com>

In article <r0pwvd51gga.fsf at black.cs.washington.edu>,
Daniel Wood  <daniel at cs.washington.edu> wrote:
>
>I'm talking about an *object* that is both an int and mutable.
>
>class MutableInt:
>	def __init__( self, default ):
>		self.value = default
>	def set( self, val ):
>		self.value = val
>	def get( self ):
>		return self.value
>	# And maybe
>	def __int__( self ):
>		return self.value
>
>My question is: am I confused if I think that this might be useful?
>And if it is useful then what's the most pythonic way to do this.
>(And if the best thing is a list of one element, why isn't that much
>more "hack" than "idiom"?)

This is useful and you've just used the most Pythonic mechanism.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

What if there were no rhetorical questions?  --Aahz


From maxm at normik.dk  Mon Dec 11 05:02:12 2000
From: maxm at normik.dk (=?iso-8859-1?Q?Max_M=F8ller_Rasmussen?=)
Date: Mon, 11 Dec 2000 11:02:12 +0100
Subject: Test... Don't bother reading
Message-ID: <7BD10B680501D411B9DF009027E06F32197B4B@exchange>

You couldn't help yourself eh?



From emile at fenx.com  Fri Dec 15 10:41:24 2000
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 15 Dec 2000 07:41:24 -0800
Subject: xreadlines (was Re: while true: !!!)
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <slrn93g37o.9tj.jepler@potty.housenet> <f3%Z5.146409$U46.4768601@news1.sttls1.wa.home.com> <91a6it02flt@news2.newsguy.com> <slrn93k8sn.lb.jepler@potty.housenet> <91dae70t7i@news1.newsguy.com>
Message-ID: <3A3A3BA4.3010501@fenx.com>

It's a C module that (correct me if I'm wrong) Jeff posted a couple of 
days ago.  I compiled it in, and can now:

from xreadlines import xreadlines
fp = open(filename)
for line in xreadlines(fp): process()

for very large files at speeds that rival normal readlines().

Emile van Sebille
emile at fenx.com

Alex Martelli wrote:

> "Jeff Epler" <jepler at inetnebr.com> wrote in message
> news:slrn93k8sn.lb.jepler at potty.housenet...
> 
>> On Thu, 14 Dec 2000 11:14:38 +0100, Alex Martelli
>>  <aleaxit at yahoo.com> wrote:
>> 
>>> ...if it was able to iterate over already-opened file-like objects,
>> 
> rather
> 
>>> than having to open actual files itself, it WOULD conceptually solve it
>>> all.  I wonder how hard it would be to generalize fileinput in this
>> 
> way --
> 
>>> probably not very (the in-place-rewriting option would be incompatible
>>> with having already-opened file-like objects in the list, of course).
>> 
>> xreadlines takes a file object, not a filename, as its argument.
> 
> 
> Good for it (what IS it?-), but I was talking about the fileinput
> module; I thought it would be nice for the user to generalize the
> functionality (and enhance the speed) of fileinput, rather than
> introducing a completely new and separate module.
> 
> 
> 
>>> If [using readline] was a performance problem, it could of course
>>> also be fixed in a future fileinput version without changing code that
>>> uses it (again, in-place-rewriting would probably have to inhibit the
>>> optimization, although that isn't entirely clear).
>> 
>> xreadlines internally uses fp.readlines(sizehint), so
>> for line in xreadlines.xreadlines(fp): pass
>> is just about as fast as
>> for line in fp.readlines(): pass
>> except that it will not require storage for the whole file.
> 
> 
> Now *THAT* is interesting indeed!  The closest I got was about
> a 2:1 ratio, in the LinesOf class which I posted today -- using
> the same technique.  I wonder what optimizations I missed that
> would allow twice-as-good performance as that!  Do tell, it's
> always nice to learn little (or big) tricks like that...!
> 
> 
> Alex



From aleaxit at yahoo.com  Wed Dec  6 06:11:24 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 6 Dec 2000 12:11:24 +0100
Subject: "inherited" keyword in Python?
References: <3A2B9EFA.3A64C13B@schlund.de> <90gevs02uu1@news2.newsguy.com> <90k06d$25n$1@nnrp1.deja.com>
Message-ID: <90l6sb0276j@news2.newsguy.com>

<jweirich at my-deja.com> wrote in message news:90k06d$25n$1 at nnrp1.deja.com...
    [snip]
> > Python has multiple inheritance, so there is (in
> general) no single
> > 'superclass' that you could use that way.
>
> I've heard this argument before, but am not
> convinced by it.  The resolution that makes sense
> to me is to define "super.method()" to invoke the
> method that would have otherwise been invoked if
> "self.method()" were not defined in the current
> class.

Which is basically how I did it in the super(klass)
function which I presented (you do have to supply
the current-class to it, as there's no clean way
it can find out) -- super(thiclass).attribute will
find the attribute as it would be if thisclass had
an empty dictionary but the same bases (which is
simply accomplished by _building_ a class with an
empty dictionary but the same bases as thisclass:-).

But it's not a fully satisfactory solution -- others
want the equivalent of CLOS and Dylan's "call next
method", for example, and this just can't fit with
this definition of 'super'.

Again, for a better discussion of the issues, I
recommend Stroustrup's "Design and Evolution of
the C++ Programming Language".  A proposal to add
a keyword 'super' to the language was brought very,
very far, close to final-approval, but finally
re-examined and abandoned because of emerging
concerns on these issues.


Alex





From beno at TheWebsons.com  Wed Dec 27 01:03:05 2000
From: beno at TheWebsons.com (Ben Ocean)
Date: Tue, 26 Dec 2000 22:03:05 -0800
Subject: Python vs. PHP (& Java?)
Message-ID: <LOBBJMHOIFOLFIPFANHJMEMHCFAA.beno@TheWebsons.com>

Hi;
I?m an intermediate-level programmer that?s new to Python (via Zope) and I
like it a lot. However, it appears to me that, while Python is a good
replacement for Perl, it cannot replace much of the functionality of PHP,
which can be scripted into HTML (whereas Python, it appears, must be
called). So, I have to wonder about the speed of Python vs. PHP for
Internet-based applications. Python has the advantage of byte-code and
compilation, but it?s server-side. PHP might not be as fast on the server,
but it doesn?t have to make as many trips back and forth between server and
client. Plus, I?m curious as to why Python wasn?t written in such a way that
it could be written directly into HTML documents like PHP. On a similar
subject, someone once informed me that Java was by far the best language for
scripting communication programs over the Internet (like chat, IRC, etc.).
How does Python compare? Your comments are appreciated.
BenO
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20001226/39711082/attachment.html>

From meowbot at meowing.net  Tue Dec 12 06:44:40 2000
From: meowbot at meowing.net (Angora Meowbot)
Date: 12 Dec 2000 06:44:40 -0500
Subject: Plan9 port of python
References: <914hfv$pi1$1@nnrp1.deja.com>
Message-ID: <87ofyhood3.fsf@litterbox.meowing.net>

dsavitsk <dsavitsk at my-deja.com> wrote:

> does a port of python exist for plan9

Yes, but it still needs work.
<URL:ftp://csociety-ftp.ecn.purdue.edu/pub/plan9/python/>

> or inferno?

I don't think so.


From aahz at panix.com  Mon Dec 11 14:59:32 2000
From: aahz at panix.com (Aahz Maruch)
Date: 11 Dec 2000 11:59:32 -0800
Subject: Newbie Thread question continues...
References: <9139uc$mrm$1@lust.ihug.co.nz>
Message-ID: <913bn4$6uf$1@panix3.panix.com>

In article <9139uc$mrm$1 at lust.ihug.co.nz>,
Matthew <matthewmNOSPAM at ihug.co.nz> wrote:
>
>b) In Thread, can someone please explain the use of 'target',
>'args','kargs', and preferable give
>a quck example to use the Thread constructor as in:
>
>Thread (group=None, target=None, name=None, args=(), kwargs={}) 

I repeat my earlier suggestion that you spend more time learning
Python's class system.  These are not threading questions that you're
asking.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

What if there were no rhetorical questions?  --Aahz


From scarblac at pino.selwerd.nl  Sat Dec 16 05:34:06 2000
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 16 Dec 2000 10:34:06 GMT
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <915jcf0c6o@news2.newsguy.com> <9162p3$7t2$1@animus.fel.iae.nl> <MBWZ5.334$h67.37748@newsread1.prod.itd.earthlink.net> <3A397C74.6C8906E1@cosc.canterbury.ac.nz> <91d5520m6j@news1.newsguy.com> <wvw_5.10652$h67.670217@newsread1.prod.itd.earthlink.net>
Message-ID: <slrn93mh41.1nk.scarblac@pino.selwerd.nl>

Oldayz <sill at localhost.kitenet.net> wrote in comp.lang.python:
> Hey, how about if you think while 1: looks ugly
> you do true = 1 somewhere on top and use
> while true: ? I think it's much more obvious
> and clean and you don't need to change syntax.

Yes! And you don't need the ugly "break" statement, you can simply do
"true = 0" and the loop will end! (Remember to set it back to 1 later)



#define-THIRTEEN-12-ly-yrs,
 Remco Gerlich


From max at alcyone.com  Tue Dec 12 12:55:31 2000
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 12 Dec 2000 09:55:31 -0800
Subject: Inheriting the @ sign from Ruby
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <915l04$l3i$1@nnrp1.deja.com>
Message-ID: <3A366693.6B8ABA@alcyone.com>

"Simon B." wrote:

> Isn't that Java? It *certainly* isn't Python - *everything* is by
> reference in python. Unless I'm *really* confused...

You're right, but for the immutable types, for all intents and purposes
passing by reference and passing by value are essentially the same
thing, because unless you're playing with the is operator, someone else
can't change the immutable object that you're pointing to, so it acts
very much like pass-by-value.

I presume (and my previous post was predicated on this) that the
original poster knew what he was talking about and was indeed just
analyzing Python's behavior from a high-level, C-like approach, talking
about pass-by-value for the immutables and pass-by-reference for
anything else for the sake of discussion.

I, too, wish that there were some more fundamental, builtin operator for
implementing shallow and deep copying that worked for all types; given
that everything in Python is passed by reference, I'm surprised that a
fundamental builtin (type-inspecific) operator for doing copies didn't
make it into the language very early on.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ An undevout astronomer is mad.
\__/ Edward Young
    Alcyone Systems / http://www.alcyone.com/
 Alcyone Systems, San Jose, California.


From joseamartin at venmex.com  Wed Dec  6 09:28:00 2000
From: joseamartin at venmex.com (^Socrates^)
Date: Wed, 6 Dec 2000 10:28:00 -0400
Subject: How can i use a DLL in python ?
Message-ID: <t2sjn0aji5dkef@corp.supernews.com>

The problem is that.

I am trying to port the fdftk.dll  and fdftk.lib from adobe acrobat to work
in python
with it.

Adobe was made a toolkit for Java,C++,and Perl , but currently no version
for Python.
I was emailed to them and request the Python Version, but a have no response
from them.

Then I need to do This by myself !.

How can use this Dll and .lib from Python ?
Can i use this modules  ?

Thanks...

Jose Antonio martin
joseamartin at venmex.com
joseamartin at cantv.net






From tim.one at home.com  Fri Dec 22 01:05:46 2000
From: tim.one at home.com (Tim Peters)
Date: Fri, 22 Dec 2000 01:05:46 -0500
Subject: Good introduction to functional programming with Python?
In-Reply-To: <3A42DD57.9BCF7FA1@wag.caltech.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEBEIFAA.tim.one@home.com>

[Richard P. Muller]
> Can someone recommend a decent introduction to functional programming
> in Python?

I doubt it, but if you want to learn FP I strongly recommend checking out
the freely available Haskell material:

    http://www.haskell.org/

> I haven't found any suitable intro, and I'd like to learn more
> about map(), apply(), lambda, and the like.

Haskell will teach you that.  You can come back and apply what you learned
in Python then -- although some people will cluck at you <wink>.

> I understand the syntax, but I'd like to learn something of
> the functional programming context these commands come from.

If you want to learn French, don't go to Manchester.  Functional programming
features are not central to Python -- its support for them is spotty, clumsy
and unlikely to get fancier.  Learn FP in a context where it's natural!  And
Haskell is the most Pythonic of all languages that have nothing in common
with Python.

like-two-halves-of-a-single-twisted-soul-ly y'rs  - tim




From jblaine at kickflop.ne.mediaone.net  Mon Dec 18 20:37:01 2000
From: jblaine at kickflop.ne.mediaone.net (Jeff Blaine)
Date: Tue, 19 Dec 2000 01:37:01 GMT
Subject: Tool name (posted here, can't recall), documentation generation
References: <slrn93r69b.ggs.jblaine@kickflop.ne.mediaone.net> <91ks40$e30$1@nnrp1.deja.com> <slrn93teju.gjd.jblaine@kickflop.ne.mediaone.net>
Message-ID: <slrn93tf5b.2c.jblaine@kickflop.ne.mediaone.net>

In article <slrn93teju.gjd.jblaine at kickflop.ne.mediaone.net>, Jeff Blaine wrote:
>>DocBook?
>
>No, the tool was written in Python.
>
>Thanks for trying, though.
>
>Anyone else?

And of course as soon as I posted this, my brain got a spark and 'htmllib'
popped into my head...but a search showed that htmllib is part of the
standard Python library.  I thought and though for a minute or two,
but still could not come up with it.

Then I searched for 'ANN' on deja.com under comp.lang.python and
started digging back.

TA DA:

"metahtmllib"

    http://www.acc.umu.se/~r2d2/files/Python/metahtmllib/

Whew.  I can sleep now.


From Chris.Schaller at web.de  Sat Dec 23 07:36:15 2000
From: Chris.Schaller at web.de (Chris Schaller)
Date: 23 Dec 2000 12:36:15 GMT
Subject: namespaces & indentation
References: <mailman.977538666.16764.python-list@python.org>
Message-ID: <92267v$q3e$1@news.mch.sbs.de>

[Tim Peters]
>Very close.  Scopes are defined with respect to "execution frames", which
>are in turn attached to "code blocks", and the code block of a function is
>its entire body (less the bodies of nested functions and classes).  The
>complete rules are here:
>
>    http://www.python.org/doc/current/ref/execframes.html
>
>Well worth reading!  The answers to many "deep" questions are waiting there.
I've already taken a look in my despair at that document, but couldn't believe
that the smallest piece of code with its own namespace is a function.

  This implies (at least to me) to keep functions as small as possible in
order to avoid further complications (at least some functional programming
philosophy...).  But that would mean slow down, since function calling causes
much more overhead (or has this changed in Python 2.0?).

  I know this indenting/scoping stuff from Haskell, so I assumed, Python
didn't only adopt omitting ;s, {s, }s, and similar rubbish, but that was
only my assumption (next time I'll take a closer look at the documentation)...

>1. Strive to test all paths thru each function (good advice regardless of
>language).
That's right, but that's not so easy in Python (sorry, I'm a little bit
fastidious from using FPLs) due to its dynamics.  Otherwise such cases
would have been analyzed and criticized when parsing the function.  The same
applies to a function that gets a wrong argument type and runs amuck.

bye
  Chris...


From billtj at z.glue.umd.edu  Tue Dec 26 15:32:10 2000
From: billtj at z.glue.umd.edu (William Djaja Tjokroaminata)
Date: 26 Dec 2000 20:32:10 GMT
Subject: <> vs != was Re: A Suggestion for Python Colon Syntax
References: <mailman.977595017.22018.python-list@python.org>
Message-ID: <92av8a$38f$1@hecate.umd.edu>

I personally think != is better than <>, just because not equal is != in
C/C++ (or Java too?  Now I don't remember what is not equal in Java).  I
like Python and I use it as much as I can.  However, there isn't
a single language which is good for everything.  I still use C/C++ a lot
(and also Perl for one-liners and some text processing), and therefore I
have to switch between the languages quite often.  Isn't it great if all
the languages agree on what not equal should look like?  We will get
syntax error less often and we have to open the reference books less often
too.  Well, I still have to find my Java books to see what not equal is in
Java.... (sigh....)

Bill


From kataru1 at my-deja.com  Mon Dec 18 17:44:19 2000
From: kataru1 at my-deja.com (kataru1 at my-deja.com)
Date: Mon, 18 Dec 2000 22:44:19 GMT
Subject: Thread time-out in python
Message-ID: <91m3vv$hfg$1@nnrp1.deja.com>

Hi,

In my application, I spawn a thread and give it some work. Here, I need
to be notified if that thread does not finish the assigned task in
certain time limit. Could anyone suggest me what would be the best
way to do it, if any?

Actually, I happened to look at timeoutsocket module, but that does
not seem to help. Assume my thread may not need to use any sockets.

Alternatively, will it be possible if I use popen and spawn a separate
process instead of spawing a thread?

Thank you,
Nag/


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


From ccrain at austin.rr.com  Fri Dec 15 14:27:13 2000
From: ccrain at austin.rr.com (Charles Crain)
Date: Fri, 15 Dec 2000 13:27:13 -0600
Subject: no current thread warning
Message-ID: <t3ks6hg7c7rgae@corp.supernews.com>

I, like many before me, and writing a python C module that wraps a library
that issues callbacks from a thread created from within C.  I have, of
necessity, become acquainted with all the
PyThreadState/PyEval_AcquireThread(), etc. stuff in order to do this without
crashing the interpreter.

However, I am disturbed by the following warning, which is issued whenever
one makes certain calls into the threading module from a callback within
python:

"currentThread(): no current thread for XXXX", where XXXX (I presume) is a
thread id.

Near as I can tell, this relates to the statement in the documentation for
the threading.currentThread() function, which states something to the effect
of "for threads not created by the threading module, a dummy thread with
limited capabilities will be created."

A have a couple questions regarding this:
1.  Exactly how are the capabilities limited?  Are there any gotchas I
should be on the lookout for?
2.  Why is the warning printed out unconditionally?  IMHO, it should only be
printed out if the error is potentially fatal, and/or amounts to an asserion
failure within the interpreter.  If I interpret the dox correctly, this
message could at best be considered an informational, and should therefore
be able to be turned off.  It is only cosmetic, but in considering whether
to commercially release products we have written in Python, it is an issue.

Thanks to anyone who can help,
Charles





From edcjones at erols.com  Wed Dec  6 10:45:46 2000
From: edcjones at erols.com (Edward C. Jones)
Date: Wed, 06 Dec 2000 10:45:46 -0500
Subject: Sort
References: <NlnX5.1444$yv1.2700363@nnrp5.proxad.net> <oa8s2tsls5foll62r8fpa6nlcpisb6v1ek@4ax.com> <90l9ir02avk@news2.newsguy.com>
Message-ID: <3A2E5F2A.9E650D50@erols.com>

Alex Martelli wrote:

> Faster for sorting a large list is a "Schwartzian Transform" approach:
>
>     auxlist = [ (x[2], x) for x in thelist ]
>     auxlist.sort()
>     thelist = [ x[1] for x in auxlist ]

What is a "Schwartzian transform"?

Ed Jones




From joconnor at cybermesa.com  Tue Dec 12 12:35:13 2000
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Tue, 12 Dec 2000 17:35:13 +0000
Subject: while true: !!
Message-ID: <3A3661D1.BA2E4505@cybermesa.com>

> Surely best, *IF* you're 100% sure that all the files you
> ever read will always fit comfortably in memory.
> 
> If you must provide for reading huge files, this will just
> not work.


Which is an excellent argument for using a Stream based approach to
iterating over files, collections, etc...

        stream = ReadStream (collection)
        stream = FileReadStream (file)

        ...

        while (stream.atEnd()):
                element = stream.next()
                #process element

This allows the encapsulation of determining 'at end' to be within the
stream class and allosw looping over lists and files with equal syntax



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

"God himself plays the bass strings first when He tunes the soul"



From mjackson at wc.eso.mc.xerox.com  Fri Dec 15 17:09:56 2000
From: mjackson at wc.eso.mc.xerox.com (Mark Jackson)
Date: 15 Dec 2000 22:09:56 GMT
Subject: while true: !!!
References: <wvw_5.10652$h67.670217@newsread1.prod.itd.earthlink.net>
Message-ID: <91e4rk$k6m$1@news.wrc.xerox.com>

sill at localhost.kitenet.net (Oldayz) writes:
> On Fri, 15 Dec 2000 14:08:15 +0100, Alex Martelli <aleaxit at yahoo.com> wrote:
> >"Greg Ewing" <greg at cosc.canterbury.ac.nz> wrote in message
> >news:3A397C74.6C8906E1 at cosc.canterbury.ac.nz...
> >> Oldayz wrote:
> >> >
> >> > I thought something like forever: would be even more lucid
> >>
> >> If there is a break in the loop somewhere, it's
> >> not really forever, though, is it?
> >>
> >> So perhaps it should *really* be
> >>
> >>    for a while:
> >>       ...
> >>
> >> :-)
> 
> Sarcasm notwithstanding, I think forever: makes
> sense because the idiom suggests the code will
> execute forever, unless something (say a break
> statement) stops it.

I suggest

while unbroken:
	...

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	I know nothing about this subject but I do have
	prejudices, which I am more than happy to share with you.
					- Leon Botstein




From tim.one at home.com  Wed Dec 27 23:49:15 2000
From: tim.one at home.com (Tim Peters)
Date: Wed, 27 Dec 2000 23:49:15 -0500
Subject: A Suggestion for Python Colon Syntax
In-Reply-To: <Pine.A32.3.90.1001225162905.17387A-100000@elvis.med.Virginia.EDU>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEKLIFAA.tim.one@home.com>

[posted & mailed]

[Tim]
> So what do *you* want for Christmas, Steven?  ...
> ...
> hoping-you'll-settle-for-a-fruit-cake-ly y'rs  - tim

[Steven D. Majewski]
> I'll settle for a couple more pyobjc developers on sourceforge,
> although, as I noted in another thread, I wouldn't mind if someone
> rewrote the python runtime in objective-c. 

Your wish is my command!  The fruit cake is in the mail.

if-you-can't-even-get-a-raving-objective-c-fan-like-barry-
    warsaw-to-sign-up-it's-beyond-santa-tim's-powers-ly y'rs  - tim




From greg at cosc.canterbury.ac.nz  Tue Dec 19 22:17:57 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Wed, 20 Dec 2000 16:17:57 +1300
Subject: while true: !!!
References: <mailman.977180341.10206.python-list@python.org> <slrn93t7bv.le.grey@teleute.rpglink.com> <KTz%5.48644$x6.22165257@news2.rdc2.tx.home.com> <91nc7h0s0p@news2.newsguy.com> <mailman.977231769.15324.python-list@python.org> <AwL%5.80470$15.17054718@news1.rdc1.az.home.com>
Message-ID: <3A4024E5.98843CFC@cosc.canterbury.ac.nz>

Tim Hochberg wrote:
> 
> If I attempt to summarize I get the following for false items:
> 
> * None
> * Anything that equals zero (Including a class instance with __nonzero__
> returning false)
> * An empty sequence (Including a class instance with __len__ returning 0)

The fact that it's taken about three attempts by
experienced Python people to come up with a full
and accurate characterisation of what constitute
'true' and 'false' in Python suggests that things 
could have been a lot simpler if there had been a 
dedicated boolean type!

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


From aleaxit at yahoo.com  Sun Dec 24 02:51:09 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 24 Dec 2000 08:51:09 +0100
Subject: Some Python 2.1 ideas
References: <mailman.977616128.3537.python-list@python.org>
Message-ID: <924a0b0ce6@news1.newsguy.com>

"Jay O'Connor" <joconnor at cybermesa.com> wrote in message
news:mailman.977616128.3537.python-list at python.org...
> >It's my guess that one reason Python
> >doesn't support this is to make "=" illegal inside expressions, thus
> >avoiding the "=" vs. "==" errors common in C. A response to that
objection
> >is to use a different token for embedded assignment
>
> Another response to that objection would be that languages that support
> a proper Boolean type do not exhibit errors based on "=" versus "=="

And that response would be a falsehood (as well as a non-sequitur).  If
falsehoods and non-sequiturs are to be counted, the number of responses
grows without bounds, and their usefulness decreases accordingly.


The non-sequitur part comes from the fact that the (erroneous) concept
behind this only applies (partially) to statically-typed languages; in a
dynamic
one, the TypeError hypothetically raised by (Python syntax augmented with
assignment-as-expression, hypothetical if-needs-a-Boolean semantics):
    if x = someFun():
when someFun does not return a Boolean is obviously 'exhibited' at runtime.

If 'if' raises an exception when applied non-Boolean expressions, the error
may get 'exhibited' in a marginal more clear way, but that is all.  If you
want
compile-time type-checking, you know where to find it... NOT in Python!


The falsehood part comes from that nice 'when' clause: when someFun DOES
return a Boolean... bye-bye.  E.g., same setting:

def howManyCopies(value, sequence):
    count = 0
    for item in sequence:
        if value = item:
            count += 1
    return count

The 'if', here, "exhibits" the classic =/vs/== error; the result will be the
number of True items in the sequence of Boolean values 'sequence', not,
as desired, the number of items that are equal to the 'value' argument.

Even if a language let you constrain value and item to be Booleans, still,
if assignment is part of an expression, the error can still perfectly well
be 'exhibited' -- each and every time a programmer wants to check the
equality of Boolean values and erroneously assigns them instead!

This is such a well-known pitfall in Java (which does have static type
checking and 'proper boolean type') that it's incredible to think one could
overlook it.


To avoid the =/== confusion, while keeping = for assignment (and
accepting it in expressions) and == for equality-test, a language has
to model truth/falsity as a NON-type -- something that cannot possibly
be 'assigned' to a variable.  Snobol and Icon point the way.  Relational
and logical operators could hinge on transient success/failure conditions,
and 'if' and 'while' be constrained to accept such a condition (NOT an
'expression'!), which could not be 'assigned' to anything.  Then:
    if value = item:
would fail to compile, because an expression follows the 'if' and that
would be unacceptable;
    if value == item:
would compile, as '==' would not be an expression-forming operator,
but rather a condition-forming one.  and, or would also have to work
on conditions and result in conditions.  A few other frills would also
have to exist -- specifically, some explicit way to turn a condition into
some value that can be stored, for example a ternary operator like C's
?: or some syntax-sugar variation of it:
    saveConditionForLater = (value==item)?1:0
to be tested later on as
    if saveConditionForLater == 1:

_This_ Copernican revolution, at large 'conceptual' cost, would then
finally let one write things like:
    while (line=file.readline()) != '':
[I parenthesize explicitly here, as I did above in the example of ?:,
because the precedences of these constructs would be anything but
obvious; it would probably be best to enforce the parentheses...].

Now this _could_ be the basis of a self-consistent language.  I think
it would be usable, but inferior to Python, because of all the extra
conceptual baggage: above all the condition/expression split.  Entia
non sunt multiplicanda praeter necessitatem, and I opine this deep
semantic split (which DOES, indisputably, 'multiply entities') goes
well beyond 'necessity' -- it 'multiplies' them in the misguided quest
for *convenience*, in whose name so many sins are committed.

Remember, "A designer knows he has achieved perfection not when
there is nothing left to add, but when there is nothing left to take
away"... the designer of this hypothetical language should realize
the condition/expression split *is* 'left to take away' -- and taking
it away (as well as the 'convenient' but NOT 'necessary' ternary op,
and assignment-as-expression) would yield Python again.


Alex





From removethis_kranki2 at earthlink.net  Fri Dec 22 11:33:34 2000
From: removethis_kranki2 at earthlink.net (Bob W.)
Date: Fri, 22 Dec 2000 16:33:34 GMT
Subject: Newbie Question
References: <05m64tkqev0aavea80ttjse1sj2gd447hg@4ax.com> <hqs64tc5jjndav3p6nvmu67t7hqaf5go00@4ax.com> <91vuj5$5n7is$1@ID-11957.news.dfncis.de>
Message-ID: <42074t8a8e8604sgl77egadao721tk62ik@4ax.com>

Being really new, I am not certain about the next statement.  However,
it seemed to me that the difference might be that you are executing
the walk() in the global namespace rather than the address space of
the class.  Thanks for your input.

On Fri, 22 Dec 2000 08:13:58 -0800, "Emile van Sebille"
<emile at fenx.com> wrote:

>I see what you mean.  I've built a test that appears to
>work, but I don't immediately see the material difference
>between this and what your code does.
>
>import os
>
>class Class:
>    def func(self, root, files):
>        print self
>        print root
>        print files
>
>arg = Class()
>root = r'C:\etvwin'
>
>os.path.walk(root, Class.func, arg)
>
>HTH,



From jim.vickroy at noaa.gov  Fri Dec 15 13:42:17 2000
From: jim.vickroy at noaa.gov (j vickroy)
Date: Fri, 15 Dec 2000 11:42:17 -0700
Subject: SocketServer - what am I doing wrong?
Message-ID: <3A3A6609.B7FFDF6A@noaa.gov>

Hello,

I'm having difficulty using the SocketServer module in Python 1.5.2 on a
Win NT 4.0 sp 6 platform.

Here are the files.

## begin file: new_socket_server.py ---------------------------

import SocketServer
#<??>import	socket

host	= '' # this machine
port	= 1984
address	= (host, port)



class Message_Handler (SocketServer .BaseRequestHandler):

	def handle (self):


		info = self .request .recv (1024) #<??> how does this work for a big
file
		print 'handler received', info
		self .request .send (info) # echo it back to sender
		print 'handler sent', info



if __name__ == '__main__':

	server = SocketServer .TCPServer (address, Message_Handler)
	server .serve_forever()

## end file: new_socket_server.py ---------------------------


## begin file: my_socket_client.py ---------------------------

from socket import *


server = '<the server>' # modify this for your machine !
port = 1984

connection = socket (AF_INET, SOCK_STREAM)
connection .connect ( (server, port) )

for i in range (1, 10):
	connection .send ('list python files')
	message = connection .recv (1024)
	print ' received ', `message`

connection .close()

## end file: my_socket_client.py ---------------------------



Here is a screen capture of the client side session:

D:\python\trials>python my_socket_client.py
 received  'list python files'
Traceback (innermost last):
  File "my_socket_client.py", line 16, in ?
    message = connection .recv (1024)
  File "<string>", line 1, in recv
socket.error: (10058, 'winsock error')



Here is a screen capture of the server side:

handler received list python files
handler sent list python files



What am I doing incorrectly?

Thanks for your time.

-- jv


From pbjorn at uswest.net  Tue Dec 12 12:35:07 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Tue, 12 Dec 2000 10:35:07 -0700
Subject: Inheriting the @ sign from Ruby
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu>
Message-ID: <3A3661CA.E494E1E6@uswest.net>

Roy Katz wrote:

> Hello,
>
> I was reading a Dr. Dobb's recently and it mentioned that Ruby uses the @
> symbol to denote instance members of a class.  This appeals to me because
> I can comprehend it faster than seeing 'self' pasted in front of a
> name, especially when the statement is buried in dense code.
>
> Why not bind @ to the first argument of method definitions? Consider:
>
>   class X:
>     def __init__( self, x ):
>       @bind(x)     # binds to 'self', because
>                    # 'self' is the first argument to __init__
>     def bind(myself, x):
>       @x = x   # equivalent to myself.x
>
> What faults or strengths do you see with this?

Besides being visually unpleasant, it wouldn't work for the following code:

   class Outer:
      class Inner:
         def __init__(self, x):
            self.x = x
      def __init__(self, x):
         self.x = x

and various other variations (like defining a class inside a method etc.)
While these cases aren't particularly common in current Python code, they
might be when/if PEP-227 (http://python.sourceforge.net/peps/pep-0227.html)
makes it into the next version (discussion on whether functions are first
class objects in Python can now comense <wink>).

-- bjorn




From aleaxit at yahoo.com  Tue Dec 19 15:04:32 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 19 Dec 2000 21:04:32 +0100
Subject: while true: !!!
References: <B43D149A9AB2D411971300B0D03D7E8B0FBCBB@natasha.auslabs.avaya.com> <v04220801b663769e1800@[192.168.11.33]> <mailman.977147642.18708.python-list@python.org> <91lclt02qr3@news1.newsguy.com> <slrn93st0m.6c1.qrczak@qrnik.zagroda> <91o3l501i59@news2.newsguy.com> <jJN%5.52075$x6.23325177@news2.rdc2.tx.home.com>
Message-ID: <91of2t01um1@news2.newsguy.com>

"Rainer Deyke" <root at rainerdeyke.com> wrote in message
news:jJN%5.52075$x6.23325177 at news2.rdc2.tx.home.com...
    [snip]
> > > OTOH Python does not equate characters with integers. C, C++ and Java
> > > do.
> >
> > Wrong!  C's types are always slightly fuzzy, but that does
> > NOT apply to C++ or Java; e.g., in C++:
    [snip]
> 'char' is an integer type, same as 'short' and 'long'.  'int' is just

It can hold some range of integral values, yes, and the
coercion, if needed, is implicit.

> another integer type.  It is possible (and often desirable) to use 'char's
> (or 'unsigned char's) to hold small numeric values.  The same does not
apply
> to single character strings in Python.

The only real difference I can see is that the coercion needs
to be explicit for Python, with ord and chr.

But to say that have implicit coercions is to EQUATE is
_deeply_ wrong; it would be like saying that C++ "equates"
floats with integers, for example.  It most assuredly
doesn't!  Despite the unfortunate fact that the language
standard lets you use a float value where an integral
one is expected, and vice versa.

If you mean 'char is a numeric type' (in C but not Python),
I wouldn't dispute that -- numeric types are NOT 'equated'
to each other, as was originally asserted of characters and
integers, but they _do_ tend to allow freer and more implicit
conversion than other types, in each language (when you
put C++ into the picture, with overloads and template
instantiations, the distinction between them is sharper;
I don't know of any distinction drawn _in C_ between
'integer' and 'non-integer' numeric types, that would make
'integer' somehow a significant C term -- am I just having
a brainstorm...?).


Alex





From sholden at holdenweb.com  Mon Dec  4 08:25:24 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 4 Dec 2000 08:25:24 -0500
Subject: General default exception handler
References: <lj0n2to9vs8bmv3oo49mqgoc6h7esh8t37@4ax.com>
Message-ID: <NZMW5.10915$lF4.165334@e420r-atl1.usenetserver.com>

Dale Strickland-Clark <dale at out-think.NOSPAMco.uk> wrote in message
news:lj0n2to9vs8bmv3oo49mqgoc6h7esh8t37 at 4ax.com...
> We have a substantial set of COM objects in Python that run a Web site.
>
> Currently, if an exception is raised, the user gets an unfriendly message
in their browser and we
> may never know about it.
>
> I want to log all unhandled exceptions but I'd rather not have to put a
try/except block around
> every entry point - because there are dozens of them.
>
> Is there a general way to catch all unhandled exceptions so I can log them
and pass back something
> nice and soothing to the user?
>
> I'm planning to use the traceback module to record the exception info.
>
> Thanks
> --
> Dale Strickland-Clark
> Out-Think Ltd
> Business Technology Consultants
>
If you're prepared to accept URLs which use an extended path to indicate
arguments, one poosibility is to direct all requests to a "framework page"
which then calls the required Python functionality and traps any errors it
raises.

See, for example,

    http://www.webtechniques.com/archives/1998/02/kuchling/

in which Andrew Kuchling describes the basic technique.  This seems to give
you what you want.

regards
 Steve





From Randy.L.Kemp at motorola.com  Fri Dec 15 09:23:56 2000
From: Randy.L.Kemp at motorola.com (Kemp Randy-W18971)
Date: Fri, 15 Dec 2000 08:23:56 -0600
Subject: Moving from Perl to Python
Message-ID: <69BC4C6AE83ED311BC9400805FA7B14AF84CF7@il93exp01.css.mot.com>

Many times, we get into camps (like Perl or Python), and refuse to cross over to the other side.  I started with Perl (and liked it), but later picked up Python (and liked it).  I work in an environment with Solaris Sun machines, and Perl comes pre-bundled with the operating system.  It would take an act of congress to get the Unix Administrators to load Python, but I have it loaded on my Windows machine.  It really broils down to a matter of pragmatic utility.  If I am presented with a programming program, I will turn to the language (Perl, Python, etc.) that allows me to quickly and easily solve the problem. 

-----Original Message-----
From: tim at degree.ath.cx [mailto:tim at degree.ath.cx]
Sent: Thursday, December 14, 2000 7:25 PM
To: python-list at python.org
Subject: Re: Moving from Perl to Python


Richard Zilavec <rzilavec at tcn.net> wrote:
> I hope I enjoy it.  I don't think I will ever stop using Perl.. but I
> really need something new.  

That was my conclusion as well.  I love using Python, but will never
drop Perl completely.  (Some people on this NG are still a bit bitter
about this, but hey...)

> >There are some posts by Tom Christiansen to this group that you can find
> >on Deja about Perl v Python, those may be useful.
> 
> I found this article at perl.com, it was quite good.

I really enjoyed this article when I read it as well.  This and the
mention in _Advanced Perl Programming_ that the author's favorite OO
language is Python, convinced me to try it.  ...  Too bad this dual
advocacy doesn't work both ways.

-- 
-Tim Hammerquist <timmy at cpan.org>
The mind of a bigot is like the pupil of the eye.
The more light you shine on it, the more it will contract.
	-- Oliver Wendell, Jr. Holmes





From pearu at cens.ioc.ee  Sat Dec 16 04:39:27 2000
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Sat, 16 Dec 2000 11:39:27 +0200
Subject: Loading startup file question
Message-ID: <Pine.LNX.4.10.10012161137360.655-100000@kev.ioc.ee>

Hi,

I have defined environment variable

PYTHONSTARTUP=~/.pythonrc.py

that loads rlcompleter stuff when python is started without arguments:

sh> python

Often I use -i option in connection with -c option. For example,

sh> python -i -c 'from Numeric import *'

But then python does not load $PYTHONSTARTUP.

So, my question is: Is there a simple way to force python to load startup
file when both -i and -c option are used?

Thanks,
	Pearu




From mballen at erols.com  Sun Dec 31 03:23:38 2000
From: mballen at erols.com (Michael B. Allen)
Date: 31 Dec 2000 08:23:38 GMT
Subject: SMB Module
Message-ID: <slrn94trot.b73.mballen@angus.foo.net>

I frequent the Samba mailing list. A couple of days ago Richard Sharp
put together an smbclient.so library of client functions for SMB. SMB
is Microsoft's native file and print sharing protocol. Someone wrote a
Perl module using it. I just thought one of you guys might want to do the
same for Python. Seems only natrual since there's a nice little library
all packaged up with open(), read(), write(), ...etc all setup easy like.

Mike

-- 
signature pending


From bowman at montana.com  Mon Dec 11 09:28:10 2000
From: bowman at montana.com (bowman)
Date: Mon, 11 Dec 2000 07:28:10 -0700
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com>
Message-ID: <jA5Z5.4112$XY.25034@newsfeed.slurp.net>

Jaroslav Gergic <j_gergic at yahoo.com>
>
> OK, it is ugly to write the same line of code twice,
> but 'while true' or 'while 1' is very non-programmer construction
> it smells like VisualBasic GOTO.

It is an very common idiom in many languages. Get used to it.
Another variant is for (; ; ) in languages that support that style
of for iteration. And I've never used VB....







From junaftnoon at nospamplzyahoo.com  Sat Dec 30 07:16:13 2000
From: junaftnoon at nospamplzyahoo.com (June Kim)
Date: Sat, 30 Dec 2000 21:16:13 +0900
Subject: how to sort a dictionary by the values ?
References: <Pine.GSO.4.10.10012292313350.21787-100000@ux11.cso.uiuc.edu>
Message-ID: <92kjor$ll7$1@news.nuri.net>

"hwan-jo yu" <hwanjoyu at students.uiuc.edu> wrote in message
news:Pine.GSO.4.10.10012292313350.21787-100000 at ux11.cso.uiuc.edu...
> Hi,
> If I sort the items() of dictionary, it is sorted by first the key and
> second the value.
> How can I sort it by the values ?
> Thanks in advance.
>
>

There are several ways but one of the simplest, not the fastest, would be:
>>> dic={ ... }
>>> z=dic.items()
>>> z.sort(lambda x,y:cmp(x[1],y[1]))

Best regards,

June



From vsbabu at erols.com  Thu Dec 21 17:59:55 2000
From: vsbabu at erols.com (Satheesh Babu)
Date: Thu, 21 Dec 2000 17:59:55 -0500
Subject: grammar & spell check
Message-ID: <91u20g$pl3$1@bob.news.rcn.net>

Hi,

Is there any suggestions on what services/software I can use
to do grammar/spell check? for English language.

requirements:
1. given a text string, it should do the checks and return
errors/suggestions report
2. Should be able to interface from Python
3. Preferably should work from Win32 systems
      (that gives an idea, can't I use  MS Word's functions for this?)

Thanks
vsb




From root at rainerdeyke.com  Thu Dec 14 21:06:36 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Fri, 15 Dec 2000 02:06:36 GMT
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <tcmay-FCA22F.14073713122000@news.got.net> <918v7a02scn@news1.newsguy.com> <tcmay-04E80C.16181313122000@news.got.net> <91a1ed02a4j@news2.newsguy.com> <3A389B1A.4F33FFD4@cfmu.eurocontrol.be> <91ae8s02llf@news2.newsguy.com> <3A397778.2E6D84D3@cosc.canterbury.ac.nz>
Message-ID: <M0f_5.27169$x6.14244092@news2.rdc2.tx.home.com>

"Greg Ewing" <greg at cosc.canterbury.ac.nz> wrote in message
news:3A397778.2E6D84D3 at cosc.canterbury.ac.nz...
> To mention something vaguely Python-relevant:
> if Python had a convenient syntax for declaring
> singletons, it could be a pretty good language for
> Interactive Fiction...

Meta-classes can turn a class statement into a singleton:

class A(Singleton):
  ...

(where Singleton is an instance of the (meta-)class that implements
singletons, not listed here because I haven't written or found it yet)

It would be nice if the Singleton above was in module __builtin__ or at
least somewhere in the standard library.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From boud at rempt.xs4all.nl  Sat Dec 16 15:18:57 2000
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 16 Dec 2000 20:18:57 GMT
Subject: Compiling PyQT-2.2
References: <TSN_5.8418$wX2.112805@typhoon.austin.rr.com>
Message-ID: <91ginh$6dm$1@news1.xs4all.nl>

Marcus Urban <MURBAN at houston.rr.com> wrote:
> I've been trying to build PyQt for several days now and it continues to
> fail during the "make install" process. SIP compiles just fine. No
> complaints during ./configure or make/install. PyQt configures fine, too,
> with no complaints. The make process succeeds. But "make install" aborts.
> And yes, QTDIR is set correctly.

<...>

It doesn't immediately suggest something to me - other than a possible
mismatch between Python versions. Perhaps you could try asking the
question on the PyQt mailing list? Maybe someone over there knows
the answer; otherwise you could always ask Phil.

The address of the mailing list is:
  http://mats.gmd.de/mailman/listinfo/pykde.

-- 

Boudewijn Rempt  | http://www.valdyas.org


From gregj at pobox.com  Mon Dec 11 01:39:22 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Mon, 11 Dec 2000 06:39:22 GMT
Subject: type vs. class
References: <3A34244A.415F4DFB@t-online.de>
Message-ID: <uE_Y5.133498$U46.4497870@news1.sttls1.wa.home.com>

"Thomas Thiele" <tnt-audio at t-online.de> wrote:

> >>> class X:
> pass
>
> >>> x = X()
> >>> type(x)
> <type 'instance'>
>
> How can I realize that I will not get type 'instance' but type 'X'? I
> have different classes, more than one, and I want to check out which
> class it is.  Are all classes type instance?

In Python a class does not create a new type. All classes are of type
ClassType, and all class instances are of type InstanceType.

The built-in function isinstance(object, class) may be what you need:

>>> class A: pass
>>> a = A()
>>> isinstance(a, A)
1

isinstance returns true if the object is a member of the specified class or
any class derived from the specified class:

>>> class B(A): pass
>>> b = B()
>>> isinstance(b, A)
1
>>> isinstance(a, B)
0

The built-in function issubclass(A, B) may be useful to you:

>>> issubclass(B, A)
1

isinstance can also be used with the built-in types:

>>> import types
>>> isinstance(3, types.IntType)
1

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From mail at codex.pe.kr  Tue Dec 26 10:17:07 2000
From: mail at codex.pe.kr (Ahn, Jae-wook)
Date: Wed, 27 Dec 2000 00:17:07 +0900
Subject: Can't load jpeg image with PIL (win32)
References: <9241ls$ks3$1@news1-2.kornet.net> <874rzsh2g9.fsf@optonline.net>
Message-ID: <92acsd$4ni$1@news1-2.kornet.net>

Dear Mr. Scahffer.

Thank you very much for your help. I downloaded your binary package and
installed it, but I can't still read jpeg images. Here I include the error
message.

PythonWin 2.0 (#8, Oct 19 2000, 11:30:05) [MSC 32 bit (Intel)] on win32.
Portions Copyright 1994-2000 Mark Hammond (MarkH at ActiveState.com) - see
'Help/About PythonWin' for further copyright information.
>>> import Image
>>> im = Image.open("c:/temp/02.jpg")
Traceback (innermost last):
  File "<interactive input>", line 1, in ?
  File "d:\program files\python20\pil\Image.py", line 894, in open
    raise IOError, "cannot identify image file"
IOError: cannot identify image file
>>>

Other formats like gif, and bmp have no problem. I extracted PIL package
into my Python2.0 directory and extracted your zip file into the same
location. Maybe I made a mistake, but I can't figure it out. Please help me.


"Les Schaffer" <schaffer at optonline.net> wrote in message
news:874rzsh2g9.fsf at optonline.net...
> > Do I have to install other components like IJG jpeg library? Please
> > let me know how I can read and handle jpeg images with Python on
> > Windows platform.
>
> you need a version of PIL with jpeg support compiled in. i have
> created some packages for Windows and MS VC++ if you want to build one
> yourself or if you just want pre-compiled modules and jpeg library,
> see:
>
> http://folks.astrian.net/godzilla
>
> les schaffer




From tjanz at   Wed Dec 13 10:18:17 2000
From: tjanz at  (Tara Janz)
Date: Wed, 13 Dec 2000 15:18:17 GMT
Subject: python
Message-ID: <976703135.615739481@news.min.net>

-- 



From rodsenra at correionet.com.br  Sat Dec  9 08:36:58 2000
From: rodsenra at correionet.com.br (Rodrigo Senra)
Date: Sat, 09 Dec 2000 11:36:58 -0200
Subject: UML and Python
References: <90o91k$eks$1@desperado.vew-telnet.net> <90stth$mea$1@newsreaderg1.core.theplanet.net>
Message-ID: <3A32357A.683FB9CD@correionet.com.br>

Franz GEIGER wrote:
> 
> I'm currently lookin' for a decent priced UML IDE. All C++ UML tools are
> rather expensive. Do you have already any experiences on ObjectDomain?

> Rainer Rasche <rrasche at dspace.de> schrieb in im Newsbeitrag:
> > is somebody out there, who is interested in developing python applications
> > with UML-CASE Tools (e. g. ObjectDomain) and wants to exchange some
> > experiences with me?

An interesting approach would be to support Python-code
generation from the dia(http://www.lysator.liu.se/~alla/dia)
environment.
Not a immediate solution though ;o)
regards,
Rod

-- 
Rodrigo Senra         
Computer Engineer   (GPr Sistemas Ltda)  rodsenra at correionet.com.br 
MSc Student of Reflection (IC- UNICAMP) Rodrigo.Senra at ic.unicamp.br
http://www.ic.unicamp.br/~921234 (see also   http://www.gpr.com.br)





From mmueller at dgfz.de  Tue Dec 12 07:21:52 2000
From: mmueller at dgfz.de (Mike Müller)
Date: Tue, 12 Dec 2000 13:21:52 +0100
Subject: passing list from C to Python
Message-ID: <9154s2$2sb@micky.ibh-dd.de>

How can I pass a list of values from C to Python an get the list back after
Python did some data manipulation on it?

I am not very familiar with C and use it only to get a call to a DLL from a
FORTRAN program . The functionality of the DLL is in Python. I am able to
call a function in python from the DLL  and get a return value like:

def test_int4(x,y):
    return x+y

I can do the same with doubles, so far everything is ok. All the win32/VC++
specific stuff seems to work.

Now my problem:
I have a lot of values which should be passed by reference to Python, i.e. I
want to change the values of all the arguments.

My approach would be to put them in a PyListObject and pass it to Python
(PyObject_CallMethod does nit work though). As return value I would get a
new PyListObject back which I could ?unpack? to the new values of the FORTAN
function arguments.

Python:

def func(list):
      newList = []
      for x in list:
            y = do stuff(x)
           newList.append(y)
     return newList

VC++ 6.0

extern "C" {  __declspec(dllexport) int  C_func_called_in DLL(double *arg1,
double *arg2)
// this works
{ Py_Initialize();
        if(PyErr_Occurred()) {
            PyErr_Print();
            return FALSE;
module_object = PyImport_ImportModule("pythonModul");
        if(PyErr_Occurred()) {
            PyErr_Print();
            return FALSE;
        }
        if(!module_object) return FALSE;

PyListObject *list = Py_BuildValue("[ff]?, *arg1, *arg2);
PyErr_Clear();

// the following does not work, need something that calls python function
with list
    PyListObject *newList = PyObject_CallMethod(module_object,

"pythonModul", list);

PyObject* dataFromPy1 = PyList_GetItem(newList, 0);
PyObject* dataFromPy2 = PyList_GetItem(newList, 1);
*arg1 = PyFloat_asDouble(dataFromPy1);
*arg1 = PyFloat_asDouble(dataFromPy2);

return 0; //everything is ok
}


This might be faulty code, it just shows my intention what I WOULD like to
do.

If there is any other simple way to do it, please let me know.
Hope things are not too confusing.
Thanks.

Mike





From amk at mira.erols.com  Thu Dec 28 17:59:55 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 28 Dec 2000 22:59:55 GMT
Subject: Chronology of the founding fathers (was: Python vs. PHP (& Java?))
References: <LOBBJMHOIFOLFIPFANHJMEMHCFAA.beno@TheWebsons.com> <mailman.977912052.14706.python-list@python.org> <54A5DC0883A6A83D.617801CCE49B11D9.FA1AC6E3617A8BBC@lp.airnews.net>
Message-ID: <slrn94ni8l.jk.amk@207-172-57-128.s128.tnt2.ann.va.dialup.rcn.com>

On 27 Dec 2000 05:00:19 -0600, 
	Cameron Laird <claird at starbase.neosoft.com> wrote:
>1991:  TBL releases Web server and WorlDwidEweb to hyper-
>       text community.  He publishes specs for HTTP,
>       HTML, and (slightly disguised) URIs.

   If you have a browser from CERN's WWW project (World-Wide Web, a
   distributed hypertext system) you can browse a WWW hypertext version
   of the manual...

   Guido van Rossum, 19 Nov 1992 [First mention of the Web on
   python-list.]

--amk



From matt at mondoinfo.com  Fri Dec  8 00:21:10 2000
From: matt at mondoinfo.com (Matthew Dixon Cowles)
Date: Fri, 08 Dec 2000 05:21:10 GMT
Subject: window placement w/ Tkinter
References: <t30nffdgb43t80@corp.supernews.com>
Message-ID: <slrn930ru6.v4.matt@happy-hour.mondoinfo.com>

On Fri, 8 Dec 2000 10:00:27 -0000, Tony <colaveaj at sirinet.net> wrote:

>I am creating GUI apps and cant figure out how to set the placement
>on the screen of my frames.  I dont see anything in the constructor
>for it.  How is this done?

To set the geometry of your windows, you need to use the geometry()
method. See (wrapped for line length):

http://www.pythonware.com/library/tkinter/introduction/
  x9489-window-geometry-methods.htm

in Fredrik Lundh's excellent An Introduction to Tkinter.

Regards,
Matt


From aleaxit at yahoo.com  Tue Dec 12 07:43:02 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 12 Dec 2000 13:43:02 +0100
Subject: A question.
References: <9152a0$acb$1@panorama.wcss.wroc.pl>
Message-ID: <9156h002uh2@news2.newsguy.com>

"DeHa" <deha at poczta.onet.pl> wrote in message
news:9152a0$acb$1 at panorama.wcss.wroc.pl...
> Hi All!
>
>     I am new to Python, and just try to write my first program. But I have
a
> small problem.
>
>     My program should find first numbers in the set from 1 to 100000. I
made

I assume you mean 'prime' numbers.

> such a code:
>
> first = range(100001)
> first = first[1:]

first = range(1,100001)

would be equivalent.  The presence of '1' there, right at
the start, incidentally means that if the following code
worked at intended it would presumably delete everything
else from the list (as everything is divisible by one).


> for i in first:
>     if i != []:
> #until end of the list

i will never equal an empty-list, as first's items are just integers.

>         for j in first[first.index(i):]:                           #find
all
> numbers divided by i
>             if j % i == 0:
>                 first[first.index(j): first.index(j)] = []    #and delete
it
> from the list

Modifying the same list on which you are looping leads to
undefined behavior.

The 'for j' starts by assigning to j the value i -- that is what
first[first.index(i)] is worth -- and of course j % i WILL then
be zero and the if will succeed -- if the body of the if did
anything, such as removing something from the list, it would
start by removing the very thing it has found.

But the body of the if does nothing.  first[x:x] is an empty
list for any x, so assigning an emptylist to it is a no-op.

All in all, then, there are many problems here.  Perhaps
foundational, the fact that you're looping on items but
then repeatedly find out you need *indices* instead.  So,
why not loop on indices -- should be faster, might be
clearer.

Let's start with a range of candidates-for-primality that
does NOT start from 1, but from 2:

prime_candidates = range(2,100001)

i = 0
while i<len(prime_candidates):
    # loop invariant: prime_candidates[i] is prime
    aprime = prime_candidates[i]
    j = i + 1    # we must NOT remove the i-th item itself!!!
    while j<len(prime_candidates):
        # note we EITHER delete one item OR increment j,
        # as len(prime_candidates)-j goes down by ONE AT A TIME...
        if prime_candidates[j] % aprime == 0:    # divisible by the prime
            del prime_candidates[j]
        else:
            j += 1
    i += 1


The remaining problem with this is its woeful inefficiency.  The
algorithm is quadratic with the top of the range, and each step
of it is horribly slow, too -- dividing, deleting from the middle
of a long list, etc.  Here are a couple of runtimes in seconds:

With a top-of-range of 1001:

D:\PySym>python prim.py
0.145141311217
found 168 primes

and with a top-of-range of 10001:

D:\PySym>python prim.py
7.39378988285
found 1229 primes

I'm not patient enough to run it for 100001, but it would
surely be well over 350 seconds.


A very similar idea can be exploited, *without* such inefficiencies,
if the 'prime_candidates' list is NOT continuously modified.  The
'deletion' can then take the form, for example, of just setting a
bit in another list 'parallel' to the main one.  This also lets us
apply Eratosthenes' original ideas -- rather than doing any
division to check the remainders, we just use the size of the
latest prime number found as the 'step' to use for further walking
on the list!  Again, we'll loop on indices -- more convenient.

Also, we don't actually need to keep 'prime_candidates' around,
since any 'prime_candidates[i]' is just i+2 -- easier to compute
it on the fly (again, Eratosthenes' idea).

total_length = 100001 - 2 + 1        # 'len(prime_candidates)'
primality_bits = [1]*total_length

for i in range(0, total_length):
    if not primality_bits[i]: continue
    # we proceed iff "prime_candidates[i]" is prime
    aprime = i + 2
    for j in range(i+aprime, total_length, aprime):
        primality_bits[j] = 0


Notice how much simpler this is now.

And, correspondingly faster... again for top values of
1001, 10001, 100001:

D:\PySym>python prima.py
0.009498131886
found 168 primes

D:\PySym>python prima.py
0.0985063468943
found 1229 primes

D:\PySym>python prima.py
1.07798139764
found 9592 primes

Not quite linear, but almost. Also: about 145 times faster
than the first version for 1001, about 740 times for 10001
(probably more still, for 100001).


Dynamical, interpreted languages such as Python are
not *mostly* about performance -- but hundreds or even
thousands of times of slowdown/speedup can easily be
had in simple cases by paying a _little_ attention to
performance issues.  I think it can be worth it...


Alex





From tgagne at ix.netcom.com  Wed Dec 20 22:42:51 2000
From: tgagne at ix.netcom.com (Thomas Gagne)
Date: Wed, 20 Dec 2000 22:42:51 -0500
Subject: How to get an (db)exception's string
References: <3A40EEBF.3262040F@ix.netcom.com> <ulmtahl1h.fsf@ctwd0143.fitlinxx.com>
Message-ID: <3A417C3B.E226C419@ix.netcom.com>

New question.

The variable I'm getting back appears to be a dictionary:

[{'server': '', 'msgnumber': 2812, 'proc': '', 'line': 3, 'text': "Stored
procedure 'xp_efloat_add' not found. Specify owner.objectname or use sp_help
to check whether the object exists (sp_help may produce lots of output).\012",
'severity': 16, 'sqlstate': 'ZZZZZ', 'state': 5}]

I need to capture the text above in a string so I can transmit it to a client
application.  The exception variable itself is not a string, but it knows how
to print itself.  How can I capture that string so I can send it?

--
.tom





From syver at NOSPAMcyberwatcher.com  Thu Dec 21 03:24:12 2000
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Thu, 21 Dec 2000 09:24:12 +0100
Subject: mail processing module?
References: <91rcig$rgu$1@nnrp1.deja.com>
Message-ID: <N7j06.27$Zq.606@news1.oke.nextra.no>

Sounds cool, don't have specific use for it just now, but...

"Preston Landers" <prestonlanders at my-deja.com> wrote in message
news:91rcig$rgu$1 at nnrp1.deja.com...
> Hi all.
>
> I'm trying to find out if there's any interest in a kitchen-sink Python
> module for mail handling.
>
> I've got some code (for my company's internal product) that, with a
> little bit of polishing, could handle most people's mail sending and
> receiving needs in a pretty straightforward way.  Also I have not got
> managerial clearance for releasing this code, as I'd like to see if
> there's any interest before approaching them.
>
> This module lets you send mail (using smtplib) and retrieve mail from
> either a POP3 or IMAP4 server.  It handles encoding and decoding MIME
> messages with attachments.  I wrote it because I needed something to
> intelligently handle sending multipart MIME attachments.  It uses the
> standard Python smtp/pop/imap/MIME modules, just basically ties them all
> together.
>
> Probable code license would be BSDish.  Post here or email me privately
> at pibble at yahoo dot com if you are interested.
>
> thanks
> Preston
>
> --
> || Preston Landers <prestonlanders at my-deja.com> ||
>
>
> Sent via Deja.com
> http://www.deja.com/




From pblanchette at pixelsystems.com  Wed Dec 13 09:18:28 2000
From: pblanchette at pixelsystems.com (Patrick Blanchette)
Date: Wed, 13 Dec 2000 14:18:28 GMT
Subject: win32net.NetWkstaSetInfo - How to use it?
Message-ID: <3A378524.1090400@pixelsystems.com>

I,
   I'm trying to set the computer name of a WinNT workstation using the 
NetWkstaSetInfo and the NetWkstaSetInfofunctions.  When I call 
NetWkstaSetInfo, It always returns the errorno 998:Invalid access to 
memory location.


 >>> info = NetWkstaGetInfo(None, 100)
 >>> info['computername'] = Unicode('NewComputerName')
 >>> NetWkstaSetInfo(None, 100, info)
Traceback (innermost last):
  File "<interactive input>", line 1, in ?
api_error: (998, 'NetWkstaSetInfo', 'Invalid access to memory location.')


Whats wrong?  How should I use these functions?


Thanks in advance.



From dustin at blacksky.org  Fri Dec  8 23:25:36 2000
From: dustin at blacksky.org (J Dustin Flesher)
Date: Fri, 8 Dec 2000 21:25:36 -0700
Subject: wxpython
Message-ID: <zuiY5.2602$kf3.429572@news.uswest.net>

How do I properly dispose of a wxgrid that uses a custom wxPyGridTableBase?
The demo application that comes with wxpython shows a leak.

Thanks for any input.




From tim at degree.ath.cx  Thu Dec 14 20:24:42 2000
From: tim at degree.ath.cx (Tim Hammerquist)
Date: Fri, 15 Dec 2000 01:24:42 GMT
Subject: Moving from Perl to Python
References: <3a3a9552.76673185@news.tcn.net> <slrn93fcdt.a8m.scarblac@pino.selwerd.nl> <3a38b798.85448901@news.tcn.net>
Message-ID: <slrn93isni.8bg.tim@degree.ath.cx>

Richard Zilavec <rzilavec at tcn.net> wrote:
> I hope I enjoy it.  I don't think I will ever stop using Perl.. but I
> really need something new.  

That was my conclusion as well.  I love using Python, but will never
drop Perl completely.  (Some people on this NG are still a bit bitter
about this, but hey...)

> >There are some posts by Tom Christiansen to this group that you can find
> >on Deja about Perl v Python, those may be useful.
> 
> I found this article at perl.com, it was quite good.

I really enjoyed this article when I read it as well.  This and the
mention in _Advanced Perl Programming_ that the author's favorite OO
language is Python, convinced me to try it.  ...  Too bad this dual
advocacy doesn't work both ways.

-- 
-Tim Hammerquist <timmy at cpan.org>
The mind of a bigot is like the pupil of the eye.
The more light you shine on it, the more it will contract.
	-- Oliver Wendell, Jr. Holmes


From sdm7g at virginia.edu  Mon Dec 18 14:44:36 2000
From: sdm7g at virginia.edu (Steven D. Majewski)
Date: Mon, 18 Dec 2000 14:44:36 -0500 (EST)
Subject: search for pattern in StringList
In-Reply-To: <slrn93shus.4tp.scarblac@pino.selwerd.nl>
References: <3a3e1935.5461298@news.skynet.be> <slrn93shus.4tp.scarblac@pino.selwerd.nl>
Message-ID: <Pine.NXT.4.21.0012181442020.11798-100000@localhost.virginia.edu>

Except it should be: filter( function, sequence )

>>> print filter( lambda x: 'a' in x, MyList )
['an', 'call']


-- Steve Majewski <sdm7g @ Virginia.EDU> 


On 18 Dec 2000, Remco Gerlich wrote:

>  
> This is a very common thing to do, select some items from a list based
> on some criterium. filter() is a built-in function that does this:
> 
> print filter(MyList, lambda x: 'a' in x)
> 
> It "filters" the list according to the simple function given by the lambda,
> that tests if 'a' in x.
> 
> Since Python 2.0, there is yet another way to write this:
> 
> print [s for s in Mylist if 'a' in s]
> 
> This is derived from set notation in math; it is 'the list of s, where
> s comes from MyList and only those where "a" in s'.
> 




From fredrik at effbot.org  Mon Dec 11 15:56:11 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Mon, 11 Dec 2000 20:56:11 GMT
Subject: Python2 distutils question: how to best "autoconfig"...?
References: <9130e60obc@news2.newsguy.com> <82aZ5.3458$Kd1.395204@newsb.telia.net> <913br4012r8@news2.newsguy.com>
Message-ID: <LbbZ5.3172$O%.511511@newsc.telia.net>

Alex Martelli wrote:
> Good point, which Tim Peters also did in his mail!

interesting.  I thought Tim was on a vacation, but his
posts show up in the python.org archives.  maybe my
ISP has installed a timbot filter?

</F>




From tim.one at home.com  Wed Dec 27 22:29:15 2000
From: tim.one at home.com (Tim Peters)
Date: Wed, 27 Dec 2000 22:29:15 -0500
Subject: A Suggestion for Python Colon Syntax
In-Reply-To: <92d72h017po@news2.newsguy.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEKIIFAA.tim.one@home.com>

[Alex Martelli]
> ...
> Because "A foolish consistency is the hobgoblin of little minds"
> (Ralph Waldo Emerson)?

Nope!  Guido van Rossum coined this for the subtitle of

    http://www.python.org/doc/essays/styleguide.html

Unfortunately, Guido took a copy of that back with him on one of his
time-machine trips, to New England in order to pave the way for me to move
there 145 years later.  He rested his head on it while making repairs
underneath, and left it there by mistake.  A young Transcendentalist found
it in the woods near Concord, MA, and his friends grew quite fond of it
despite not understanding much.  Thoreau went on to make that principle the
basis of his form of poetry:

# Local Variables:
# py-indent-offset: 8
# End:
# My Captain:  Oh, My Captain!

Emerson liked the subtitle so much that he borrowed it for his ironically
named essay "Self-Reliance", but tacked on "adored by little statesmen and
philosophers and divines" to calm his guilty conscience.

The rest-- as they say --is history.

no-coincidence-that-emerson-and-thoreau-eschewed-tab-characters-ly
    y'rs  - tim




From grey at despair.rpglink.com  Sun Dec 31 03:10:11 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Sun, 31 Dec 2000 08:10:11 -0000
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <slrn94smds.150.grey@teleute.rpglink.com> <FOt36.17$vd1.2224@e420r-atl2.usenetserver.com> <slrn94t1vj.150.grey@teleute.rpglink.com> <mqw36.50$9s1.10042@e420r-atl2.usenetserver.com>
Message-ID: <slrn94tqf3.150.grey@teleute.rpglink.com>

On Sun, 31 Dec 2000 01:59:46 GMT, Kragen Sitaker <kragen at dnaco.net> wrote:
>Some of us continue to gain new insights about programming by listening
>to those who disagree with us, despite having 20 years of experience
>programming.  Our humility removes limits to our knowledge and achievement.

    Yes, I have gained insights from others on this newsgroups and I respect
them for it.  OTOH, there are certain people, certain statements which simply
are too stupid to let slip.  That was one of them.  

>It's sad to watch, but funny, too, in a sad way.

    Whatever.  *plonk*

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From tim.one at home.com  Sat Dec 23 12:35:50 2000
From: tim.one at home.com (Tim Peters)
Date: Sat, 23 Dec 2000 12:35:50 -0500
Subject: <> vs != was Re: A Suggestion for Python Colon Syntax
In-Reply-To: <20001223180819.E5515@xs4all.nl>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEEDIFAA.tim.one@home.com>

[Tim]
> Best guess is that "<>" will go away in P3K.  I jokingly told
> Guido it was clearly wrong for an inequality operator to look
> so symmetric, and he forgot to laugh <wink>.

[Thomas Wouters]
> Well, excuse me, but it's clearly wrong for an equality test
> operator to look so much like an augmented assignment statement !
> I definately prefer '<>' myself, anyway, I just wish I could
> use it in C without turning towards preprocessor macros :)

You're missing the Big Picture, though:  Guido's brother Just is a <> fan
too.  That means <> is doomed.  It's a sibling rivalry thing <wink>.

just-killed-stackless-and-"for-indexing"-too-ly y'rs  - tim




From root at rainerdeyke.com  Sat Dec 16 22:26:33 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Sun, 17 Dec 2000 03:26:33 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <r0pzoi11lhz.fsf@black.cs.washington.edu> <5jvZ5.3588$Kd1.418770@newsb.telia.net> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com> <kuu28971kp.fsf@lasipalatsi.fi> <3A380E4E.8537A0FB@san.rr.com> <91a1vr02b02@news2.newsguy.com> <Co9_5.25465$x6.13756579@news2.rdc2.tx.home.com> <91bbq102mtk@news1.newsguy.com> <rWd_5.26820$x6.14119694@news2.rdc2.tx.home.com> <91d5el0me7@news1.newsguy.com> <rOr_5.31120$x6.15241856@news2.rdc2.tx.home.com> <slrn93o7j4.tu0.quinn@dinar.ugcs.caltech.edu>
Message-ID: <JnW_5.37708$x6.18119061@news2.rdc2.tx.home.com>

"Quinn Dunkan" <quinn at dinar.ugcs.caltech.edu> wrote in message
news:slrn93o7j4.tu0.quinn at dinar.ugcs.caltech.edu...
> On Fri, 15 Dec 2000 16:38:47 GMT, Rainer Deyke <root at rainerdeyke.com>
wrote:
> >Yes, in-place modification should be supported.  I had something like
this
> >in mind:
> >
> >class UserInt:
> >  def __init__(self, value):
> >    self.data = value
> >  def __add__(self, other):
> >    if isinstance(other, UserInt):
> >      return self.__class__(self.data + other.data)
> >    else:
> >      return self.__class__(self.data + other)
> >  __radd__ = __add__
> >  def __iadd__(self, other):
> >    if isinstance(other, UserInt):
> >      self.data += other.data
> >    else:
> >      self.data += other
> >  # Repeat for all other operations
> >
> >UserInt as written is more properly UserNumber, since it supports longs
ints
> >and floats.
>
> Just as a matter of style, would it be superior to add a __coerce__ method
> that knows how to coerce ints, longs, and floats into UserInts, or make an
> explicit check on each method?  __coerce__ saves a lot of typing, but I'm
> generally wary of implicit things like coercion.  When I implemented a
class
> like this (a vector, actually), I used __coerce__, but after trying to
wrap my
> brain around the control flow, I was thinking maybe I wouldn't next time,
even
> though it seems to work fine.

__coerce__ would probably have been superior, since it eliminates duplicate
code and more closely resembles the way real integers work in Python.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From aleaxit at yahoo.com  Wed Dec 20 06:54:28 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 20 Dec 2000 12:54:28 +0100
Subject: while true: !!!
References: <mailman.977180341.10206.python-list@python.org> <slrn93t7bv.le.grey@teleute.rpglink.com> <KTz%5.48644$x6.22165257@news2.rdc2.tx.home.com> <91nc7h0s0p@news2.newsguy.com> <mailman.977231769.15324.python-list@python.org> <AwL%5.80470$15.17054718@news1.rdc1.az.home.com> <3A4024E5.98843CFC@cosc.canterbury.ac.nz>
Message-ID: <91q6mb01d22@news2.newsguy.com>

"Greg Ewing" <greg at cosc.canterbury.ac.nz> wrote in message
news:3A4024E5.98843CFC at cosc.canterbury.ac.nz...
> Tim Hochberg wrote:
> >
> > If I attempt to summarize I get the following for false items:
> >
> > * None
> > * Anything that equals zero (Including a class instance with __nonzero__
> > returning false)
> > * An empty sequence (Including a class instance with __len__ returning
0)
>
> The fact that it's taken about three attempts by
> experienced Python people to come up with a full
> and accurate characterisation of what constitute
> 'true' and 'false' in Python suggests that things
> could have been a lot simpler if there had been a
> dedicated boolean type!

I disagree that it suggests that.  There is a dedicated
int type, but charactezing accurately all expressions
that yield an integer 0 (when fed to the 'int' function)
is still a bother -- even "all _objects_ that yield an
integer 0 when fed to 'int'" takes quite a bit.

If there was a dedicated bool type, and of course a
'bool' function, it would take *exactly* as much effort
to enumerate all X's such that:
    if bool(X):
        print "bu!"
prints 'bu', as it now takes to enumerate all X's
such that the simpler:
    if X:
        print "bu!"
prints the same.  That's *IF*, of course, the precious
characteristic of polymorphism is maintained with its
current power!  Polymorphism _could_ be taken away, at
a considerable loss of power and handiness of course,
and _that_ would no doubt ease the specific task of
'listing all object that behave in this way' -- without
polymorphism, 'objects that behave in this way' are
a finite and predetermined set; WITH polymorphism,
the answer to "what objects behave in this way" always
has to be "all, and those only, which WANT to behave
in this way".  I dispute that lack of polymorphism is
'simpler', because the typical use-case is NOT that of
enumerating objects that behave in certain ways, but
rather that of _designing_ an object to behave in a
certain way (which for the boolean case translates to
'implement __nonzero__, or, missing that, __length__')
or _using_ an object you've received without worrying
WHY it chooses to behave thus and thus (and in THIS
very typical situation, polymorphism *simplifies* your
life enormously!).


So, summarizing: whether a dedicated boolean type exists
or not has no real bearing on the issue; polymorphism
does (if one was so crazy as to want to remove it, one
might perfectly well remove it without introducing a
'dedicated boolean type'), but it's an optical illusion
to thing that removing it would 'simplify' things!


Alex





From jim at interet.com  Wed Dec 13 12:59:51 2000
From: jim at interet.com (James C. Ahlstrom)
Date: Wed, 13 Dec 2000 12:59:51 -0500
Subject: zipfile.py
References: <3A37AD3E.7233E7B7@ex.ac.uk>
Message-ID: <3A37B917.6AC49FEA@interet.com>

"M.J.Fisher" wrote:
> 
> Hi,
> 
> I have been having some trouble with using zipfile.py, I need to unzip
> .zip files some created in windows and
> some in unix.
> 
> zlib.error: Error -3 while decompressing: incomplete dynamic bit lengths
> tree
> 

It looks like there is a version incompatibility in your
version of zlib on Windows or Unix.  Zipfile depends on zlib
to compress/uncompress files.  You can test this by specifying
a compression method of ZIP_STORED for all files so they are not
compressed.  Or you could check zlib versions.

Jim Ahlstrom


From aleaxit at yahoo.com  Fri Dec 22 06:03:49 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 22 Dec 2000 12:03:49 +0100
Subject: JPython embed question
References: <3A42AA88.D9BD321A@ncsa.uiuc.edu>
Message-ID: <91vcf5027on@news2.newsguy.com>

"Randy Heiland" <heiland at ncsa.uiuc.edu> wrote in message
news:3A42AA88.D9BD321A at ncsa.uiuc.edu...
> In the JPython demo/embed example, we find how to create an instance of
> a class, e.g.:
>
>   interp.set("foo1", new Foo());
>   interp.set("bar1", new Bar());
>
> I'd like to be able to dynamically create an instance of a class - e.g.,
>
>   createInstance(String instName, String className)

Java's "reflection" abilities is what you use for this -- most
of them are in package java.lang.reflect, although the crucial
java.lang.Class class is in java.lang for historical reasons.

Static method forName(String className) of class java.lang.Class:
"Given the fully-qualified name for a class, this method attempts
to locate, load, and link the specified class. If it succeeds,
returns the Class object representing the class".

Once you do have the java.lang.Class object, representing the
class whose fully-qualified name you started with, you call
its newInstance() method, and get an Object "just as if" you
has executed a 'new Foo()' expression.

All of this is really core Java stuff, with marginal specific
relevance to JPython.  I'd recommend the Java tutorial,
specifically http://java.sun.com/docs/books/tutorial/reflect/
for reflection issues.


Alex








From aleaxit at yahoo.com  Thu Dec 28 17:58:49 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 28 Dec 2000 23:58:49 +0100
Subject: Python for prototyping of the user interface
References: <20001228.17165633@mis.configured.host>
Message-ID: <92ggrs02jfb@news2.newsguy.com>

"Volkmar Busch" <vbusch at gmx.de> wrote in message
news:20001228.17165633 at mis.configured.host...
"""
I took my first steps in using Python, and I really like the language. I
read a lot about using Python for prototyping. As a basis for discussion
with our users we use GUI prototypes. Building complex user interfaces in
Python to me seems quite complicated, because

- I am missing a Gui builder
"""

Try Boa Constructor, on sourceforge.  Early, but promising.  I think
there are good commercial products, too.

Or, depending on the GUI's you need, dynamic HTML can be good:
there are umpteen HTML generators and editors, and, at least on
Windows platforms, with active-scripting, 'behaviors', 'HTML Applications',
etc, adding Python code to script the HTML is quite easy.

"""
- I am missing controls like notebooks and tables
"""
wxPython, the cross-platform toolkit Boa Constructor uses, has
them I think.  HTML surely does, of course.


Alex





From leonb at pqafrica.co.za  Thu Dec  7 03:09:08 2000
From: leonb at pqafrica.co.za (Leon Booyens)
Date: Thu, 7 Dec 2000 10:09:08 +0200
Subject: PIL binary for Python 2.0 for Windows ?
Message-ID: <3a2f472f$0$235@hades.is.co.za>

Where can I find a binary distribution for Windows of the Python Image
Library 1.1.1. (It is not as yet available from Pythonware as a binary and I
unfortunatly have no means of compiling the code on a windows platform)

Any help would be appreciated.

Thanks
Leon
(leonb at pqafrica.co.za)





From nas at arctrix.com  Tue Dec 19 05:36:28 2000
From: nas at arctrix.com (Neil Schemenauer)
Date: Tue, 19 Dec 2000 02:36:28 -0800
Subject: Walk through dictionary keys?
In-Reply-To: <3A3F8B89.1DD9E01F@schlund.de>; from cg@schlund.de on Tue, Dec 19, 2000 at 05:23:37PM +0100
References: <3A3F8B89.1DD9E01F@schlund.de>
Message-ID: <20001219023628.B737@glacier.fnational.com>

On Tue, Dec 19, 2000 at 05:23:37PM +0100, Carsten Gaebler wrote:
> Is there a way of 'walking' through the keys of a dictionary? If I have
> a large dictionary I'd like to prevent 
> 
> for key in d.keys():
>     do_something

Python 2.1 will have a d.popitem() method for just this purpose.

  Neil



From sragsdale at my-deja.com  Tue Dec  5 18:06:09 2000
From: sragsdale at my-deja.com (sragsdale at my-deja.com)
Date: Tue, 05 Dec 2000 23:06:09 GMT
Subject: Rounding error in Python 2.0
Message-ID: <90jscs$uno$1@nnrp1.deja.com>

Can anyone explain to me why python2 can't read a float from a string?
Python 1.5.2 didn't have this problem.

[2:58:01pm]bach:/d/bach/proj/gridTools/rnd/ 31> python2
Python 2.0 (#2, Nov 13 2000, 20:09:09)
[GCC 2.95.2 19991024 (release)] on irix646
Type "copyright", "credits" or "license" for more information.
>>> foo = 1242.34
>>> float(str(foo))
1242.3399999999999

[2:58:20pm]bach:/d/bach/proj/gridTools/rnd/ 32> python
Python 1.5.2 (#8, Jul 13 1999, 17:39:45) [C] on irix646
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> foo = 1242.34
>>> float(str(foo))
1242.34


Sent via Deja.com http://www.deja.com/
Before you buy.


From hinsen at cnrs-orleans.fr  Fri Dec  8 06:05:03 2000
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 08 Dec 2000 12:05:03 +0100
Subject: Parallel Computing
References: <3A300B1D.223630F3@bioeng.ucsd.edu>
Message-ID: <m3ofynqilc.fsf@chinon.cnrs-orleans.fr>

Curtis Jensen <cjensen at bioeng.ucsd.edu> writes:

> I haven't had much luck on finding anything about parallel computing via
> Python.  Does Python support this?  Or do we just need to write a C
> module to run parallel code?  Thanks.

It depends on what you mean exactly by "parallel code", but I suppose
that for most interpretations the answer is that you can do it in
Python using existing modules.

I'll limit myself to the kind of parallel computing that I do myself,
which is algorithm parallelization implemented via MPI. Version 2.1 of
ScientificPython (available at
ftp://dirac.cnrs-orleans.fr/ScientificPython/) contains an interface
to the most commonly used MPI functions plus the necessary
infrastructure to make Python run as an MPI application, and to permit
C extension modules to use MPI as well without linking to it
(important if you use dynamic libraries). With this package, you can
use MPI in Python in exactly the same way you would do in C or
Fortran. The data objects that can be passed between processes are
NumPy arrays and strings.

However, most real-life parallel code will also contain C code for
speed; you wouldn't use a parallel machine if speed weren't important.
But you can end up with an efficient program in which most code is
Python. For an example, see the development releases (2.1.x) of the
Molecular Modelling Toolkit (available at
ftp://dirac.cnrs-orleans.fr/MMTK/).

Konrad.
-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.56.24
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------


From j.c.yip at computer.org  Mon Dec 25 15:55:48 2000
From: j.c.yip at computer.org (Jason Che-han Yip)
Date: Mon, 25 Dec 2000 13:55:48 -0700
Subject: [ANN] XP2001 - Final Call for Papers
Message-ID: <928c8a$lrm$1@nserve1.acs.ucalgary.ca>

XP2001 - FINAL CALL FOR PAPERS

2nd International Conference on eXtreme Programming
and Flexible Processes in Software Engineering - XP2001

21-23 May, 2001, Villasimius, Sardinia, Italy

Home Page: http://www.xp2001.org

General Chair:  Kent Beck, First Class Software, Inc., USA
Program Chair:  Michele Marchesi, University of Cagliari, Italy
Program Co-chair:  Giancarlo Succi, University of Alberta, Canada


Following the big success of XP2000, the Second International Conference on
eXtreme Programming and Flexible Processes in Software Engineering is to be
a forum for discussing theories, practices and tools, and experiences on XP
and other flexible software processes in the industry, in academia, and in
research institutions.

The conference will consist of technical presentations, practitioner's
reports, panels, poster sessions, activity sessions, demonstrations,
workshops, tutorials.


IMPORTANT DEADLINES:

PAPER SUBMISSION: January 15, 2001
Acceptance notification: March 1, 2001
Final manuscript submission: April 1, 2001

Workshops and active sessions proposals: January, 15, 2001
Demo and Practitioner's Reports submission: April 1, 2001
Demo and Practitioner's Reports acceptance notification: April 30, 2001

For more information, including conference topics and instructions to
authors, please refer to XP2001 Web site: http://www.xp2001.org

or send an e-mail to:  michele at diee.unica.it

[Apologies if you receive this more than once]

--
"The most exciting phrase to hear in science, the one that heralds new
discoveries, is not 'Eureka!' (I've found it!), but "That's funny...".
(Isaac Asimov)
www.quase.ualberta.ca/yip




From gbreed at cix.compulink.co.uk  Wed Dec 20 07:58:19 2000
From: gbreed at cix.compulink.co.uk (gbreed at cix.compulink.co.uk)
Date: 20 Dec 2000 12:58:19 GMT
Subject: while true: !!!
References: <3A4024E5.98843CFC@cosc.canterbury.ac.nz>
Message-ID: <91qadb$1di$1@plutonium.compulink.co.uk>

In article <3A4024E5.98843CFC at cosc.canterbury.ac.nz>, 
greg at cosc.canterbury.ac.nz (Greg Ewing) wrote:

> The fact that it's taken about three attempts by
> experienced Python people to come up with a full
> and accurate characterisation of what constitute
> 'true' and 'false' in Python suggests that things 
> could have been a lot simpler if there had been a 
> dedicated boolean type!

Ah, but that's because the question was asked the wrong way round.  I hope 
all experienced Pythonistas would know that "" evaluates to false.  
Remembering something is easier than remembering you forgot something.

The general rule is simple: nothing is false.  Take any type, and if 
there's a way of it being nothing, it evaluates false.  So what does no 
string look like? ""  No list?  [].  And so on.  There are some anomolies, 
like [[]] looks like nothing, but such is life.  If you can't think what 
it would mean for an object to be nothing, you have no business using it 
as a boolean.

Adding a boolean type wouldn't make anything simpler.  It would only 
change the question to "What objects x will boolean(x) return false?"  
Restricting falsehood to 0 would make the language simpler, but also less 
powerful.


              Graham


From aleaxit at yahoo.com  Fri Dec 29 18:43:22 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 30 Dec 2000 00:43:22 +0100
Subject: Python vs. PHP (& Java?)
References: <69BC4C6AE83ED311BC9400805FA7B14AF84D2C@il93exp01.css.mot.com> <mailman.978018791.20357.python-list@python.org> <92fv4a$l2l$1@nnrp1.deja.com> <92gghv02iq1@news2.newsguy.com> <92iie9$lg3$1@nnrp1.deja.com>
Message-ID: <92j7q0033j@news1.newsguy.com>

<rturpin at my-deja.com> wrote in message news:92iie9$lg3$1 at nnrp1.deja.com...
    [snip]
> Turpin:
> >> Knowing one-tenth of Perl is not enough to understand
> >> most Perl scripts.
>
> Martell:
> > Of course not, but, it IS enough to _write_ a lot of
> > somewhat useful scripts ..
>
> Your use of a computer language is fairly limited if
> you can only WRITE it. Software engineering requires
> the ability to READ. And not just what you once wrote,
> but what others have written, including others who have
> been using it for years and know all the idioms and
> tricks. Even though they are long gone. This is why

Russell, are we talking software engineering, or are we
talking Perl?  You're making me dizzy with these continuous
switches between these completely unrelated subjects.

I'm not saying that Perl is a write-only language, not for
ALL, but I do maintain it is for many of its 'millions' of
'casual' users -- the ones who keep posting to c.l.perl
groups about how to do perl-unrelated CGI tasks even
though they DO get flamed to a crisp each & every time.

Such users (I claim, without rigorous empirical studies to
back me up, but, a LOT of anecdotal evidence:-) master
10% or less of the total language.  There being an average
of 7.2 ways to get any given task done, that fraction still
lets them do over half of what could be done in all, which
ain't bad, though they DO get headaches.

They don't claim to be software engineers any more than
typical foreign speakers of English claim to be scholars
in that demanding field.  They just get by... but they DO.
Mostly.


> an after dinner conversation. The interesting and hard
> productivity question is not how long it takes a
> beginner to complete a task, but how much effort it
> requires a group of professionals to maintain a sizable
> product, over its lifecycle. (This isn't meant as
> criticism of  those who want to use Python for teaching
> or personal use. It is a very good language for both
> purposes. But my interest and most of my comments are
> from the viewpoint of professional use.)

So, both questions are interesting, given that both users
(beginners and professionals) are in the language's target
market.  There need not be as much contrast as one
might think between their interests, I believe.

> simultaneously. A rich language that admits multiple
> meaning and implication is very useful. If you want to
> look for adaptation, sexual selection may have more to do
> with this than natural selection. A silver tongue will
> not save you from the lion or bag the deer, but it will
> win a heart.

Sexual selection, particularly female-choice, IS very much
a part of natural selection -- why male peacocks have
absurdly long and rich tails, etc.  "Survival" of the fittest
is the least of it -- *reproduction* is the name of the
natural selection game, and survival only adaptive in as
much as it serves reproductive purposes.

(Though in our specific cases, I'm convinced memes took
over most control a long time ago, but that's _another_
debate:-0).


> > Alternatively, of course, great complexity in implementation
> > MIGHT be spent towards such targets as optimization, as
> > you were interestingly arguing Python should be doing now.
>
> Yep. Keep in mind that only the implementation team has to

Plural, please.  Think how many C++ compilers there are, or
where, on the market.  Or, for Python: CPython, Jython, .NET,
Vyper, Stackless, other possible variants; also, the reference
implementation's clarity and simplicity encourages others who
are 'at the margins' of "the implementation team" (e..g, me)
to get gradually more involved.

> tackle the complexity of optimization, whereas the world has
> to tackle the complexity of language and semantics.

But complexity of language/semantics influences complexity
of optimization; and since the latter brings optimizer bugs, it
can affect all users (of a given implementation).  So it's not
as cut-and-dry as all that.

> > .. (from your excellent diagnostic-manual on quack
> > theories) ..
>
> It's amazing how one post in sci.med fifteen years ago has
> taken on a life of its own. I would rather you criticize
> my recent proposal for caching attribute and method
> references.

I won't, because it seems quite reasonable to me, and I'm
trying to puzzle over criticisms made by people who understand
the internals far better than I do.  Also, I do NOT understand
exactly where the time is going -- instrumented runs are
telling me that quite clearly.  E.g., on list comprehensions: I
*thought* tweaking allocation would help a lot; Tim Peters
reminded me of a thread back in August, in which I had entered
though I didn't recall that, which must have been what gave
me the idea, since somebody reported a simple patch with
huge benefits; but, having applied that patch, I cannot for
the life of me reproduce the claimed performance pluses.  So,
I'm busy trying to revise my mental performance model of
Python -- which makes my generic ideas about "caching's
always nice" not a very sharp contribution at this time.


> BTW, didn't we meet once?

Yes, we did.  Well before I met Python, actually.


Alex





From jcm at bigskytel.com  Fri Dec  8 08:30:03 2000
From: jcm at bigskytel.com (David Porter)
Date: Fri, 8 Dec 2000 06:30:03 -0700
Subject: My first attempt at python
In-Reply-To: <n2j13tcc0il8k1i3dtjkdckglo871ln744@4ax.com>; from kahuna01@hotmail.com on Fri, Dec 08, 2000 at 12:02:37PM +0000
References: <n2j13tcc0il8k1i3dtjkdckglo871ln744@4ax.com>
Message-ID: <20001208063003.A2569@bigskytel.com>

* Jason Stewart <kahuna01 at hotmail.com>:

> 	This same functionality might be in a module somewhere, but I

string.split()

> wrote it for practice. Its a string tokenizer. If you could analyze it
> and tell me where I could improve I would be grateful.

It looks good to me...

> def tokenize(string_to_chop):
> 	"Break a string up into individual words"
> 	last_space = 0
> 	token = []
> 	for index in range(len(string_to_chop)):
> 		if string_to_chop[index:index+1] == " ":

the above can be string_to_chop[index]

> 			token.append(string_to_chop[last_space:index])
> 			last_space =  index + 1
> 		elif (index+1 == len(string_to_chop)):
> 		    token.append(string_to_chop[last_space:index+1])
> 	return token

Here are two alternatives.

import re
def tokenize(s):
    r = re.compile(' ')
    return r.split(s)

def tokenize(s):
    ls = list(s)
    tokens = []
    while ls:
        try:
            space = ls.index(' ')
        except:
            space = len(ls)
        tokens.append(s[:space])
        del ls[:space+1]
        s = s[space+1:]
    return tokens

Of course, the best way is:

import string
tokens = string.split(s)


(I wonder if this will reach the newsgroup...)

  David    





From ben.hutchings at roundpoint.com  Fri Dec 15 13:42:17 2000
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 15 Dec 2000 18:42:17 +0000
Subject: launching a python program from GNU make (Win2K)
References: <918lp0$6q4$1@nnrp1.deja.com> <slrn93jn66.p9r.hat@se-46.wpa.wtb.tue.nl>
Message-ID: <ug0jpzfue.fsf@roundpoint.com>

hat at se-46.wpa.wtb.tue.nl (Albert Hofkamp) writes:

> On Wed, 13 Dec 2000 20:22:02 GMT, jschmitt at vmlabs.com <jschmitt at vmlabs.com> wrote:
> >From my experience, GNU make has the most features that I want to use.
> >However, I find it so hard to use under Windows.  For example, I wrote
> 
> No kidding :-)
> 
> >a small python program that I can launch from the command line and
> >execute without problems.  I can't seem to launch it from GNU make.  I

It might be easier to diagnose the problem if you would post the rule
and any relevant variable definitions from the Makefile.

> We had the same kind of problems, ever since we switched from Novell
> to NT servers.
> Until the arrival of the latter servers, we had a number of command-line
> utils that worked fine within the limits of 'fine' as we know it on W*.
> 
> When the Novell server was replaced by an NT server, long filenames
> suddenly stopped working, both on the network and locally.
> Also, it appeared that 1 command-line program (e.g. GNU-make), could not
> start another program (e.g. a compiler), without giving an error of any
> kind.

Which version of GNU make?  The Cygwin version works as expected here,
as long as it's set to Unix mode (set MAKE_MODE=unix).  I never tried
to work out how the native mode is supposed to work; Makefiles for
typical Unix software don't work at all in native mode.

> The core of the problem was tracked to the NT server that does not handle
> long filenames (even in the 21st century), although they are listed.

I think some other tool is at fault.  If you use the command
interpreter to run any DOS or Win16 programs, it switches into a
compatibility mode where it displays short names to avoid breaking
those programs.  I don't think it's possible to switch back later.

> I don't know about W2K, but given the fact that it is an NT-based version,
> I would not be surprised to see that the situation did not improve.
> In fact, I consider command-line oriented programs DEAD on W* systems.

Meanwhile, I'm writing and running command-line programs on Windows
2000 in C, C++, Python and Perl.

> >What could the reason be and how could I fix it?  Does it have
> 
> I think the reason is W2K. The only fix I know is complain to B.G.

I think you're falling into Microsoft-bashing mode.

> If that would fail for some reason, there are only 2 options left imho.
> 1) Switch to visual programming (aka do it the MS way),
> 2) Drop the platform.
> 
> FYI: We did the 2nd option.
> 
> >something to do with GNU make going out of it's way to launch sh.exe if
> >it's on your hard drive, rather than launching command.exe to shell out

Setting the SHELL to point to cmd.exe might fix this.

> Nope, it is simply not possible to launch a command-line program from
> another command-line program.

On what evidence do you base this assertion?

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From syver at NOSPAMcyberwatcher.com  Fri Dec 15 10:54:56 2000
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Fri, 15 Dec 2000 16:54:56 +0100
Subject: Python advocacy
References: <91bln3$lmr$1@nnrp1.deja.com> <91d63k0msl@news1.newsguy.com> <8iak3t4a3lrqcqp9iqftmekdoahuico2b0@4ax.com>
Message-ID: <91df1c$j85$1@troll.powertech.no>

"Steve Horne" <sh at ttsoftware.co.uk> wrote in message
news:8iak3t4a3lrqcqp9iqftmekdoahuico2b0 at 4ax.com...
> <SNIP>
>
> >Makes me wonder why Larry Wall didn't think to include at least
> >ONE strangeness, wart, or inconsistency *somewhere* in Perl...
> >thus making it so boringly perfect and un-idiosyncratic as to
> >leave its advocates without a suitable 'scapegoat feature'!
>
> Erm - either you haven't used Perl, or your an advocate yourself!
> Personally, if I was selling Python, Perl would be involved in my
> scapegoat manouvre.
>
> In fact, one such wart was mentioned in that article above (a length
> function that gives the 'wrong' result due to a scalar context).

Steve, I think Alex was joking :-)




From dyo at shinbiro.com  Tue Dec 26 03:54:25 2000
From: dyo at shinbiro.com (dyo)
Date: Tue, 26 Dec 2000 17:54:25 +0900
Subject: Duplicated List...
Message-ID: <929lq7$ds1$1@hiline.shinbiro.com>

Hi everyone.

>>> a=[1,2,3]
>>> b=[a,10,20,30]
>>> b[0].append('x')
>>> b
[[1, 2, 3, 1, 2, 3, 'x'], 10, 20, 30]
>>> a
[1, 2, 3, 1, 2, 3, 'x']
>>> a.append('xxx')
>>> a
[1, 2, 3, 1, 2, 3, 'x', 'xxx']
>>> b
[[1, 2, 3, 1, 2, 3, 'x', 'xxx'], 10, 20, 30]

Duplicated List work above..


But in this case..

>>> a=[1,1000]
>>> a
[1, 1000]
>>> b
[[1, 2, 3, 1, 2, 3, 'x', 'xxx'], 10, 20, 30]
>>>

Why do not change List 'b' ?
What happen List 'b' ?





From aleaxit at yahoo.com  Mon Dec 18 10:33:23 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 18 Dec 2000 16:33:23 +0100
Subject: search for pattern in StringList
References: <3a3e1935.5461298@news.skynet.be>
Message-ID: <91lap002od4@news1.newsguy.com>

"Daniel" <Daniel.Kinnaer at AdValvas.be> wrote in message
news:3a3e1935.5461298 at news.skynet.be...
> As a Python newbie, I was trying a very simple exercise, but got
> stuck. Is there someone who can help me complete this code, please?
>
> Thanks.
> Daniel
>
>
> MyList = ['an','been','call','door']
>
> print "Scanning a StringList"
>
> for i in range(0, len(MyList)):
>     print i, MyList[i]

Note that
    for i in range(len(MyList)):
is equivalent (0 is the default starting value for range).

> #try to show every item in MyList which contains an 'a'
> #that is 'an' and 'call'.   In Pascal :
> #for i:=0 to MyList.Items.Count -1 do
> #    if pos('a',MyList.Items[i])>0 then writeln(MyList.Items[i])
> #in Python : ???

In Python, it can be a little bit more "conceptually concise",
thanks to the for-in statement (one can iterate on the
items of a sequence directly, rather than iterating on
the indices and using indexing in the body, when the
index itself isn't needed):

for item in MyList:
    if item.find('a')>0:
        print item


It's not *necessary* to break the 'if', since its body
is a single, simple statement -- you *could* also write:

for item in MyList:
    if item.find('a')>=0: print item

but the normal Python convention is to end the line at
any ':' (which concludes such statements as for, while,
if, def, ... -- 'compound statements', basically), and
then (this is not just convention, but a key Python
lexical rule) you have to indent the statement's body
uniformly.


Alex





From mwh21 at cam.ac.uk  Thu Dec 14 10:06:29 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 14 Dec 2000 15:06:29 +0000
Subject: strip weirds
References: <QI2l1FAXBMO6EwD6@jessikat.demon.co.uk>
Message-ID: <m3y9xjoxe2.fsf@atrus.jesus.cam.ac.uk>

Robin Becker <robin at jessikat.fsnet.co.uk> writes:

> Can somebody explain why strip behaves differently with and without an
> argument?

Because:

> with the default whitespace arg [...] a run of whitespace is
> [...] treated as a single character.

This is usually what I want, at least, eg:

>>> 'a bunch of  words.  and stuff'.split()
['a', 'bunch', 'of', 'words.', 'and', 'stuff']

This has always been true as far as I know.

Cheers,
M.

-- 
  In short, just business as usual in the wacky world of floating
  point <wink>.                        -- Tim Peters, comp.lang.python


From root at rainerdeyke.com  Sat Dec 30 12:17:22 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Sat, 30 Dec 2000 17:17:22 GMT
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <3A4D6A05.EEC97FA6@uswest.net> <slrn94qqvg.mgg.grey@teleute.rpglink.com> <ije36.96813$x6.43604873@news2.rdc2.tx.home.com> <slrn94qt4f.mgg.grey@teleute.rpglink.com>
Message-ID: <CMo36.97857$x6.44474719@news2.rdc2.tx.home.com>

"Steve Lamb" <grey at despair.rpglink.com> wrote in message
news:slrn94qt4f.mgg.grey at teleute.rpglink.com...
> On Sat, 30 Dec 2000 05:23:26 GMT, Rainer Deyke <root at rainerdeyke.com>
wrote:
> >At that point it's better to use 'while 1', which is at least idiomatic
and
> >explicit in its inelegance.
>
>     I'm sorry, but I don't agree.  Even though I have no problem with
while 1:
> initing variables before the loop is more preferable to a break.  Esp.
when
> one can put the condition in the while; unlike reading lines from a file
where
> one cannot.

The example was something like this:

T = 0
while T < 25:
  T = ...

to simulate somethjing like this:

do:
  T = ...
until t >= 25

In other words, you need to artifically manipulate the 'while' condition to
always be true the first time through.  If you change the 'while' condition,
you may have to change the manipulation.  In other words, there's an
additional way the code could break.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From thomas at xs4all.net  Sat Dec 23 15:21:59 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Sat, 23 Dec 2000 21:21:59 +0100
Subject: Some Python 2.1 ideas
In-Reply-To: <005f01c06d01$d8114820$74eb0b18@C322162A>; from bobalex@home.com on Sat, Dec 23, 2000 at 09:00:25AM -0800
References: <005f01c06d01$d8114820$74eb0b18@C322162A>
Message-ID: <20001223212159.H5515@xs4all.nl>

On Sat, Dec 23, 2000 at 09:00:25AM -0800, Bob Alexander wrote:

> ----
> New operator: embedded assignment

Not very likely. It would mean changing assignment from a statement to an
operator (you can do that without actually introducing a new operator, by
the way, if it wasn't for the =/== confusion issue.) I looked at something
slightly similar while doing the augmented-assignment work (a
'content-filling' statement, ':='. The main difference with '=' being that
it would allow container objects to provide their implementation of '=', and
it wouldn't actually re-bind the name.) I didn't really like it myself,
though, and it was out of that particular scope, so I didn't do anything
with it.

However, I doubt an *operator* doing assignment would be accepted by Guido.
It goes against several of Python's implicit and explicit guidelines (one of
them being 'Guido should like it' :) Of course, stranger things have
happened. As the purpose of PEPs is to close discussion on issues such as
these once and for all, one way or another, it's not a bad idea at all to
PEP this.

> ----
> Variations on "for"
> Here are some possible additional forms of "for" that seem appealing:

> Allow slice-like syntax as shorthand for xrange():
>     for i in 10:                # equivalent to "for i in xrange(10)"
>     for i in 5:10:             # equivalent to "for i in xrange(5, 10)"
>     for i in 20:10:-1:       # equivalent to "for i in xrange(20, 10, -1)"

The 'range-literal' proposal (also one of mine ;) was rejected after being
on the "check this in when you're ready" list for a while. In fact, if Guido
had had those second thoughts a couple of days later, they might have made
it anyway :-) In any case, one of the reasons it was rejected was because
the slice-like syntax is just a bit *too* confusing. It looks a lot like
slicing, but it isn't, really, due to small details, and that makes it a bad
idea. (Strange as it may seem, that is actually my opinion too, and I don't
regret one bit that the range-literal proposal was rejected.)

Funnily enough, I also implemented the 'indexing-for' almost-PEP (which was
written by Just van Rossum, but not accepted as a PEP) which was rejected
because of uhm, one reason or another. Mainly the Guido-line above. I have
since considered proposing 

 for i in <x> .. <y>:

(or perhaps three dots instead of two) as an alternative to the
'indexing-for' syntax, but I wanted to catch Guido in a sufficiently good
mood to try it. And perhaps I should first have a public fight with Just,
just so show that I'm not conspiring with him :-)

Guido may disagree, but I still believe the idea behind PEPs is to stop
discussion of closed issues (or waiting-for-implementation ones) and you
can't do that if you only PEP subjects that have a chance or two in hell to
get accepted.

> ----
> String method to remove trailing newline

>     <string>.chomp()    # sorry, borrowed the method name from Perl

This suggestion I kind of like. Not the method name, though, since it
implies the string itself is modified (which it can't, as you should all
know ;) I don't know howmuch of a chance this method has, but it *is* an
often requested function. This is just the first time I've seen the request
phrased as a string method rather than a string module function :) I
personally don't have Python code that needs something that strips newlines
but not trailing whitespace, but there is probably code that does.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From idot at vt.edu  Sat Dec  9 22:41:44 2000
From: idot at vt.edu (Carl Banks)
Date: 10 Dec 2000 03:41:44 GMT
Subject: sgmllib nit
References: <eyErVLAEHlM6EwA4@jessikat.fsnet.co.uk> <EAsY5.2808$O%.435892@newsc.telia.net> <JagDRIAEbuM6EwxY@jessikat.fsnet.co.uk>
Message-ID: <90uu1o$t7i$1@solaris.cc.vt.edu>

Robin Becker <robin at jessikat.fsnet.co.uk>,
exactly what drugs were you under the influence of when you wrote:
>
> I realise that it does that! Unfortunately Zope's dtml isn't HTML.
> 
> Which parser should one use to analyse dtml which can contain tags like
> <dtml-var standard_html_header>?
> 
> The above analysis won't work easily if I have <tag X> as then I'm
> presented with attribute 'x' having value 'X'.
> 
> I merely want to copy the input to the output whilst making some trivial
> changes to the URL's in <A href=URL> style tags. I guess I need to write
> my own parser that doesn't make unnecessary assumptions.


I've noted this problem myself, when I was using sgmllib for an HTML
preprocessor for my web site.

Evidently, what you're "supposed" to do is keep track of which
attributes are boolean, and which take values.  While this may be
appropriate for an HTML renderer or syntax checker, it seems like a
lot to ask of someone writing a simple filter.  Even though it's not
sgmllib's advertised job, I think setting the attribute's "value" to
None when no value was given would be nice.


Yet, upon further reflection, I realized that I would almost never use
an attribute equaling its value.  Even if I did want to do that, I had
a good workaround.  So I just took attributes equaling their values to
mean that no value was given.


-- 
CARL BANKS                   http://filebox.vt.edu/users/cabanks1/

"I like nothing better than a good challenge.
 This is not one of them."


From fredrik at effbot.org  Fri Dec 15 12:43:40 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 15 Dec 2000 17:43:40 GMT
Subject: Python advocacy
References: <91bln3$lmr$1@nnrp1.deja.com> <3A39AC3E.FA75C4FF@engcorp.com> <nl5k3t03on9ehf9d50nov3coi71bu8tgo4@4ax.com>
Message-ID: <gLs_5.36$u17.4121@newsb.telia.net>

Steve Horne wrote:
> List comprehensions are copied directly from functional languages
> such as Haskell, which in turn copied them from mathematical set
> notation.

You forgot SETL...

</F>




From phrxy at csv.warwick.ac.uk  Mon Dec 11 08:00:34 2000
From: phrxy at csv.warwick.ac.uk (John J. Lee)
Date: Mon, 11 Dec 2000 13:00:34 +0000
Subject: C API mismatch with 2.0
In-Reply-To: <7qTY5.3054$O%.484335@newsc.telia.net>
References: <Pine.SOL.4.30.0012102209040.29674-100000@mimosa.csv.warwick.ac.uk> <7qTY5.3054$O%.484335@newsc.telia.net>
Message-ID: <Pine.SOL.4.30.0012111210490.29266-100000@mimosa.csv.warwick.ac.uk>

On Sun, 10 Dec 2000, Fredrik Lundh wrote:

> John J. Lee wrote:
> > I compiled Michael Miller's _tableio.c (the C extension bit of TableIO.py)
> > on windows using cygwin (with Robert Kern's help), and got the following
> > warning about the C API.  I used the python20.lib from the same
> > distribution of python 2.0 that I'm trying to import TableIO into.
> >
> > Can somebody tell me what's going on?  Where did this API version number
> > come from for _tableio?
>
> API version 1007 is 1.5.2, version 1009 is 2.0.
>
> looks like you failed to recompile (or install) the module.  to
> see which file you're loading, use:
>
> >>> import _tableio
> >>> print _tableio.__file__

Yes, I'm certainly loading the right file.  I've never compiled it (or
installed a binary of it) before, so I certainly compiled it.  The module
doesn't seem to be working properly - I've used it on unix before with, if
not the exact same file, at least one of the same format.

I presume the problem is that I compiled it with Paul Sokolovsky's patched
1.5.2 headers.  Is this still necessary to compile with gcc on windows,
and if so, are there patched 2.0 headers around?  Paul doesn't have any on
his page.


John



From howard at eegsoftware.com  Wed Dec 20 14:13:40 2000
From: howard at eegsoftware.com (howard at eegsoftware.com)
Date: Wed, 20 Dec 2000 19:13:40 GMT
Subject: object deletion order question
Message-ID: <3a41016d.3397213@news-server.socal.rr.com>

While locating a crash in a Windows Python extension I am working on
    ( a Python wrap around a DirectX game library called CDX)
I discovered my problem as being the order in which Python objects
were being deleted.  

In this case, I had created a Sound object and a series of SoundBuffer
objects.  INSIDE DirectX, deleting the Sound object implicitly
releases the SoundBuffer objects.  When Python terminated (or at other
deletions), it called the delete method of the SoundBuffer  objects
after deleting the Sound objects, leaving DirectX calling into unknown
data space (I know..it's a Microsoft tradition).

I patched around this problem by managing a list of buffer objects
attached to a Sound object (explicitly specifying such objects) and
deleting them explicitly.

However, that led my to question:
        What ordering does Python use for object deletions? 

I somehow *assumed* that deletion was the reverse order of creation
(in a global sense) but clearly my problem was a case of
first-created, first-deleted.  Of course, I realize (now) I should not
have *assumed* ANY deletion order.  

For a "good" fix, I will probably have to build some linkage between
the two 'independent' classes in the C++ code rather than rely on
working around the problem in Python code.
Howard Lightstone
EEGSoftware
howard at eegsoftware.com


From mjjacob at my-deja.com  Wed Dec 13 11:39:44 2000
From: mjjacob at my-deja.com (mjjacob at my-deja.com)
Date: Wed, 13 Dec 2000 16:39:44 GMT
Subject: Problem with += (augmented assignment)
Message-ID: <9188of$q55$1@nnrp1.deja.com>

Hi folks!

I tried this under Python 2.0 running on Win 2k:

x = 0
while x <= 42:
	x += 1

Python always complains a Syntax Error: Invalid
Syntax on this:

    x += 1
       ^
SyntaxError: invalid syntax

Augmented assignment only seems to work when used
in first indentation level (script level).
Otherwise, I always get that syntax error, it
does not matter whether I am using if, while or
whatever.

Is there something terribly wrong with the +=
operator or is that all my fault?

Help appreciated.


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


From aleaxit at yahoo.com  Wed Dec 27 10:38:10 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 27 Dec 2000 16:38:10 +0100
Subject: Controlling the Parallel Port
References: <mvuj4tkmn22gbth1a6t8fag39902qeeu9l@4ax.com> <mailman.977928790.25688.python-list@python.org> <pe1k4tsr7u017n45gf09f20fc8k1int2ci@4ax.com>
Message-ID: <92d2do0138o@news2.newsguy.com>

"Simon Faulkner" <News at Titanic.co.uk> wrote in message
news:pe1k4tsr7u017n45gf09f20fc8k1int2ci at 4ax.com...
> Cheers Bill,
>
>
> I have found the link and, indeed, several other c links.
>
> So, pardon my Python ignorance, but what is SWIG?

See www.swig.org:

"""
SWIG is a software development tool that connects programs written in C,
C++, and Objective-C with a variety of high-level programming languages.
SWIG is primarily used with common scripting languages such as Perl, Python,
Tcl/Tk, Ruby, and Guile but has also been extended to include languages such
as Java and Eiffel. SWIG is most commonly used to create high-level
interpreted programming environments and as a tool for building user
interfaces. SWIG may be freely used, distributed, and modified for
commercial and noncommercial use
"""


Alex





From parish at ikb.mavt.ethz.ch  Sun Dec 31 09:31:08 2000
From: parish at ikb.mavt.ethz.ch (Yogi Parish)
Date: Sun, 31 Dec 2000 09:31:08 -0500
Subject: Problems with the re module on IRIX
Message-ID: <3A4F432C.386EFFE3@ikb.mavt.ethz.ch>

Hello, everybody

I cannot get the re module to run on an SGI. (1.5.2) Has anybody had
similar experiences
with such a problem (make test fails in test_re)....

what can I do?

thanx
yogi




From leonrom at hypermath.org  Wed Dec  6 02:13:58 2000
From: leonrom at hypermath.org (leonrom)
Date: Tue, 05 Dec 2000 23:13:58 -0800
Subject: Beginers question
References: <_3NW5.44729$nh5.3218436@newsread1.prod.itd.earthlink.net> <3A2DD311.25AF9EE4@pehr.net>
Message-ID: <3A2DE736.80BFBF54@hypermath.org>

Look here too:
http://www.csulb.edu/~murdock/dosindex.html

pehr anderson wrote:
> 
> I think this isn't a big problem.
> You should probably get somebody to
> walk you through how to use the
> command line.
> 
> It is hard to explain all the
> concepts involved in navigating
> files and directories without
> being there with you an introducing
> the commands.
> 
> If you can, find a reference on the
> DOS/Windows command line commands.
> These haven't changed much in 15 years.
> 
>         -pehr
> 
> robert george wrote:
> >
> > I started the beginers tutorial. My operating system is Winme.
> > The tutorial says I should get a" C:\WINDOWS> " prompt. But I get
> > a"C:\WINDOWS\DESKTOP>" prompt.
> > Why is that? Will Python work with Winme?
> > Thanks for your help
> > An extreme novice.
> > Rob


From blosh at tsbrass.com  Fri Dec 15 15:02:34 2000
From: blosh at tsbrass.com (Brad)
Date: Fri, 15 Dec 2000 20:02:34 GMT
Subject: smtp and Groupwise
Message-ID: <91dtck$fgo$1@nnrp1.deja.com>

I am having a problem sending an email using smtplib.  we have an HPUX 10.20
box and a novell server with groupwise for our pobox.  from unix, i am trying
to send an email using by smtp server on the novell server which works if it
is in house, not if it out house.  what am i overlooking?  tried the same
thing from windows, same result....

any reasons?


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


From teyc at bigfoot.com  Sun Dec 24 16:01:06 2000
From: teyc at bigfoot.com (Chui Tey)
Date: Mon, 25 Dec 2000 07:01:06 +1000
Subject: xml-rpc server in python returning python lists...
References: <hih16.85504$yR4.2492277@news1.rdc1.tx.home.com>
Message-ID: <925o9s$pqk$1@bunyip.cc.uq.edu.au>

As far as I know, VB takes it as an array of Variants. Worked for me before.
Which XMLRPC Client are you using?

Scott Hathaway <slhath at home.com> wrote in message
news:hih16.85504$yR4.2492277 at news1.rdc1.tx.home.com...
> Hello All,
>
> I am writing an xml-rpc server in python that has a function which returns
a
> python list.  When I use the python client to connect to the python
server,
> it works great.  When I try from VB, ASP, PHP, and Delphi, they all have
> trouble with the return type of the variable.  By the way, all the clients
> mentioned work fine if the server's method returns a string or a number.
>
> Any ideas?
>
> Scott
>
>




From ludwig_callewaert at frontierd.com  Fri Dec  8 10:24:45 2000
From: ludwig_callewaert at frontierd.com (Ludwig Callewaert)
Date: Fri, 08 Dec 2000 16:24:45 +0100
Subject: Q for regexp experts
References: <yv2aea77xw2.fsf@lionsp093.lion-ag.de>
Message-ID: <3A30FD3D.E3DA4639@frontierd.com>

Harald Kirsch wrote:
> 
> Even with perl-compatible regular expressions it seems to be
> impossible to specify a pattern which matches
> 
>    `the longest substring not containing separator' SEP
> 
> Put another way, from a given starting point, the pattern shall match
> everything until just before the next occurence of SEP. A first
> attempt is something like
> 
>    [^SEP]*
> 
> but this disallows S', 'E' and 'P' in between the separators.
> 
> I tried something like (.(?!SEP))+ but it somehow does not work. Is it
> at all possible to get this to work with regular expressions?
> 
>   Harald Kirsch
> 
> --
> ----------------+------------------------------------------------------
> Harald Kirsch   | kirschh at lionbioscience.com | "How old is the epsilon?"
> LION Bioscience | +49 6221 4038 172          |        -- Paul Erd?s

What about:

set s "the longest matching String not containing SEP"

regexp -- ".*(?=SEP)" $s match

puts $match

> the longest matching String not containing 

Ludwig

-- 

Ludwig Callewaert       e-mail: ludwig_callewaert at frontierd.com
Senior Software Engineer        Frontier Design, Belgium
__________________________________________________________________


From darrell at dorb.com  Wed Dec 20 20:35:47 2000
From: darrell at dorb.com (Darrell Gallion)
Date: Wed, 20 Dec 2000 20:35:47 -0500
Subject: re different between US and German versions of NT
Message-ID: <04ba01c06aee$589f27c0$6401a8c0@home>

Found a re pattern that behaved differently on NT4 (US version) than it did
on NT4 (German version)
When I switched to use pre instead of sre it worked the same on both.
I assume this is a Unicode thing ?
Is this to be expected ?


Sorry, left the example at work. Also don't have access to the German
machine any longer.

--Darrell











From peter at schneider-kamp.de  Thu Dec 28 08:43:48 2000
From: peter at schneider-kamp.de (Peter Schneider-Kamp)
Date: Thu, 28 Dec 2000 14:43:48 +0100
Subject: Is this a BUG?
References: <3A4B0C5E.1E7976D2@wielun.dhs.org>
Message-ID: <3A4B4394.83A6B8B@schneider-kamp.de>

QdlatY schrieb:
> 
> print "!",form["index"].value,"!"
> ! 1 !
> Why the hell there are spaces? index property has value '1'!
> not ' 1 '?!

They are put there by the print statement. To learn more about
print see the Language Reference at
  http://www.python.org/doc/current/ref/print.html

Hope that helps,
Peter



From qrczak at knm.org.pl  Wed Dec  6 16:18:55 2000
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 6 Dec 2000 21:18:55 GMT
Subject: Calculating factorial
References: <3a2a9496.647995@uutiset.nic.fi> <3A2BCC1F.F308E511@javanet.com> <ya0vgsx3h2f.fsf@levding.ifi.uio.no> <zytX5.54052$15.11439984@news1.rdc1.az.home.com>
Message-ID: <slrn92tb9v.o4q.qrczak@qrnik.zagroda>

Wed, 06 Dec 2000 16:11:11 GMT, Tim Hochberg <tim.hochberg at ieee.org> pisze:

> def factorial(n):
>     result = 1L
>     while n:
>         result *= n
>         n -= 1
>     return result
> 
> It's not any faster than the other choices, but it has the dubious,
> theoretical distinction of using less memory than all but the xrange version

OTOH counting upwards lets it work on smaller numbers on average.

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


From ldw at us.ibm.com  Tue Dec  5 14:40:33 2000
From: ldw at us.ibm.com (Larry Whitley)
Date: Tue, 5 Dec 2000 13:40:33 -0600
Subject: Graphs from python
References: <90gse2$14uo$1@news.rchland.ibm.com> <HOTW5.4117$xW4.39082@news-server.bigpond.net.au>
Message-ID: <90jgbg$17bi$1@news.rchland.ibm.com>

I hadn't thought of that.  I'll give it a shot.

Thanks,
Larry

"Neil Hodgson" <neilh at scintilla.org> wrote in message
news:HOTW5.4117$xW4.39082 at news-server.bigpond.net.au...
> Larry Whitley:
>
> > I'm looking at using win32com (by Mark Hammond) to drive excel or
> > piddle+graphite to generage graphs directly from python.  While I'm sure
> it
> > will get the job done, the excel object model is a bit overwhelming.
>
>    The easy way to create Excel code and learn the object model is to turn
> on the macro recorder and perform by hand the steps you need to create the
> graphs. Then look at the generated code and generalise it to suite your
> needs.
>
>    Neil
>
>
>




From kahuna01 at hotmail.com  Fri Dec  8 07:02:37 2000
From: kahuna01 at hotmail.com (Jason Stewart)
Date: Fri, 08 Dec 2000 12:02:37 GMT
Subject: My first attempt at python
Message-ID: <n2j13tcc0il8k1i3dtjkdckglo871ln744@4ax.com>

Greetings,
	This is the first piece of code that I wrote in Python. So far
I think that Python is wonderful. Being inexperienced as I am in
python, I would like your help in analyzing my code and telling me
where I can improve. 
	This same functionality might be in a module somewhere, but I
wrote it for practice. Its a string tokenizer. If you could analyze it
and tell me where I could improve I would be grateful.

Thanks,
Jason

def tokenize(string_to_chop):
	"Break a string up into individual words"
	last_space = 0
	token = []
	for index in range(len(string_to_chop)):
		if string_to_chop[index:index+1] == " ":
			token.append(string_to_chop[last_space:index])
			last_space =  index + 1
		elif (index+1 == len(string_to_chop)):
		    token.append(string_to_chop[last_space:index+1])
	return token



From m.faassen at vet.uu.nl  Sat Dec 30 21:23:23 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 31 Dec 2000 02:23:23 GMT
Subject: Python vs. PHP (& Java?)
References: <mailman.977897470.4734.python-list@python.org>
Message-ID: <92m5ar$cfm$2@newshost.accu.uu.nl>

Ben Ocean <beno at thewebsons.com> wrote:

[snip]
> Python has the advantage of byte-code and
> compilation, but it?s server-side. PHP might not be as fast on the server,
> but it doesn?t have to make as many trips back and forth between server and
> client.

It appears as if you're misunderstanding something here. Firstly, 
I don't know which of the languages is faster; it probably depends on the
task as much on the language itself. I don't know much about PHP, but I thought
it could be compiled to bytecode or at least some intermediate format as well?

PHP and Python when used for website programming are both server side only.
There are no 'trips back and forth' between server and client more than
the regular HTTP requests coming from the client and the resultant web
pages being delivered from the server. There is no difference between Python
and PHP in this respect at all.

Perhaps you got confused by DTML in Zope? DTML allows you to embed calls
to Python methods (and callable Python objects in general) in a web page,
for instance like this:
  
 <p><dtml-var "myfunction()"></p>

Where myfunction is some Python function (defined as an external method,
for instance). But the client never calls myfunction() in this example;
the *server* does the calling before the resultant web page is sent out
to the client.

Anyway, I saw some other misunderstandings in your post as well, but 
from my brief reading of some answers I didn't see anyone go into this
one yet. Good luck and have fun with Zope and Python!

Regards,

Martijn  
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From rjroy at NO-SPAM.magma.ca  Wed Dec 27 11:54:05 2000
From: rjroy at NO-SPAM.magma.ca (Robert Roy)
Date: Wed, 27 Dec 2000 16:54:05 GMT
Subject: Exception Handling and Inheritance
References: <921sdr$mlq$1@newsreaderg1.core.theplanet.net>
Message-ID: <806k4tc5t9js74aqv81u214ijiqsni9q2j@4ax.com>

On Sat, 23 Dec 2000 10:45:57 +0100, "Franz GEIGER" <fgeiger at datec.at>
wrote:

>
>I'd like to specialize Exceptions. So I derived a TException from Exception
>and added all the things I think I need to have.
>
>But now there is a problem: I have created a second branch in the
>inheritance tree:
>Exception
>    StandardError
>        IOError
>        ...
>    TException
>        TDataException
>        ...
>
>When I want to catch TExceptions I write
>try:
>    obj.meth(anyStuff)
>except TException, e:
>    print >> sys.stderr, "Panic: %s!" % e.Formatted()
>
>But what about Exceptions? They are not caught by my TException handler! And
>Exception does not have my Formatted method.
>

You could always add __str__ = Formatted to your class definintion
since the published interface to Exception supports the str()
function.

Then  you could write 
try:
    obj.meth(anyStuff)
except Exception, e:
    print >> sys.stderr, "Panic:", e

Since TException is derived from Exception this would work.

>Do I really have to catch them in a 2nd handler? Or can I add some
>functionality somehow to Exception instead of implementing it in a derived
>TException?

As Aahz mentionned, it is usually considered a Bad Idea to use
Exception as a catch all since it tends to mask programming errors. Of
course you could always Exception then do:
if e.__class__ == AException:
	handle a
elif e.__class__ == TException:
	handle t
etc...

but this does not really improve your lot in life over multiple
handlers

>
>Many thanks in advance and best regards
>Franz GEIGER
>
>

Bob


From gregj at pobox.com  Sat Dec 30 01:15:50 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Sat, 30 Dec 2000 06:15:50 GMT
Subject: Question re: eval()
References: <978109654.339714327@news.silcom.com> <bB636.2482$Qb7.271736@newsb.telia.net>
Message-ID: <q4f36.204086$U46.6288617@news1.sttls1.wa.home.com>

"Fredrik Lundh" <fredrik at effbot.org> wrote in message
news:bB636.2482$Qb7.271736 at newsb.telia.net...

[cool trick using compile and eval]

Thanks! I started working on simple expression evaluation for a mail merge
program, and I need to refer to database column names in the expressions.
This solves 90% of my problem.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From cjensen at bioeng.ucsd.edu  Wed Dec 27 14:39:17 2000
From: cjensen at bioeng.ucsd.edu (Curtis Jensen)
Date: Wed, 27 Dec 2000 11:39:17 -0800
Subject: Python core dumping
References: <92deoj$b2q$1@bob.news.rcn.net>
Message-ID: <3A4A4565.7322DA7B@bioeng.ucsd.edu>

David Allen wrote:
> 
> I have a medium sized python application which is causing
> python to segfault under Linux at very odd times.  I'd
> like to get a core dump, so I can try to figure out
> what on earth is going on with it, and possibly post
> a bug report or something similar.  But as odd as this
> sounds, the application will NOT dump core.  The
> program I'm running does not use any C extensions,
> it's written in pure python.
> 
> I've tried using:
> Python 1.5.2 (#0, Apr  3 2000, 14:46:48)  [GCC 2.95.2 20000313 (Debian GNU/Linux)] on linux2
> and
> 
> Python 2.0 (#1, Dec 17 2000, 18:36:14)
> [GCC 2.95.2 20000220 (Debian GNU/Linux)] on linux2
> 
> Both after issuing "ulimit -c unlimited" which gives
> this:
> 
> [x at test pscr]$ ulimit -a
> core file size (blocks)     unlimited
> data seg size (kbytes)      unlimited
> file size (blocks)          unlimited
> max locked memory (kbytes)  unlimited
> max memory size (kbytes)    unlimited
> open files                  1024
> pipe size (512 bytes)       8
> stack size (kbytes)         8192
> cpu time (seconds)          unlimited
> max user processes          256
> virtual memory (kbytes)     unlimited
> 
> So the lack of core dumping is not due to my
> environment as far as I can tell.
> 
> I did not compile 1.5.2, that's the default version
> that's in Debian GNU/Linux (Woody) right now.  As
> for 2.0, I only "./configure"'d it with a --prefix
> option.
> 
> How can I get python to dump core so I can figure out
> what's wrong with my program, or alternatively, what's
> wrong with python?  Any help would be appreciated.
> 
> --
> David Allen
> http://opop.nols.com/

Try using pdb (python's debuger)  or even something like gdb or ddd. 
pdb would be more useful if the problem is in the python script code. 
gdb would be more useful if the problem is in the C code.  The problem
is probably in the C code.

-- 
Curtis Jensen
cjensen at bioeng.ucsd.edu
http://www-bioeng.ucsd.edu/~cjensen/
FAX (425) 740-1451


From robin at alldunn.com  Sat Dec 16 02:57:29 2000
From: robin at alldunn.com (Robin Dunn)
Date: Fri, 15 Dec 2000 23:57:29 -0800
Subject: Compiling Python 2.0 with Berkeley DB 3.0.55?
References: <3A39DF89.D76B6415@schlund.de> <u4s_5.168$SKV4.36635003@news.randori.com> <3A3A54C8.E85CB507@schlund.de>
Message-ID: <YfF_5.525$SKV4.85000270@news.randori.com>

> > Can you create and use new files with dbhash.open()?
>
> Yes. But if I run db_dump on these files, it says "unexpected file type
> or format", regardless which db_dump version I use (2.0.77, 3.0.55,
> 3.1.17)
>

It may be that db_dump isn't compatible with version 1.85 databse files.  I
can't remember.  I seem to recall that there was an option to build 1.85
versions of db_dump and db_load.  Check the configure options for BerkeleyDB
to find out.  (Also, while you are there, make sure that BerkeleyDB was
built the same on both of your platforms...)


>
> >  Try running db_verify (one of the utilities built
> > when you compiled DB) on the file and see what it tells you.
>
> There is no db_verify among my Berkeley DB utilities.

There should have been a bunch of them built when you compiled DB.  I've got
these:

-r-xr-xr-x  1 rd       users     343108 Dec 11 12:11 db_archive
-r-xr-xr-x  1 rd       users     342580 Dec 11 12:11 db_checkpoint
-r-xr-xr-x  1 rd       users     342388 Dec 11 12:11 db_deadlock
-r-xr-xr-x  1 rd       users     342964 Dec 11 12:11 db_dump
-r-xr-xr-x  1 rd       users     349348 Dec 11 12:11 db_load
-r-xr-xr-x  1 rd       users     340372 Dec 11 12:11 db_printlog
-r-xr-xr-x  1 rd       users     341076 Dec 11 12:11 db_recover
-r-xr-xr-x  1 rd       users     353284 Dec 11 12:11 db_stat
-r-xr-xr-x  1 rd       users     340340 Dec 11 12:11 db_upgrade
-r-xr-xr-x  1 rd       users     340532 Dec 11 12:11 db_verify



--
Robin Dunn
Software Craftsman
robin at AllDunn.com
http://wxPython.org     Java give you jitters?
http://wxPROs.com        Relax with wxPython!





From arcege at shore.net  Wed Dec 27 14:18:52 2000
From: arcege at shore.net (arcege at shore.net)
Date: Wed, 27 Dec 2000 19:18:52 GMT
Subject: newbie : array initialization, anyone ?
References: <3a4a1352.9818548@news.t-online.de>
Message-ID: <wgr26.1691$JG.213419@news.shore.net>

Gerson Kurz <gerson.kurz at t-online.de> wrote:
: you can initialize a fixed-size array using something like

:   array = [0] * 10

: now, one would think that you can initialize a multidimensional array
: like this:

:   multidim = [ [0] * 10 ] * 3

: but now, all 3 elements in multidim are references to the same single
: array, so that if you type

:   multidim[0][0] = 1
:   print multidim 

: you actually get three 1 in all those 0s. Where did my thinking go
: wrong along these lines ?

This is in the FAQ (#4.50).  You will want to create each inner dimension
individually:

>>> multidim = [ None ] * 3
>>> for i in range(3):
...   multidim[i] = [ 0 ] * 10
...
>>>

  -Arcege



From tim.one at home.com  Wed Dec 20 14:11:55 2000
From: tim.one at home.com (Tim Peters)
Date: Wed, 20 Dec 2000 14:11:55 -0500
Subject: while true: !!!
In-Reply-To: <91qada$1dh$1@plutonium.compulink.co.uk>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEMMIEAA.tim.one@home.com>

[gbreed at cix.compulink.co.uk]
> ...
> If you want to empty a sequence, but preserve the type, try
> filter(lambda x:0, asequence).

Picking a nit:  that returns an empty list, regardless of type(asequence).
To get an empty sequence of the same type, do

    asequence[:0]

or even

    asequence * 0

there's-only-one-way-to-do-it-ly y'rs  - tim




From aleaxit at yahoo.com  Mon Dec  4 11:45:46 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 4 Dec 2000 17:45:46 +0100
Subject: Calculating factorial
References: <3a2a9496.647995@uutiset.nic.fi>
Message-ID: <90ghoh0lh@news2.newsguy.com>

"Olli Rajala" <ollir at linuxfreak.com> wrote in message
news:3a2a9496.647995 at uutiset.nic.fi...
> Hi!
> I've learned a few time Python and programming.  I wrote a function
> that calculates a factorial of N and I was wondering if it's the best
> (I think it's not. =) way to do that. I'm glad to hear your comments.
>
> <script>
>
> def factorial(f):
>     result = 1
>     for number in range(1, f + 1):
>         result = result * number
>     return result
>
> print factorial(10)

It's not too bad, the main limit being that it
will only work for a *VERY* small f; you can
remedy that by changing the very first statement:
    result = 1L
*Now* it will work for decently-sized f.

Attempts to get more sophisticated, such as

def factorial(f):
    import operator
    return reduce(operator.mul, range(1,f+1), 1L)

will not necessarily bring happiness.  Measuring
1000 calls for f==13, on my machine this appears
substantially slower -- about 0.85 milliseconds
per call (net) vs 0.72 for your simple approach
(just modified to do the computations with longs).


I _think_ the only way to accelerate this quite
substantially is...:

import gmpy

def factorial(f):
    return gmpy.fac(f)

which gives me about 0.09 milliseconds per call
in the same situation.  But I guess this is sort
of cheating, since gmpy is a module I'm just now
developing (http://gmpy.sourceforge.net/) and it
'just happens' to have .fac (factorial) as one of
its built-in functions (I do enough combinatorial
arithmetic that I really appreciate these kinds
of order-of-magnitude speedups:-).


Alex





From aleaxit at yahoo.com  Mon Dec  4 11:23:51 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 4 Dec 2000 17:23:51 +0100
Subject: Mutual import?
References: <r16n2tkqepahf288gmu6tku5d2sctb72o0@4ax.com>
Message-ID: <90ggf9030hi@news2.newsguy.com>

"Dale Strickland-Clark" <dale at out-think.NOSPAMco.uk> wrote in message
news:r16n2tkqepahf288gmu6tku5d2sctb72o0 at 4ax.com...
> Is it considered safe and acceptable for two modules to import each other
> or are there likely to be unwelcome side effects of that?

Even if you could avoid any Python-specific problems, mutual dependencies
between components are one of the worst banes of software development.  I
would go to VERY great lengths to refactor any system based on mutual
dependencies (and highly recommend Lakos' excellent "Large Scale C++
Development" for a thorough treatment of this, albeit in a C++ context).


> The problem is that both require access to values and routines in the
other.

Refactor the system so that they don't... there are many ways to do so.

For example, the "two" modules, if they depend on each other, are 'really'
one.  Merge them into a single module.  You may still provide separate
"views" of that module 'from the outside' ('client-code'), if you want,
via two auxiliary modules (which are the ones client-code sees) which just
'from mergedmodule import' the specific list of identifiers to be exported
as belonging to the specific auxiliary-module.

Note that the "modularity" you seem to lose here (with the merging)
was never really there in the first place, because two alleged 'modules'
that actually depend on each other are too intertwined and need to be
changed/released/tested as ONE unit (i.e., you only actually have one
module anyway) -- and you may preserve appearances with the auxiliary
modules (which are the ones client-code will be seeing).


For higher-sophistication solutions, consider Robert Martin's "dependency
inversion principle" (read all he has to say about OO at his site,
http://www.objectmentor.com/publications/articlesBySubject.html -- it's
well worth the time even if one has trouble following his C++ oriented
examples, IMHO).  The general idea, using -> to denote dependency:

    we currently have A->B and B->A
    we hate that for lots of excellent reasons
    specifically, we want to remove (say) B->A, because
        B is more stable than A thus should NOT depend
        on it (that's the bad half of the mutual dependency)
    (if x->y, y should be more stable than x).

Solution: introduce another (abstract) entity C that just exposes
the "abstract interface" of A (but no implementation).  Change
the dependencies to:
    A->B    unchanged, but that's OK
    B->C    B depends on C because it *uses* that interface
    A->C    A depends on C because it *implements* that interface

This eliminates the loops and leaves all right-hand sides of a
depends-on relation as stabler than the left-hand sides (since
just-abstract-interface components can be made pretty stable
much more easily than ones doing implementation).  [Note that
Lakos' key idea for eliminating mutual dependencies is also
based on abstract-interfaces -- what he calls "protocol classes"].

Classes are the natural way to implement this (in Python just
as well as in C++), but you can do it with modules if you wish.
For example, C may be just a list of the functions A will expose,
each in the form:

-- C.py

def foo(*args, **kwds):
    global foo
    import A
    foo = A.foo
    return A.foo(*args, **kwds)

def bar(*args, **kwds):
    global bar
    import A
    bar = A.bar
    return A.bar(*args, **kwds)

-- &c

Cookie-cutter code (needed just because there is no __getattr__
in a module -- which is why a class would be handier), quite
stable indeed (it would need to change only if and when the
abstract interface of A, i.e. the set of functions A exports,
should change; its stability is thus [basically] that of an
abstract interface, as desired).

Note that this will work even if B does 'from C import *', as
it shouldn't:-), but it will be pretty inefficient then as you
will be paying a double-indirectness on each call (while, if
B has 'import C' and calls explicitly 'C.bar(fee,fie)' etc,
the double-indirectness is only paid once, at the first call,
then C's entry for 'bar' becomes identical with A's).


Alex





From robin at jessikat.fsnet.co.uk  Sun Dec 10 06:13:18 2000
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sun, 10 Dec 2000 11:13:18 +0000
Subject: sgmllib nit
References: <eyErVLAEHlM6EwA4@jessikat.fsnet.co.uk>
 <EAsY5.2808$O%.435892@newsc.telia.net>
 <JagDRIAEbuM6EwxY@jessikat.fsnet.co.uk>
 <y_IY5.2910$O%.457391@newsc.telia.net>
Message-ID: <HiCA$IAOV2M6EwyT@jessikat.fsnet.co.uk>

In article <y_IY5.2910$O%.457391 at newsc.telia.net>, Fredrik Lundh
<fredrik at effbot.org> writes
  helpful stuff elided
>
>> I guess I need to write my own parser that doesn't make unnecessary
>> assumptions.
>
>well, if you think that assuming SGML is an "unnecessary assumption"
>for something called sgmllib, there's not much I can do to help.
>
no I meant just look for <a href=> ignoring the fact that it's SGML
except for commenting quotes etc etc. 

As for zope's dtml I think it's not really designed to be a single
language as it can contain constructs like

<a href=<dtml-var b>>

which I think this can only be parsed by considering the dtml and then
the resultant html.
-- 
Robin Becker


From sab at NOSPAM.freeuk.com  Thu Dec 21 14:19:19 2000
From: sab at NOSPAM.freeuk.com (G. Willoughby)
Date: Thu, 21 Dec 2000 19:19:19 -0000
Subject: 2 question from a python newbie!
Message-ID: <91tl43$rko$1@neptunium.btinternet.com>

1). Can python scripts be compiled to 'stand alone' .exe files?
2). What's the best book/s available for learning python from a total
beginers point of view?

Thanks in advance and i apologise if the faq covers these, i didn't see
them.

G. Willoughby.




From qdlaty at wielun.dhs.org  Mon Dec 25 13:17:09 2000
From: qdlaty at wielun.dhs.org (QdlatY)
Date: Mon, 25 Dec 2000 19:17:09 +0100
Subject: html filtering
References: <3A46900D.BD118A0A@wielun.dhs.org> <ku7l4pt7sw.fsf@lasipalatsi.fi>
Message-ID: <3A478F25.1A8CCC4E@wielun.dhs.org>

Hello!

> no. there have been bugs in some browsers making them interpret
> '<' and '>' with the 8th bit set as those characters also.
> there may be others. this was a popular topic on bugtraq at one
> point, i suggest you check the archives.

If all webmasters will check all borwsers on their pages, they will
never create good page.

btw. If someone wants to put html tags without permission, he's got
problem with post.

 
> | Maybe You've got better ideas?
> 
> no, but good luck :)


Thanks! My multi-channel message board will be ready soon. I'm going to
give it's source free.

Yours
QdlatY


From rzilavec at tcn.net  Wed Dec 13 13:05:21 2000
From: rzilavec at tcn.net (Richard Zilavec)
Date: Wed, 13 Dec 2000 18:05:21 GMT
Subject: Moving from Perl to Python
References: <3a3a9552.76673185@news.tcn.net> <slrn93fcdt.a8m.scarblac@pino.selwerd.nl>
Message-ID: <3a38b798.85448901@news.tcn.net>

On 13 Dec 2000 17:30:56 GMT, scarblac at pino.selwerd.nl (Remco Gerlich)
wrote:

>Personally I don't like Programming Python much. It's becoming old (there
>are rumors of a new version soon) and not very well structured. But then,
>it's a lot of book, there are many examples, and the important things are
>probably all in there somewhere.

I hope I enjoy it.  I don't think I will ever stop using Perl.. but I
really need something new.  

>I started with Learning Python, and I like that book. It explains the

Wish I could see this book first, I don't want anything too basic
maybe just a quick start, then more advance features.

>Python Essential Reference is a nice reference book. It has a short
>tutorial, then a good reference on the language and of the standard
>library. One of the best books out there at the moment.
>
>More books at http://www.python.org/psa/bookstore/
>
>There is good online documentation at http://www.python.org/doc/current/ .

Thanks!

>There are things out there on the web that explain Python to Perl programmers,
>but I can't find them at the moment. 

I just found a man pages on another server, should be a good start..

>
>There are some posts by Tom Christiansen to this group that you can find
>on Deja about Perl v Python, those may be useful.

I found this article at perl.com, it was quite good.

--
 Richard Zilavec
 rzilavec at tcn.net


From dberlin at redhat.com  Thu Dec  7 15:57:48 2000
From: dberlin at redhat.com (Daniel Berlin)
Date: 07 Dec 2000 15:57:48 -0500
Subject: Python legalities
In-Reply-To: <LNBBLJKPBEHFEDALKOLCIEOGICAA.tim.one@home.com>
References: <LNBBLJKPBEHFEDALKOLCIEOGICAA.tim.one@home.com>
Message-ID: <m3sno0eypf.fsf@dan2.cygnus.com>

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

> > Bruce Dawson
> > comments at cygnus-software.com
> 
> That's a bizarre mailing address, given the substance of your post <wink>.
> 

Not really, they aren't Cygnus (now part of RedHat).
They make a program called Fractal Extreme, apparently.
--Dan




From barry at digicool.com  Wed Dec 20 10:39:30 2000
From: barry at digicool.com (Barry A. Warsaw)
Date: Wed, 20 Dec 2000 10:39:30 -0500
Subject: IGNORE - this is only a test
Message-ID: <14912.53938.280864.596141@anthem.concentric.net>

Testing the new MX for python.org...



From MarkH at ActiveState.com  Sat Dec 30 19:25:09 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sun, 31 Dec 2000 00:25:09 GMT
Subject: Embedding Python in Windows Applications
References: <3a4e28bf.16852652@news.t-online.de>
Message-ID: <3A4E7BC4.8040101@ActiveState.com>

Gerson Kurz wrote:

> If I embedd Python in a Win32 console application (using
> Demo\embed.c), everything works fine. If I take the very same piece of
> code and put it in a Win32 Windows application (not MFC, just a plain
> WinMain()) I see no output (and more importantly so, no errors),
> because the application does not have a stdout/stderr set up. 
> 
> I have made the following "hack" to make it work (in VisualStudio)
> 
>     // stdout, stderr as defined in the MSVCRT
>     _iob[1] = *fopen("D:\\stdout.log","w");
>     _iob[2] = _iob[1];
> 
> but that surely cannot be a longterm solution. I don't want the output
> to be written to a temp file, I want to see it in a callback. Is that
> possible, and if so, how ? 

There is not much Python can do here, but it does provide tools to help you.

When Python initialized, "sys.stdout" and "sys.stderr" will be set to 
the C runtime stdout and stderr handles.

As you discovered, a GUI app doesnt have a valid stderr or stdout. 
Therefore, you may choose to set a valid C runtime stderr and stdout as 
your app initializes.  However, only you know what makes sense for you app.

Python itself always prints via "sys.stdout" and "sys.stderr".  Thus, if 
the first thing you do is to set sys.stdout and sys.stderr to some 
object of yours, again you have full control over where this output 
goes.  But again, only you know what makes sense.


IMO, you best solution is for your app to:
* Execute some Python "bootstrap" code.  This code should set sys.stderr 
etc to something useful.  If necessary, this Python implemented code 
could still callback into your application via your existing extension 
mechanisms.

* Once this bootstrap code has executed, you can then run your "normal" 
code in the knowledge that all errors etc will go somewhere useful (ie, 
to the object created in the bootstrap phase).

You probably would have special handling for the bootstrap phase - eg, 
maybe you _could_ still temporarily set stderr to "%TEMP%\stderr.log". 
If the bootstrap code fails, this log will tell you why.  If it 
succeeds, you could delete the log file, safe in the knowledge that 
subsequent errors will end up to your new sys.stdout/stderr objects.

Mark.



From jp at ulgo.koti.com.pl  Wed Dec 27 15:37:05 2000
From: jp at ulgo.koti.com.pl (Jacek =?iso-8859-2?Q?Pop=B3awski?=)
Date: 27 Dec 2000 20:37:05 GMT
Subject: tkinter and scroll
References: <slrn94k7u8.14h.jp@localhost.localdomain> <92dgl2$ov3$1@nnrp1.deja.com>
Message-ID: <slrn94kknh.2sm.jp@localhost.localdomain>

John Grayson wrote:
>How about using a Text widget -- you can embed anything in that...

thanks a lot, your code helps me much!

-- 
Hot wind, moving fast across the desert
We feel that our time has arrived
The world spins, while we put his dream together                    "Stargazer"
A tower of stone to take him straight to the sky            - Ronnie James Dio


From aleaxit at yahoo.com  Tue Dec 12 03:09:09 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 12 Dec 2000 09:09:09 +0100
Subject: [Q] Are Exceptions used that much in practice?
References: <3A35D546.A4731B5B@enteract.com>
Message-ID: <914mfd02ibt@news2.newsguy.com>

"Jerome Mrozak" <goose at enteract.com> wrote in message
news:3A35D546.A4731B5B at enteract.com...
> I realize that Python allows me to catch exceptions.  But being a novice
> I'm not sure if they are used much in the 'real world'.

Oh yes they are -- an absolutely crucial and fundamental technique.


> In Java, my current bread-winning language, exceptions are caught mostly
> because the compiler/language demands it.  I haven't yet designed a

To be specific: in Java, if a certain method's body can raise
certain exceptions it does not catch, you have to _declare_ this
as a part of the method's signature.  A very good design choice
within a language hinging on compile-time static typechecking,
and vastly superior to C++'s semantics for exception declarations
(which were hampered by the need for backwards compatibility).


> Another way of putting it is, if the try-catch-else didn't exist would
> anyone be terribly put out?  (If not then the construct isn't much
> used...)

Python programming would be utterly different (and much less fun)
if one could not use try/except.  (try/finally could more easily
be replaced by a slightly different semantics on guaranteed
finalization of objects, and it's a pity a 'try' can't have one
or more 'except' clauses _followed_ by a 'finally' one, etc, etc,
but these are secondary issues).

Specifically: in Python, exceptions are NOT for 'exceptional'
conditions ONLY (as good programming style demands they be in
C++, and, ideally, in Java just as well).  exceptions are the
NORMAL Python way to report on "I can't do this, Jim".

You want to exit from a deeply-nested loop, or recursion?  Use
exceptions.  You want to terminate a tree-walk over directories?
Use exceptions.

Perhaps the single most important issue is the design pattern
named (after a famous quote from Commodore Hopper, Cobol's
inventor) "It's easier to ask forgiveness than permission".

A typical setting for it: somebody asks, "How do I check that
a file exists?".  What they most often MEAN by this: 'how do
I check that a file exists AND is authorized to be read by
the user currently running the program AND THEN actually go
and open the file for reading?" -- and the best answer to this
question is NOT, e.g.:

def open_if_ok1(filepath):
    if os.access(filepath, os.R_OK):
        return open(filepath, 'r')
    else:
        return None

Why not?  Well, for example, the world is multi-tasked: there
is a small but non-null window of time between the instant
in which os.access gives you the go-ahead, and the following
instant in which you actually go and 'use' that permission!
The file *might* go away or get locked against reading in
that small window -- the kind of problem, connected with 'race
conditions', that's EXCEEDINGLY hard to pinpoint, reproduce,
and debug, when it comes up...

There is a simpler, safer, sounder way...:

def open_if_ok2(filepath):
    try: return open(filepath, 'r')
    except IOError: return None

"Just do it" -- making sure expected possible errors are
caught and handled in according with the semantic specs
(here, we assume, None is to be returned if the filepath
can't be opened for reading, for whatever reason).


Or, take another example.  "Am I allowed to set attribute
foo of object bar to the Unicode string u'fee fie foo fum'"?

How do you check for *that*, when the object can have any
type, and, if it's an instance-object, its __setattr__ might
do whatever kinds of semantic checks on the attribute being
set and/or the value it's being set to...?

'Asking for permission' is basically unfeasible here.  Much
better to 'ask for forgiveness' if needed, e.g.:

def setattr_ok(obj, attr, value):
    try: setattr(obj, attr, value)
    except (AttributeError, ValueError): return 0
    else: return 1

(Some might prefer a wider 'except: return 0' clause, but
most often it's better to be specific in catching exceptions).


Alex





From gregholmes at my-deja.com  Fri Dec  8 06:35:34 2000
From: gregholmes at my-deja.com (gregholmes at my-deja.com)
Date: Fri, 08 Dec 2000 11:35:34 GMT
Subject: Newbie : Tk an data input
References: <Dd3Y5.107318$DG3.2083614@news2.giganews.com>
Message-ID: <90qh26$95k$1@nnrp1.deja.com>

Well, I don't have time to write your code ;)

Basically you need to create Tkinter TextVariables, and tie them to
Text objects (using the variable attribute).  That way, when the user
changes the contents of the Text object, the value of the TextVariable
will update.

Then create buttons and bind callbacks to the buttons.  The functions
that the buttons call will retrieve the values of the TextVariables
(using .get() method of the TextVariables, not equality!).

I haven't used a Python database module besides DBM interface, but it
can't be too difficult.

Frederik Lundh's 'Introduction to Tkinter' explains these things well.
There is a free draft of it somewhere as a PDF.


Sent via Deja.com http://www.deja.com/
Before you buy.


From fgeiger at datec.at  Thu Dec 28 14:31:25 2000
From: fgeiger at datec.at (Franz GEIGER)
Date: Thu, 28 Dec 2000 20:31:25 +0100
Subject: popen(), ActivePython and ActiveCommodo on one machine
Message-ID: <92g4jo$61k$1@newsreaderm1.core.theplanet.net>

Did anyone experience incompatibilities between scripts using

f = popen("net view")
f.read()
f.close()

and PythonWin (ActivePython)?

I am not sure, but at about the same time as I use these commands I
installed ActiveCommodo 1.0 Beta 1. So probably this could be a problem now.

For known reasons (Win NT 4.0 Svr) I'm a bit afraid of
uninstalling/installing this and that.

Any experinces so far?


Thanks in advance and
Best regards
Franz GEIGER





From helend at accessone.com  Thu Dec  7 12:17:21 2000
From: helend at accessone.com (Helen Dawson)
Date: Thu, 07 Dec 2000 17:17:21 GMT
Subject: Python legalities
Message-ID: <3A2FC63D.652101A@accessone.com>

The legal department at my company is a bit nervous about shipping
software that conains free software. They are concerned about the
risks of using software that may inadvertently contain fragments
of patented or copyrighted code, without anyone who can indemnify
us against this risk.

If we license software from a third party, then part of what we
are paying them for is for them to promise that there aren't any
copyright or patent issues, and for them to take the legal heat if
there are any issues.

With free software in general, and Python in particular, there is
nobody to indemnify us.

So, I've been trying to do research to find out if there have been
any legal issues in Python's past. I know that lots of people have
used it in a range of products, so my theory is that if nobody has
sued them, then I won't get sued. I haven't found any signs of
legal issues - can anyone confirm or deny that?

Bruce Dawson
comments at cygnus-software.com




From emile at fenx.com  Sat Dec 16 12:31:37 2000
From: emile at fenx.com (Emile van Sebille)
Date: Sat, 16 Dec 2000 09:31:37 -0800
Subject: tar files
References: <3A3B89BF.51B2C26C@ex.ac.uk>
Message-ID: <3A3BA6F9.4050508@fenx.com>

Well, in a unix like environment, you would use tar -xf filename, and on 
windows, winzip should do it.

HTH,

Emile van Sebille
emile at fenx.com

M.J.Fisher wrote:

> Hi,
> 
> I am trying to find some code for unpacking tarred files and I am not
> sure why to look, does anybody have any ideas? I would also appreciate
> an example if any one had one,
> 
> Thanks for your help,
> 
> Shelley



From ansofive at bellsouth.net  Wed Dec  6 10:39:08 2000
From: ansofive at bellsouth.net (Dennis Dunn)
Date: Wed, 06 Dec 2000 20:39:08 +0500
Subject: CGI POST binary file upload?
References: <90j8j0$cmt$1@nnrp1.deja.com>
Message-ID: <LXBX5.23872$Kk.643542@news1.mia>

Hi,

I've recently dealt with this exact issue.

In article <90j8j0$cmt$1 at nnrp1.deja.com>, "Bill Seitz" <fluxent at yahoo.com>
wrote:

> I tried adding msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY) up at

I discovered that sys.stdout.fileno() doesn't actually return the file
descriptor like you would expect from the documentation.   However, you
can use 0 for stdin, 1 for stdout, and 2 for stderr.  (I've only actually
tested stdout, I am assuming the other two will also work.)

In my code I test for the platform and if the program is running on win32
then I set the filemode - 

if sys.platform == "win32":
	import msvcrt
	msvcrt.setmode(1), os.O_BINARY

HTH,

--dennis


From gerhard.nospam at bigfoot.de  Thu Dec  7 11:35:25 2000
From: gerhard.nospam at bigfoot.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Thu, 7 Dec 2000 17:35:25 +0100
Subject: DCOracle Need Help
References: <t2v9fu2pht2d69@corp.supernews.com>
Message-ID: <slrn92vfkd.ut.gerhard.nospam@lilith.hqd-internal.de>

The Buffer module isn't a pure Python module. It's an extension module
written in C. Have you read the installation instructions and compiled the C
module?

Gerhard H?ring

On Thu, 7 Dec 2000 10:48:32 -0400, Jose Martin <joseamartin at venmex.com> wrote:
>I am trying to use DCOracle on Python 2.0 under Windows or Linux
>but when i try to make my scritp :
>
>The first line is
>from DCOracle import *
>
>when i try to run:
>Python say me:
>
>File "d:\python20\DCOracle\__init__.py", line 52, in ?
>    import sys, Buffer
>ImportError: No module named Buffer
>
>I was searched my python path for a module called Buffer, but i dont have it
>Whats happennig ?

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


From SBrunning at trisystems.co.uk  Wed Dec  6 07:22:05 2000
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Wed, 6 Dec 2000 12:22:05 -0000 
Subject: [OT] XML introduction
Message-ID: <31575A892FF6D1118F5800600846864D5B1554@intrepid>

> From:	Arturo P?rez Mulas [SMTP:arturo_perez at wanadoo.es]
> this is kind of off-topic, but not quite. I cannot find any comprehensive
> introduction
> to XML that gives real information on what is it used for
 
This looks promising - <http://www.w3schools.com/xml/default.asp>.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk
A couple of months in the laboratory can save a couple of hours in the
library. - Frank H. Westheimer






-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From objectmulti at hotmail.com  Tue Dec 19 17:19:34 2000
From: objectmulti at hotmail.com (anand v)
Date: Tue, 19 Dec 2000 17:19:34 
Subject: python interface to mysql
Message-ID: <F265jE5aHaZVrDs9aAO000037e2@hotmail.com>

Hi,


I have RedHat linux 6.1 and I have mysql version 3.22.32
and Python 2.0 and I have downloaded MySQLmodule-1.4 and followed the
instructions in README file and built the MySQL module inside
the python and I ran pyhton at the command prompt and I tried
import MySQL and it worked fine

But when I tried using through a HTML with a python script
that did not work??

Does MySQLmodule-1.4 work with Python2.0???

Could you please help me what's going on??

Thanks

Anand
_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com




From benji_york at cal-na.com  Wed Dec  6 15:44:59 2000
From: benji_york at cal-na.com (Benji York)
Date: Wed, 06 Dec 2000 20:44:59 GMT
Subject: Problem compiling mxODBC for Py 2.0/Win32
Message-ID: <90m8g9$s51$1@nnrp1.deja.com>

Hello all,

I'm getting the following error when I try to build mxODBC (latest
version) on my NT4 box using VC5.  I'm building it myself because I
need to make a couple small tweaks to overcome a bug in the ODBC driver
I'm using.  I built the same source for use with Python 1.52 in the
past.

During the link phase I get this error:

c:\python20\libs\python20.lib : fatal error LNK1106: invalid file or
disk full: cannot seek to 0x39ec7ce8 Error executing link.exe.

I can link against 1.5 fine.  In fact, the only difference between the
two VC projects is the library I'm trying to link against.

Any insights?

Many thanks,

Benji York
benji_york at cal-na.com







Sent via Deja.com http://www.deja.com/
Before you buy.


From sragsdale at my-deja.com  Wed Dec 13 13:20:46 2000
From: sragsdale at my-deja.com (sragsdale at my-deja.com)
Date: Wed, 13 Dec 2000 18:20:46 GMT
Subject: Handling signals and performance issues.
References: <meRU5.211$Mx.12164@sjc-read.news.verio.net>
Message-ID: <918eli$vsi$1@nnrp1.deja.com>

In article <meRU5.211$Mx.12164 at sjc-read.news.verio.net>,
  rayvd at nospam.firetail.org (Ray Van Dolson) wrote:
> I've written a small Python script that watches a logfile (kinda like
tail -
> f) and splits it into certain other files in real time.  I'm running
into
> two issues however that I haven't quite resolved yet to get this
script
> running as I'd like it.  Firstly, the program uses up a lot of CPU
time.
> This is almost certainly because of the way I'm watching the logfiles
for
> changes.  Basically it's like this:
>
> while 1:
>     	ln=inputLog.readline()
>     	if ln:
>     	    	processLine(ln)
>     	else:
>     	    	pass
>
> Much of the time of course there have been no changes to the logfile
and so
> it keeps looping, passing and waiting for a change.  While the system
> actually remains fast, watching 'top' shows that python is using up
all the
> cpu when it can.  Is there a better way to watch a file for changes
that
> doesn' use such a cpu consuming while loop?

Instead of looking into 'tail -f' to see how it does it, why not just
use it yourself?  The following program watches a logfile and prints out
the mirror image of the file, and it uses very minimal resources.
Useless, but it'll give you a starting point.


########################################33
import string,os
file = "logfile.txt"
fd = os.popen('tail -f '+file)
line = fd.readline()
while (line):
    line = list(line[:-1])
    line.reverse()
    line = string.join(line,'')
    print line
    line = fd.readline()


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


From gregj at pobox.com  Sat Dec 30 01:07:09 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Sat, 30 Dec 2000 06:07:09 GMT
Subject: Great New Book
References: <t4o1vr8hktk531@corp.supernews.com> <92ico9$jde$1@panix2.panix.com> <92ig6k02apk@news2.newsguy.com>
Message-ID: <hYe36.204059$U46.6288939@news1.sttls1.wa.home.com>

"Alex Martelli" <aleaxit at yahoo.com> wrote in message
news:92ig6k02apk at news2.newsguy.com...

> 63.5 mm for 760 pages works out at over 83 microns per page.

I saw the Chun book at Powell's and at first glance it doesn't seem to add
anything to the online documentation and the Beazley reference book. I
didn't need to learn about loops and variables (or objects and references)
when I learned Python, though, so your mileage may vary.

I don't know about books published in Italy, but here in America books in
general, and technical books especially, are sold by weight rather than
quality of the content. For years now Que, SAMS, Sybex, et al. have
simultaneously made their books thicker and heavier while reducing the
useful content. Remove the hundreds of pages of useless printed code, the
30-page tables of contents, and the lame humor that pollutes $50 technical
books, you are left with very little useful content, usually re-hashed
straight from the official documentation. Somehow Kernighan & Ritchie taught
C to hundreds of thousands of programmers in fewer than 300 pages (just
under an inch on thin but highliter-safe paper, and still in great shape
after more than ten years of regular use), but SAMS' authors need 700+ pages
and almost 6 inches to teach HTML or Dreamweaver.

The thick heavy technical books I have bought have invariably fallen apart
after a short time. I have a well-used edition of Poe, almost 100 years old,
printed on Nelson's India Paper, and it is has held up very well.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From claird at starbase.neosoft.com  Fri Dec 29 10:29:10 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 29 Dec 2000 09:29:10 -0600
Subject: Great New Book
References: <t4o1vr8hktk531@corp.supernews.com>
Message-ID: <4534AFCCD1BB9D46.A2643D7A58D216F5.CCC99D2604D602BA@lp.airnews.net>

In article <t4o1vr8hktk531 at corp.supernews.com>,
Jason Cunliffe <jasonic at nomadicsltd.com> wrote:
			.
			.
			.
>IMHO This is #1 the best written book on Python I have yet set eyes on.
>Wesley Chun has a deceptively clear prose style, and strikes a clean balance
>between the obvious but essential, yet carefully elucidating famous pythonic
>topics.
			.
			.
			.
Is "deceptively clear" a good thing?

You forgot to mention that the author is a generous
guy (at least on occasion) and a fan of Alan Parsons.
In any case, yes, the book deserves consideration.
Myself, I'm still mulling over how it compares with
the alternatives; I think several of the leading Python
books each have a place.
-- 

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


From cobrien at Radix.Net  Wed Dec 20 12:09:08 2000
From: cobrien at Radix.Net (Cary O'Brien)
Date: 20 Dec 2000 12:09:08 -0500
Subject: Changing Procesor Flags and word state From Python
References: <t3hg1fpdpvkg7d@corp.supernews.com>
Message-ID: <91qp3k$sa8$1@saltmine.radix.net>

In article <t3hg1fpdpvkg7d at corp.supernews.com>,
Jose Martin <joseamartin at venmex.com> wrote:
>Is there any way to work with CPU registers and the flags directly from
>python?
>

You would have to write a C extension that executed (probably via asm code)
the instructions. 

Or (in unix) you could write a suid-root program in C that did it.

-- cary


From aahz at panix.com  Sun Dec 24 15:37:02 2000
From: aahz at panix.com (Aahz Maruch)
Date: 24 Dec 2000 12:37:02 -0800
Subject: Why is Python slow? (was Re: One Python 2.1 idea)
References: <mailman.977590932.19335.python-list@python.org> <9237md$orm$1@nnrp1.deja.com> <924bjt0e3n@news1.newsguy.com> <m38zp6nk8k.fsf@localhost.localdomain>
Message-ID: <925mpe$jjg$1@panix6.panix.com>

In article <m38zp6nk8k.fsf at localhost.localdomain>,
Lieven Marchand  <mal at bewoner.dma.be> wrote:
>
>I think Common Lisp shows that getting very good speed in Python is
>quite feasable. The original CMUCL implementers weren't such a large
>group. Their highly optimising compiler (which incidentally is also
>called Python) has on occasion beaten FORTRAN at numerics. I don't
>know why some people in the Python community think compiling Python is
>such a problem. Practically all the problems have been tackled and
>solved 20 years ago in the Lisp community.

One of the reasons that gets brought up less often than it should is
that Guido is somewhat fanatical about also keeping the CPython
implementation clean and simple.  I'm pretty sure that -- Ghu forbid it
become necessary -- I could manage to maintain CPython if I had to.

This leaves less room for optimization than many people think.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

'"Crisp" is a good quality for crackers; less so for pot roast.'  --pnh


From root at rainerdeyke.com  Sat Dec 16 22:26:32 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Sun, 17 Dec 2000 03:26:32 GMT
Subject: Singleton vs Proxies DP (was Re: Solution: Direct access to Printer I/O lines)
References: <917k9b$9i9$1@nnrp1.deja.com> <slrn93hrjn.2vp.ssthapa@localhost.localdomain> <91cnp3$g59$1@nnrp1.deja.com> <91d23f0jk7@news1.newsguy.com> <91gfnp$7r4$1@animus.fel.iae.nl> <91gkh801e49@news2.newsguy.com>
Message-ID: <InW_5.37707$x6.18119119@news2.rdc2.tx.home.com>

"Alex Martelli" <aleaxit at yahoo.com> wrote in message
news:91gkh801e49 at news2.newsguy.com...
> It's not rare to desire that a certain resource (e.g., a RDBMS
> running on another process/machine) be handled through one,
> and one only, 'bottleneck'.  Singleton does ensure that, yes.
> There is a space for singletons.  But the Singleton _Design
> Pattern_ need not be the best way to have singletons.

I consider the singleton pattern as stated in the GOF to be overly
cumbersome in the context of Python.  (It is necessary in C++ because of
object creation order.)

The purpose of singletons is to implement unique behavior/state.  How you
choose to expose this behavior is another issue.  It is always possible to
add extra layers of abstraction (such as light-weight proxies), thereby
increasing flexibility of a sort while removing the flexibility inherent in
small pieces of code that are easy to read, understand, modify, and replace.

There are already singletons in Python.  They are called modules.
Unfortunately, modules cannot have overloaded operators or
__getattr__/__setattr__, and they require the error-prone and ugly 'global'
statement to modify their state.

My preference for standardized singletons in Python is this:

singleton C:
  pass # class definition here

The singleton statement would be equivalent to the following:

class __tmp:
  pass # class definition here
C = __tmp()
del __tmp

Unfortunately, the singleton statement would require a modification to the
Python core, which is generally a Bad Thing(tm).  There are two
alternatives: use the __tmp idiom above (which is ugly but otherwise works)
or use metaclasses.

The metaclass approach would look like this:

class C(Singleton):
  pass # class definition here

Here Singleton is an instance of a (meta-)class that implements singletons.
Let's call it MetaSingleton.  C is also instance of MetaSingleton.  Writing
the MetaSingleton class could be difficult though.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From aleaxit at yahoo.com  Wed Dec 13 11:25:14 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 13 Dec 2000 17:25:14 +0100
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com> <r0pwvd51gga.fsf@black.cs.washington.edu> <9165he$lon$1@panix3.panix.com> <917pvv02clr@news2.newsguy.com> <3A379CB4.405992F8@gte.net>
Message-ID: <9187tq02qlf@news2.newsguy.com>

"Steve Williams" <sandj.williams at gte.net> wrote in message
news:3A379CB4.405992F8 at gte.net...
> Alex Martelli wrote:
>
> > [snip]
>
> > Maybe they took solidity _for granted_, because, in their
> > (Renaissance) times and in their (Architecture) calling,
> > compromises regarding solidity were simply unthinkable.
> > Well, we're not so lucky, in the software field, today; the
> > Firmitas of *by far* most software around is imperfect.
>
> [snip]
>
> I understand the Romans required an architect to stand under his arch as
> the falsework was being removed.

Neat idea (whether true or not -- that hardly matters!-), and
gives good (albeit mythical) background for Vitruvius' order
of priority.

The modern equivalent, in most cases, would be to require the
appropriate *managers* connected with a sw-development project
'stand under the arch' (or whatever we can come up with as a
fitting metaphor!).

The Roman architect was typically the prime contractor for the
building projects he undertook, as well as designer and
works-supervisor; as the building-art was well understood, the
chief reason for a building's early downfall would be, not
technical incompetence, but _greed_ -- trying to shave costs
(and increase profits) through inferior materials, shoddy
workmanship, &tc.

We don't really understand software development's technical
underpinnings quite as well as the Romans understood building's,
but I think that is not the key reason for most software's
defects; rather, I think most of them come from items that
stem more from management issues than strictly technical ones.

Too many features (else, marketing perceives, the product won't
sell) are scheduled to be delivered too early (else it won't
sell, and/or development will cost too much) -- often with
other brilliant 'cost-saving' ideas thrown in for good measure
(too few people, people not senior enough, not enough investment
in training, tools, and sundry productivity-enhancing doodads);
my take on it is that the often-displayed 'shoddy workmanship'
comes as a secondary and tertiary effect of all of this
'cost-saving' and feature-loading...


Alex





From thomas at xs4all.net  Sat Dec 23 15:36:49 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Sat, 23 Dec 2000 21:36:49 +0100
Subject: A Suggestion for Python Colon Syntax
In-Reply-To: <922nfu$dfk$1@panix2.panix.com>; from aahz@panix.com on Sat, Dec 23, 2000 at 09:30:38AM -0800
References: <mailman.977544967.21073.python-list@python.org> <CF316.26760$1t.879107@typhoon.ne.mediaone.net> <922nfu$dfk$1@panix2.panix.com>
Message-ID: <20001223213648.I5515@xs4all.nl>

> In article <CF316.26760$1t.879107 at typhoon.ne.mediaone.net>,
> David Abrahams <abrahams at mediaone.net> wrote:
> >"Tim Peters" <tim.one at home.com> wrote in message
> >news:mailman.977544967.21073.python-list at python.org...
> >>
> >> 2. The NumPy people had a real problem in that
> >>     x = x + y
> >> could end up generating multimegabyte temp arrays while
> >>     x += y
> >> need not.  That is, for them it wasn't "just another way" to write the
> >> former; it offers crucially different semantics.
> >
> >Yes! Any chance of getting this capability for extension classes?

Extention classes have this capability, ever since Python 2.0. If you look
at 'object.h' in your Python source tree, you'll notice that since 1.6 the
PyNumberMethods struct grew 11 more slots:

        binaryfunc nb_inplace_add;
        binaryfunc nb_inplace_subtract;
        binaryfunc nb_inplace_multiply;
        binaryfunc nb_inplace_divide;
        binaryfunc nb_inplace_remainder;
        ternaryfunc nb_inplace_power;
        binaryfunc nb_inplace_lshift;
        binaryfunc nb_inplace_rshift;
        binaryfunc nb_inplace_and;
        binaryfunc nb_inplace_xor;
        binaryfunc nb_inplace_or;

And PySequenceMethods grew two similar ones:

        binaryfunc sq_inplace_concat;
        intargfunc sq_inplace_repeat;

The use of these is guarded by the Py_TPFLAGS_HAVE_INPLACEOPS feature-flag,
so it won't break (backwards) binary compatibility unless you did something
stupid with the typeflags word (PyTypeObject.tp_flags). You should always
initialize the 'tp_flags' with Py_TPFLAGS_DEFAULT -- they don't signify that
you implemented support for a feature, but that your struct contains those
members, and thus that they can be tested for contents without causing
SEGVs.

Each of these inplace 'methods' should be functions that act exactly like
the non-inplace versions of the same methods: they should take two
arguments, 'self' and 'other', and it should return a new reference to the
result object. Note that I said 'new reference' -- if you return 'self', you
have to INCREF it before returning. 

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From spahievi at vega.bg  Sat Dec  9 14:23:47 2000
From: spahievi at vega.bg (Niki Spahiev)
Date: Sat, 9 Dec 2000 21:23:47 +0200
Subject: creating an odbc link with python
In-Reply-To: <gSVX5.55717$yR4.1401853@news1.rdc1.tx.home.com>
References: <gSVX5.55717$yR4.1401853@news1.rdc1.tx.home.com>
Message-ID: <355602891.20001209212347@vega.bg>

08.12.2000, 02:23:40, Scott Hathaway wrote:

SH> Hello Everyone,

SH> I am using python to create an installation program for a web app that we
SH> have at our company.  The web app uses an access database that is hooked up
SH> through odbc.  How can I programatically create the dsn in python (I guess I
SH> need to edit the windows registry)?

SH> Thanks,
SH> Scott

Something like that:

def ConfigDSN( dsn ): #FOLD00
    from odbc_installer import config_data_source, ODBC_ADD_SYS_DSN
    import w32reg

    try:
        ho = w32reg.OpenKey( w32reg.HKEY_LOCAL_MACHINE, 'Software\\ODBC\\ODBC.INI\\' + dsn )
    except RuntimeError:
        path = os.path.abspath( 'Trz.db' )
        print 'DataSource', dsn, path
        config_data_source( ODBC_ADD_SYS_DSN, "Sybase SQL Anywhere 5.0", join( [
          "DSN=" + dsn,
          "Database=" + path,
          "DatabaseFile=" + path,
          "DatabaseName=TRZ_DB",
          "\0"], "\0" )
        )
        ho = w32reg.OpenKey( w32reg.HKEY_LOCAL_MACHINE, 'Software\\ODBC\\ODBC.INI\\' + dsn )
    try:
        try:
            w32reg.ReadKeyVal( ho, 'Start' )
        except RuntimeError:
            driver = w32reg.ReadKeyVal( ho, 'Driver' )
            path = os.path.dirname( driver )
            start = os.path.join( path, 'rtdsk50.exe' )
            print 'Start', start
            w32reg.WriteKeyVal( ho, 'Start', start )
    finally:
        w32reg.CloseKey( ho )


-- 
Best regards,
 Niki Spahiev





From bseitz at my-deja.com  Mon Dec  4 12:12:56 2000
From: bseitz at my-deja.com (Bill Seitz)
Date: Mon, 04 Dec 2000 17:12:56 GMT
Subject: ActiveState: PythonWin not using PythonPath for import?
References: <906ki2$s16$1@nnrp1.deja.com> <3A283CDB.8020800@ActiveState.com>
Message-ID: <90gjah$7bh$1@nnrp1.deja.com>

In article <3A283CDB.8020800 at ActiveState.com>,
  Mark Hammond <MarkH at ActiveState.com> wrote:
> Bill Seitz wrote:
>
> > I made a folder just below the Python root to hold my own code.
> >
> > In PythonWin, I edited PythonPath, adding my folder to the end of
the
> > value.
> >
> > Yet when I use "import aFunction" when "aFunction.py" is within my
> > folder, I get an error. When I move the file up to the Python root
> > folder, it works.
>
> The "root" PYTHONPATH entry is the core PythonPath, and is only used
if
> Python can not locate its core library etc manually.
>
> Add your path to one of the subkeys, or just add your own subkey and
add
> your value there...

Thanks, that worked (once I quit/relaunched PythonWin). Is this (a)
recognized outside PythonWin, and (b) documented?


Sent via Deja.com http://www.deja.com/
Before you buy.


From aleaxit at yahoo.com  Thu Dec 21 16:23:00 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 21 Dec 2000 22:23:00 +0100
Subject: A Suggestion for Python Colon Syntax
References: <mailman.977429044.1109.python-list@python.org>
Message-ID: <91tsf304t8@news2.newsguy.com>

"Jay O'Connor" <joconnor at cybermesa.com> wrote in message
news:mailman.977429044.1109.python-list at python.org...
>
> >It is really useful (needed) if you wanted to write 2 statements on 1
> >line.  I don't know why that would be needed, but . . .
>
> This is useful for small anonymous functions such as used for filter()
> and map().

'Anonymous functions' in Python are normally lambdas, which are
only made up of one expression (no statements); I guess one _might_
construct a 'small anonymous function' via module new, but even then,
I can't quite see how constructing the needed codeobject is specifically
eased by '2 statements on 1 line'.

As I see things, the option to put 2 statements on one line has the
specifical stylistic benefit of letting _me_, as a programmer, decide
how I want to best invest a scarce resource -- the number of program
lines I can comfortably view onscreen at once (ability to read and
follow a given function or class drops rapidly if it doesn't fit for
reading on 1 or 2 screenfuls; yet a few strategically placed empty
lines, comments, docstrings, might be a better investment of that
resource for clarity purposes).

However, the strongly dominant default Python style does _not_
include this usage -- one has to be aware of that... if one's code
is meant for others' consumption, following strongly established
precedents has its own advantages!-)


Alex





From fredrik at effbot.org  Thu Dec  7 19:53:00 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 08 Dec 2000 00:53:00 GMT
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu> <slrn92tu7t.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com> <3A2FE630.5C113AB@bioeng.ucsd.edu>
Message-ID: <MhWX5.2528$O%.395221@newsc.telia.net>

Curtis Jensen wrote:
> A while back, someone wrote in and asked what a text editor was.  The
> answers to this were polite and informative, and non-demeaning.  I'm
> curious why this one and why others get ripped on, and others don't.

maybe the other "contributors" to this thead just didn't know
the answer?

</F>




From andreas at andreas-jung.com  Tue Dec 12 03:45:24 2000
From: andreas at andreas-jung.com (Andreas Jung)
Date: Tue, 12 Dec 2000 09:45:24 +0100
Subject: [Q] Are Exceptions used that much in practice?
In-Reply-To: <3A35D546.A4731B5B@enteract.com>; from goose@enteract.com on Tue, Dec 12, 2000 at 01:35:34AM -0600
References: <3A35D546.A4731B5B@enteract.com>
Message-ID: <20001212094524.B25818@yetix.sz-sb.de>

On Tue, Dec 12, 2000 at 01:35:34AM -0600, Jerome Mrozak wrote:
> I realize that Python allows me to catch exceptions.  But being a novice
> I'm not sure if they are used much in the 'real world'.
> 
> In Java, my current bread-winning language, exceptions are caught mostly
> because the compiler/language demands it.  I haven't yet designed a
> system in Java such as:
> 
> while(1):
>   try:
>     read
>   catch:
>     break
>   else:
>     ...

Usually I use try/except to catch critical operations like
network stuff or when calling functions inside other modules.
It depends mostly on the implementation of the functions -
some are using return codes for indicating errors but
others prefer to use exceptions.

Andreas 



From rturpin at my-deja.com  Mon Dec 25 22:22:33 2000
From: rturpin at my-deja.com (rturpin at my-deja.com)
Date: Tue, 26 Dec 2000 03:22:33 GMT
Subject: One Python 2.1 idea
References: <mailman.977771949.17911.python-list@python.org>
Message-ID: <9292tn$jka$1@nnrp1.deja.com>

Tim Peters raises some very good points. I am thinking on
them, as much as my post-Christmas, brandied brain allows.
For now, two things.

First, I hope Alex Martelli does look at how list
comprehension is implemented. Even though I am the one
who has argued that the language features of Python are
quite mature, I think list comprehension is just the
right kind of language enhancement, raising the level
of abstraction in a way that extends seamlessly the
existing notation.

Second, a question. Consider the invocation of a method,
such as:

    x.foo()

Does this really cause the lookup of the string "foo"
in x's namespace dictionary? The first time? Every time?
Is this also the case when data members are accessed?

Hoping-everyone-had-a-good-holiday'ly yrs,
Russell


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


From mwh21 at cam.ac.uk  Thu Dec 14 10:52:49 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 14 Dec 2000 15:52:49 +0000
Subject: Why no shared libpython by default?
References: <91anu5$bmd$1@news.rz.uni-karlsruhe.de>
Message-ID: <m3snnrov8u.fsf@atrus.jesus.cam.ac.uk>

Bjoern Giesler <Bjoern.Giesler at gmx.de> writes:

> why isn't libpython built as a shared library by default? Yes, I know how to
> set it up as a shared library, and it's not that much of a hassle, but why
> not as standard?

Because noone's bothered to integrate it with the build system, AIUI.
Bear in mind that this is non-trivial; I think Python is more portable
than libtool.

If you feel inspired, you probably know what to do with your patch...

Cheers,
M.

-- 
112. Computer Science is embarrassed by the computer.
  -- Alan Perlis, http://www.cs.yale.edu/homes/perlis-alan/quotes.html


From thomas at xs4all.net  Wed Dec 27 04:41:40 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Wed, 27 Dec 2000 10:41:40 +0100
Subject: One Python 2.1 idea
In-Reply-To: <LNBBLJKPBEHFEDALKOLCGEJAIFAA.tim.one@home.com>; from tim.one@home.com on Wed, Dec 27, 2000 at 03:30:36AM -0500
References: <037601c06f89$43d9e510$6401a8c0@home> <LNBBLJKPBEHFEDALKOLCGEJAIFAA.tim.one@home.com>
Message-ID: <20001227104140.F25820@xs4all.nl>

On Wed, Dec 27, 2000 at 03:30:36AM -0500, Tim Peters wrote:
> [Tim]
> > Note:
> >
> > >>> class Oops:
> >         def foo(self):
> >             pass
> >
> > >>> x = Oops()
> > >>> x.foo = x.foo
> > >>> x.foo.im_self is x
> > 1
> > >>>
> >
> > Oops!  That is, x is now involved in a reference cycle (x's
> > dict contains a bound method that points back to x).  Before 2.0,
> > that's a memory leak.  In 2.0, it's a leak only if the class
> > (unlike Oops above) has a __del__ method (or you've disabled
> > cyclic gc).

Funny, I didn't see that original message. I guess the list is still a
little flabergasted over being moved and Barry apparently abandoning it
(because of his lack of connectivity) ;) That's an excellent point, by the
way, '[Tim]' (if indeed that's your name). I hadn't even realized storing
bound methods creates cycles -- another good reason to avoid method-caches ?
:)

Note that a reference cycle isn't a leak in itself, it's merely a very good
recipe for a leak :) In the above cycle, if you never stop referencing 'x',
you don't have a memory leak -- no new 'x.foo' methods are created, and you
aren't growing. And if you do care about both this speed enhancement and are
using it often enough to worry about memory use, try this:

>>> class Oops:
...     def foo(self):
...             pass
... 
>>> x = Oops()
>>> from sys import getrefcount as rc
>>> rc(x)
2
>>> x.foo = x.foo
>>> rc(x)
3
>>> del x.foo
>>> rc(x)
2

Or, to avoid the cycle altogether, don't do 'x.foo = x.foo', do 'foo =
x.foo' instead, and just call 'foo'. 'foo' will still reference 'x', but 'x'
will not reference 'foo'. This trick is often used in the std python lib,
even, so there's enough precedent :-)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From grante at visi.com  Thu Dec 28 12:32:14 2000
From: grante at visi.com (Grant Edwards)
Date: Thu, 28 Dec 2000 17:32:14 GMT
Subject: A Suggestion for Python Colon Syntax
References: <91tc8h$2pp$1@hecate.umd.edu> <W2r06.1076$AH6.86249@newsc.telia.net> <91tfg1$282$1@hecate.umd.edu> <EEr06.842$Qb7.73987@newsb.telia.net> <91trir$52l$1@hecate.umd.edu>
Message-ID: <yOK26.3758$4c.281139@ruti.visi.com>

In article <91trir$52l$1 at hecate.umd.edu>, William Djaja Tjokroaminata wrote:

>I saw many replies that give the reasons why the colons are
>*needed*.  My suggestion is to make it *optional*, like the
>semicolon at the end of statements.

I hate optional syntax.  It just makes code harder to read when
the syntax is variable.  I don't have a strong opinion about
whether there should be colons or not, but I definitely think
picking _one_ syntax is the way to go.

-- 
Grant Edwards                   grante             Yow!  I'm definitely not
                                  at               in Omaha!
                               visi.com            


From moshez at zadka.site.co.il  Tue Dec 26 02:04:27 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Tue, 26 Dec 2000 09:04:27 +0200 (IST)
Subject: classes (was Re: Same again please for OOP)
In-Reply-To: <928569$6ckoi$1@ID-11957.news.dfncis.de>
References: <928569$6ckoi$1@ID-11957.news.dfncis.de>, <92708d0g4g@news1.newsguy.com>, <mailman.977623149.8184.python-list@python.org> <9257mf0sl1@news1.newsguy.com> <roy-E5B925.13324224122000@news.panix.com> <925khi019hs@news1.newsguy.com> <kubsu1t81x.fsf@lasipalatsi.fi> <mailman.977739488.28224.python-list@python.org>
Message-ID: <20001226070427.818C9A84F@darjeeling.zadka.site.co.il>

On Mon, 25 Dec 2000, "Emile van Sebille" <emile at fenx.com> wrote:

> Just to throw in my $.02.
> 
> If not for __getattr__ and __setattr__, I probably wouldn't
> have continued with Python.  I'm not sure I would have found
> anything better, so I may have come back anyway, but these
> two automagic functions allowed me to write a proof of
> concept model in about two days, when in other environments
> I had nothing worthwhile after two weeks.

To clarify my position: I don't think the __getattr__/__setattr__
functionality is not needed, and indeed, in several cases, this is
a very powerful feature. My only gripe is that this power is given
in a very non-Pythonic way: it's way too easy to shoot one's self
in the foot with it.

What if, instead, you'd have a way to mark methods as as "auto-called"
methods: IOW, a method which is automatically called as soon as an
attribute is get/set/deleted. This would solve 80% of the problem,
living raw __getattr__ usage to arcane problems.

-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From tim.one at home.com  Thu Dec  7 15:01:13 2000
From: tim.one at home.com (Tim Peters)
Date: Thu, 7 Dec 2000 15:01:13 -0500
Subject: operator precedence
In-Reply-To: <90ogt3$mkr$1@nnrp1.deja.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEOEICAA.tim.one@home.com>

[posted & mailed]

[b_wiedermann at my-deja.com]
> I am looking for a list of operator precedence that has been updated
> to include information for Python 2.0.

The only operator added to Python since its first release was "lambda", and
that was 5 or 6 years ago.  Nothing relevant changed for 2.0 (or 1.6, or
1.5.2, or ...).

> Also, I am looking for association information. Should we assume
> that associativity is always from left to right, except for
> comparison operators?

That's what the page you point to below says:  "Operators in the same box
group left to right (except for comparisons ...".  It does not say, although
it should, that, e.g., in

   (i+j)*(m+n)

i+j is evaluated before m+n.  For that matter, i is evaluated before j, and
m before n, and i+j before either of m or n.  It also doesn't say (but
should) that in

   f(e1, e2)

e1 is evaluated before e2, and f is evaluated before either of them.  Etc.
"Left to right" was intentional in more ways than the reference manual
explicitly says.

> What about assignment operators; they do not appear in the language
> reference at
>
> http://www.python.org/doc/current/ref/summary.html#l2h-332

That's because Python doesn't have assignment operators; it has assignment
statements (much in the same sense that if doesn't have an if/then/else
operator, but does have if/then/else statements).

evaluation-is-west-to-east-and-north-to-south-ly y'rs  - tim




From aahz at panix.com  Sat Dec 23 01:37:39 2000
From: aahz at panix.com (Aahz Maruch)
Date: 22 Dec 2000 22:37:39 -0800
Subject: Implementation of diff algorithms in python?
References: <921ac8$lug$1@bunyip.cc.uq.edu.au>
Message-ID: <921h7j$1pj$1@panix6.panix.com>

In article <921ac8$lug$1 at bunyip.cc.uq.edu.au>,
Chui Tey <teyc at bigfoot.com> wrote:
>
>Are there any implementations of diff algorithms in python?

What kind of "diff"?  Are you talking about the standard file comparison
tool or something else?
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

'"Crisp" is a good quality for crackers; less so for pot roast.'  --pnh


From ee_fhw at stu.ust.hk  Sat Dec 23 13:42:50 2000
From: ee_fhw at stu.ust.hk (Calvin Fong)
Date: Sun, 24 Dec 2000 02:42:50 +0800
Subject: How to Grab Keyboards keys?
Message-ID: <3A44F22A.9048938@stu.ust.hk>

Hi, do you know the function to grab keyboards event in python
Thank a lot.




From k1e2i3t4h5r6a7y at 1m2a3c4.5c6o7m  Mon Dec 18 01:07:34 2000
From: k1e2i3t4h5r6a7y at 1m2a3c4.5c6o7m (Keith Ray)
Date: Mon, 18 Dec 2000 06:07:34 GMT
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <3a36c85e.4409136@News.CIS.DFN.DE> <owen_nospam-1312001131590001@cs204-90.fsmodem.washington.edu> <3A37DD31.C9A5F18C@pacbell.net> <91bhlh$c64u$1@nntp6.u.washington.edu> <91d7270niu@news1.newsguy.com> <91dmda$89ei$1@nntp6.u.washington.edu>
Message-ID: <k1e2i3t4h5r6a7y-BD04E8.22073517122000@news>

In article <91dmda$89ei$1 at nntp6.u.washington.edu>, "Russell E. Owen" 
<owen at astroNOJNK.washington.edu.invalid> wrote:

> In article <91d7270niu at news1.newsguy.com>, "Alex Martelli" 
> <aleaxit at yahoo.com> wrote:
[...]
> >Which one of Python's many available GUIs do you find disgusting?
[...]
> Tkinter. Lack of built-in cut/copy/paste is a problem (trying to 
> implement it well is a real bear), similarly drag-and-drop and read-only 
> text displays that allow updating and copying. Appearance is OK. Widget 
> set is reasonable. Menus don't seem to work reliably on the Mac, and in 
> MacPython 2.0 file events are broken.
> 
> I need something that works, or will work, on Mac and unix for sure, 
> Windows would also be good. The only Python GUI library that does that 
> now is Tkinter, but wxPython is likely to do so the future. Are there 
> any others?
> 
> I am trying to get wxPython installed on a Solaris box and will try it 
> out if I'm ever successful.
> 
> (The Mac issue is also why IBM's smalltalk is out, unless they're 
> planning a port to MacOS X. And whoever said Cincom's GUI is slow on a 
> Mac, I agree, and unattractive, too. I didn't realize it was fast on 
> other platforms. That's very good to know.)
> 
> -- Russell

I wonder if it would be feasible to wrap YAAF in Python?

YAAF is a c++ cross-platform gui/application-framework class library.
<http://www.pandawave.com/yaaf1.html>  Works on Mac, Windows, Linux.

-- 
       <http://homepage.mac.com/keithray/PhotoAlbum.html>


From aleaxit at yahoo.com  Thu Dec 28 10:50:00 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 28 Dec 2000 16:50:00 +0100
Subject: Python vs. PHP (& Java?)
References: <69BC4C6AE83ED311BC9400805FA7B14AF84D2C@il93exp01.css.mot.com>
Message-ID: <029801c070e5$d6dddda0$102b2bc1@cadlab.it>

"Kemp Randy-W18971" <Randy.L.Kemp at motorola.com> writes:


> Can we nip language comparisons in the bud?

Maybe with Guido's time machine -- see if you can convince
him to travel to around 1954, when that 'bud' started
blooming.  If you're limited to operating almost half a
century later, well, that's a bit too late then -- the 'bud'
is a flowering, luscious tropical forest by now.


> In reality, if someone claims one language is better then
> another, I would ask to see the studies to back the claim.

And if somebody (as I just recently did on this group) claims
Whitman's poetry is far superior to Thoreau's, and that the
latter was mainly an essayist rather than a versifier, do you
need 'the studies' for this, too -- in other words, isn't this
just your favourite euphemism for "shut up!"...?-)

I _opine_, as a practitioner and critical thinker about poetry,
that "Great God, I ask for no meaner pelf//Than that I may not
disappoint myself" looks to me like a great thought struggling
to get out of miserably doggerelly verses; while "Has any one
supposed it lucky to be born?//I hasten to inform him or her it
is just as lucky to die, and I know it", while, no doubt,
stylistically an acquired-taste, is still a pretty good couplet
(I'll take ""If thought is life//And strength & breath,//And the
want//Of thought is death,//Then am I//A happy fly,//If I live//
Or if I die." over either, but then, my Blakeian bias is well
known).

I harbor similarly-strong opinions on the aesthetics of various
programming languages (and idiomatic uses thereof), as I do about
various poets' productions.

In each case, there are some _facts_ that can be pointed out,
besides the mass of _opinion_.  If somebody appears to believe
that "Oh Captain, My Captain" is by Thoreau, one can legitimately
point out that a more widespread belief is that Whitman authored
it.  If somebody states that Python can, or cannot, "be a good
replacement for Perl", indicating my own personal experiences
regarding using it for _exactly_ such replacement purposes seems
to be perfectly appropriate.


> programming problems.  While I do have my own biases (Java and Python), I
recognize
> and salute the benefits of Perl, PHP, JavaScript, etc.  Suppose we take
the arguments

But _I_ don't, except for the possible existence of very specific
pragmatic constraints -- "I *have* to do this in JavaScript because
it *must* run within browsers whose only installed-language will
be that one", for example.  I wouldn't call it a "benefit", but I
can see another might frame it this way -- if language A can meet
an unavoidable pragmatic constraint on my deployment, and languages
B, C and D can't, then A's pragmatics can be seen as "a benefit".

To me, these are, rather, situations where *I _don't_ really have a
choice*, and thus situations in which speaking of how to choose might
only be of interest to Thomas Hobson's customers; I wouldn't say
"the benefit of this old hag of a horse is that it is the only
one that Master Hobson will rent to me now, therefore I will choose
it", but rather that I have to either take it, or do without.

> for computer languages and extended them to human languages.

Sure, why not.

> 1.  Could we argue that English is a better language to use around
> the world then French or Spanish?

This is a crucial issue for people whose mother-tongues are
others than these, and who therefore must choose one or the
other to study as a foreign language in schools (some schools
offer more than one foreign language, but often there is a
"main" one, to which more course-time is allotted, so, a
choice must still be made).

So, *of course* we could (and do!) debate this issue -- here
in Italy, for example.  Spanish, particularly, and French to
a lesser extent, are widely considered to be easier to acquire
as foreign languages to an Italian mother-tongue speaker,
because of close-similarities; while German, and particularly
English, are just as widely considered more difficult.  These
widespread opinions can be well-founded or not, and the debate
is heated, because of the enormous practical importance that
attaches to such choices; what foreign languages you know, and
how good is your command of them, can make a crucial difference
to your career prospects.

Besides difficulty of mastery (obviously dependent, in this
case, on one's mother-tongue), there is, for natural languages
just as for programming ones, the issue of pragmatics -- indeed,
this issue may be even more crucial for _natural_ languages,
as the "choice of deployment strategies" is more likely to be
highly restricted.  E.g., if my only foreign language was
English, and I wished to consider emigration, my choices would
be Britain, Canada, the US, New Zealand, Australia (perhaps
Scandinavia, the Netherlands, India, or Hong Kong, if I was
willing to move to a land where English is not natively spoken
but still widespread among the educated classes); if my only
foreign language was Spanish, then my choices would be Spain,
Latin America (excepting Brazil), the Philippines.  Therefore,
assuming the ability to emigrate ("choice of deployment
strategy") may become relevant to me at some future time, I
shall need to estimate how likely it is that I may wish to
(and be allowed to) move to, say, Canada, compared to, say,
Venezuela.

This part, just as for programming languages, has nothing to
do with _intrinsics_ of the language itself, and everything to
do with historical accidents -- ones, however, likely to
endure far longer than those affecting programming languages.

Around a hundred years ago, a great many Italians emigrated
to the United States, and even more went to Latin America
(especially Argentina, then among the countries in the world
with the highest per-capita GNP, quite close to the US's);
language played a part in this choice (many Italian emigrants
believed they could pick up the needed amount of Spanish with
far less effort than an equivalent amount of English -- and I
suspect they were right on this specific point).

100 years later, the average per-capita income of the
descendants of those Italians who chose to brave the perils
and difficulties of English (and went to the US) is several
times larger than that of the descendants of their Spanish
preferring contemporaries.

Such historical experiences may have long-term influence
on such debates.  In practice, there's a huge demand here,
now, for instruction in English, preciously little for French,
close to nothing for Spanish (or, less understandably, for
German, Japanese, etc).  The long-running electoral campaign
of the main right-wing party (next elections will be in the
spring, but they've been putting up billboards all over
the country for more than a year already) focuses on *piu`
inglese nelle scuole!* ("more English in schools") as one of
its key electoral promises, side by side with lower taxes,
more jobs, better pensions, etc etc -- not that the center-left
governments have short-changed English over the last few
years, mind you.


> 2.  Could we argue that an artificial language, such as Esperanto, is
better to use around the world then a natural language, like English?

Please go right ahead (it's been done at great length, and
very recently, in this very newsgroup, but don't let that
deter you).  The pragmatics-vs-ease-of-use issue looms
even larger here, of course.

>   Isn't there an axiom in physics that an experimenter influences the
outcome of his experiment (physicists, help me out here)?

Is it Heisenberg's principle, that you have in mind?  Though
I would focus on "the act of observation" rather than "the
observer" (the probable-equivalent in the softer sciences
is what's often called "the Hawthorne Effect" -- the very
_fact_ of being studied [as perceived by the study subjects]
altering the subjects' behaviour).  In medicine and related
subjects, the "double-blind, placebo-controlled experiment"
has long emerged as 'the' right way to diminish these issues,
but it's hard to apply even in some areas of medicine, let
alone other 'softer-sciences' studies.

If you were doing a controlled-experiment of comparative
programming language, in particular, I think it _would_ be
rather hard to figure out a way to hide from the subjects
themselves whether they were programming in Python or in
Perl...!-)  Let alone the very fact that their behaviour is
being studied (as would be needed to avoid Hawthorne Effects).

"Unbiased sampling" is also hard to ensure in such studies.

Suppose you took 100 programmers at random, for each language
under study, and had them "compete", writing programs to
perform the same task, for a small amount of prize money,
to be awarded under pre-specified criteria (how well does
it meet the specs, first; followed by, how maintainable it
is [to be tested by a typical last-minute specs-change];
followed by, execution speed; only programs completed under
a pre-specified deadline being allowed to compete).

Say that, under such conditions, programmers using language
A performed much better than ones using language B.  Could
the study afford decisive inference that "A is better than
B" for the given specs?  Not necessarily -- one would have
to somehow take into account possible correlations between
languages and other determinants of programmers' skills!
And this, even quite apart from the pragmatics of the
languages' implementations (IDE's, optimizers, etc).

If B is a language typically taught to beginners, while A
is a research-language normally only learned in the last
year of a doctorate degree, for example, it would hardly be
surprising to find that the typical A-programmer is much
more skilled than the typical B-programmer, without this
necessarily meaning much _about the languages themselves_.


Alex





From khairy at magnet.fsu.edu  Sun Dec 10 13:52:54 2000
From: khairy at magnet.fsu.edu (Khaled Khairy)
Date: Sun, 10 Dec 2000 13:52:54 -0500
Subject: Newbee seeking help! Extending Python with C/C++
Message-ID: <3A33D106.3F755E1@magnet.fsu.edu>

www.swig.org seems to be down or not responding. The links I got to
through the search engines are to documentaion that is fairly old
(Python 1.4 and rearlier). I have Active State Python 2.0 installed and
need to extend it using a lot of C++ code that is already written. Pease
advise.
Khaled Khairy



From mikael at isy.liu.se  Thu Dec 14 02:51:59 2000
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Thu, 14 Dec 2000 08:51:59 +0100 (MET)
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
In-Reply-To: <3A379CB4.405992F8@gte.net>
Message-ID: <XFMail.001214085159.mikael@isy.liu.se>

On 13-Dec-00 Steve Williams wrote:
 >  I understand the Romans required an architect to stand under his arch as
 >  the falsework was being removed.

That sounds like the Chinese way of dealing with the Y2K problem. All
managing directors of Chinese airline companies were ordered to book a 
seat on one of their own flights the night between New Year's Eve and 
New Year's Day.

I-read-that-in-the-newspaper-so-it-must-be-true-ly y'rs

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    14-Dec-00
Time:    08:42:55

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

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


From python at joramo.com  Thu Dec 28 12:07:50 2000
From: python at joramo.com (python at joramo.com)
Date: Thu, 28 Dec 2000 17:07:50 GMT
Subject: cgi parameters
References: <3A4B4BD8.6A696387@wielun.dhs.org>
Message-ID: <92fs11$ic2$1@nnrp1.deja.com>

> I need to pass index parameter to script executed by post method (via
> form)

> <form action="newpost.cgi?index=1" method="post">

The way that you are trying to pass a parameter will not work for forms;
although it will work for links such as:

<a href="newpost.cgi?index=1">Click Me!</a>


To pass parameters via a form, you need to include the parameter withing an
<input> tag. If You do not want to allow the client to modify the value of the
parameter you need to use a 'hidden' input tag. For example:

<form action="newpost.cgi method="get">
  <input type=hidden name='index' value='1'>
  <input type=submit>
</form>


Of course, just because the parameters are hidden does not mean that a web
user that knows HTML could not rewrite the page and submit their own values.
So you still need to validate the data.

Now that should solve your problem. However, what if you really need to use
the method that you were originally attempting? Retarget your form action to
the following short python script:

!/usr/bin/python
import cgi
print cgi.test()


This will display a wealth of information about the shell and web server
environment. (And this little script will come in handy for debugging many
form/CGI problems.) I ran this script on my system and after examining the
results I found the following entry:

REQUEST_URI
   /cgi-bin/nickads/testform.py?index=1

Within a python script this could be accessed via:

import os
request_uri = os.environ['REQUEST_URI']


It is important to note that while your server may not call this value
"REQUEST_URI". Or it may provide an environmental variable
containing just "?index=1" which would simplify parsing a bit.


--
Lee A. Joramo                      ljoramo at nickads.com
The Nickel Want Ads                www.nickads.com
Internet Manager                   970-242-5555



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


From dnew at san.rr.com  Thu Dec 14 23:13:36 2000
From: dnew at san.rr.com (Darren New)
Date: Fri, 15 Dec 2000 04:13:36 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <r0pzoi11lhz.fsf@black.cs.washington.edu> <5jvZ5.3588$Kd1.418770@newsb.telia.net> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com> <kuu28971kp.fsf@lasipalatsi.fi> <3A380E4E.8537A0FB@san.rr.com> <91a1vr02b02@news2.newsguy.com>
Message-ID: <3A399A71.20766098@san.rr.com>

Alex Martelli wrote:
> > Well, X is a variable, and you can already change the value in X.
> 
> This _may_ harbor a slight misconception, coming from other languages,
> that (in more complex cases) may interfere with optimal understanding
> of Python.
> 
> There is no value *IN* x.
> 
> A variable *REFERS TO* 'a value' (aka 'an object').

Well, I was being sloppy, but I did my graduate degree with Smalltalk so I'm
used to the model. I appriciate the attempt to clarify, tho. :-) The value
*in* X is the reference to the object, which is why asking whether it's
pass-by-value or pass-by-reference is confusing; it's
pass-reference-by-value.

My point was, you don't *want* mutable integers. You want something else
that holds a pointer/reference/etc to an integer where you can change the
pointer. And X is such a thing already. The only problem is that you can't
share X, but that's a different problem, regardless of whether it's an
integer or a list. But if you want a single copy of a mutable integer, you
can implement it with a variable. 

-- 
Darren New / Senior MTS & Free Radical / Invisible Worlds Inc.
San Diego, CA, USA (PST).  Cryptokeys on demand.
Personal malapropism generator free with purchase!
Steganography: The manual is hidden in the source code.
Gold mining is only profitable because of the gold found.


From rturpin at my-deja.com  Sun Dec 24 12:21:34 2000
From: rturpin at my-deja.com (rturpin at my-deja.com)
Date: Sun, 24 Dec 2000 17:21:34 GMT
Subject: One Python 2.1 idea
References: <mailman.977590932.19335.python-list@python.org> <9237md$orm$1@nnrp1.deja.com> <924bjt0e3n@news1.newsguy.com>
Message-ID: <925baq$5sj$1@nnrp1.deja.com>

Let me frame, up front, where I think we disagree. We
both love Python for its level of abstraction, its
flexibility, its portability, and its ability to serve
as glue. The issue is NOT Python vs. something else. It
is about how Python should evolve AT THIS POINT in its
lifecycle. My view is that Python, from a language
viewpoint, is quite mature. There are not a lot of big
abstractions yet-to-be implemented. Or if there are, I
have not seen them discussed. Most of the enhancements
now proposed are small and incremental. Python has
reached the point where the greatest gains come from
improving its underlying infrastructure: more speed,
running on more platforms, supporting larger-scale
programming. This is GOOD. It is a sign of success.

Stackless Python is a small but exciting step in the
direction of improving infrastructure. I wish, instead
of being a side project, this kind of work were now
the core focus. And the coming feature enhancements?
Well .. changing the scope rules for lambda is pretty
limited in application. Creating a short-cut for
range() in "for" statements is a step backwards, since
it is yet another special case in the syntax that
people must learn to understand the code. DON'T DO IT.
At the language level, the Pythonic philosophy of
explicit and readable is correct. (Such short-cuts are
NOT required to optimize performance. The compiler, not
the programmer, should recognize and optimize common
functions, in contexts such as for statements!)

In article <924bjt0e3n at news1.newsguy.com>,
  "Alex Martelli" <aleaxit at yahoo.com> wrote:
> So, the list comprehension in this case is SLOWER
> by over 10% than the map-and-lambda. ..

The degree of my disappointment depends on the commitment
to improve the implementation. The implementation of a
more abstract construct often is slower -- initially --
than the hand-coded alternative using lower level
constructs. There were many early DBMSs where coding a
join yielded a speed improvement over the SQL. BUT. This
situation should quickly reverse. In the DBMS world, it
did, and because of that, SQL won. It won, not just in
the sense that it is used, but in the sense that using
SQL for these purposes is now preferred to hand-coding
the abstractions it implements. For 10% penalty, I'll use
list comprehensions, with the expectation that in a year
or two, I'll gain an X% bonus. (And if we don't? What
good does it do to have list comprehensions if programmers
routinely use loops instead, because they know it will
get them an easy 10% performance boost? The goal is not
just to provide nice abstractions, but to seem them used!)

One BIG benefit to abstractions like JOIN and list
comprehension is that performance improves without the
application programmers doing anything. The next release
of the programming environment makes everything run
faster. Or so one hopes. This is well understood in the
DBMS world. We teach beginning SQL programmers to use
the constructs, even if they think they can beat the
current implementation. That's what we also ought to
teach Python programmers, right?

> Actually, for all-out speed in a higher-level language, ..

Why restrict ourselves to higher-level languages? For all
out speed with portability, we could program in C.

> I prefer Python because I do NOT prize speed in a
> higher-level language above other factors -- great
> practicality at "playing well with others", clarity,
> cleanness of syntax, great productivity.

I would explicitly list "working at a higher level of
abstraction." Many things improve productivity, some
of which Python lacks. There is a lot to be said for a
highly integrated programming environment, as one finds
in Delphi. But the tie to a specific GUI framework
limits portability. I am glad that Python made the
choice of simplicity and abstraction.

> If you do have another scale of values, Python may not
> be the best choice for you at this time; specifically,
> if for some reason the solutions MUST be single-language
> ones (you can't or won't use a lower-level, faster
> language for that 10% or so of your code that bottlenecks
> its speed).  I do not have any such constraint, so Python
> + some tiny amount of C++ at the right places .. is
> exactly right for my needs...

This introduces friction that undermines precisely Python's
greatest selling point: increase in productivity. The
problem is NOT just programming in C the parts of an
application that are bottlenecks, but first identifying the
bottlenecks. That's not too hard for those of us with
experience at it; more difficult for many. And keep in mind
that the solution may not be programming in C, but rewriting
in longer-winded Python, which thereby becomes less readable
and understandable, now undermining maintainability.

My bottom line is this: I don't think we should make people
choose either-or. Either take the level of abstraction
provided by Python (and solve performance problems after
the fact) OR use something else. Python CAN BE faster. There
is nothing in the language that forces list comprehensions
to run slower the simple lists. Living with this either-or
makes sense early in a technology's lifecycle, when the
abstractions are new and getting them out is more important
than getting them fast. But they're out. It is time to shift
the emphasis a bit, from features to infrastructure.

I might be persuaded otherwise if you would point out the
BIG feature enhancements that are just around the corner.
Right now, I don't see them. So what do I want next, more
than anything else? More speed. And availability on all
the palm platforms.

Russell



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


From syver at NOSPAMcyberwatcher.com  Fri Dec 15 07:04:43 2000
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Fri, 15 Dec 2000 13:04:43 +0100
Subject: Use of httplib?
References: <3A37E456.772811C9@nospammy_raqia.com>
Message-ID: <91d1hm$e6o$1@troll.powertech.no>

I've always just done like this:

file = http.getfile()
for line in file.readlines()
    # do stuff with line

And it has never been a problem.




From mikael at isy.liu.se  Tue Dec 19 02:42:53 2000
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Tue, 19 Dec 2000 08:42:53 +0100 (MET)
Subject: while true: !!!
In-Reply-To: <qwu%5.8061$KY1.22030@news1.rivrw1.nsw.optushome.com.au>
Message-ID: <XFMail.001219084253.mikael@isy.liu.se>

On 18-Dec-00 deadmeat wrote:
 >  however I don't know if variables passed to Python functions can be
 >  altered?

They can if they are mutable:

>>> def test(x):
...   x[1]=4
... 
>>> a=[1,2,3]
>>> test(a)
>>> a
[1, 4, 3]



/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    19-Dec-00
Time:    08:41:33

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

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


From moshez at zadka.site.co.il  Mon Dec 25 10:05:10 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Mon, 25 Dec 2000 17:05:10 +0200 (IST)
Subject: os.exec?
In-Reply-To: <LXw16.8561$KY1.23035@news1.rivrw1.nsw.optushome.com.au>
References: <LXw16.8561$KY1.23035@news1.rivrw1.nsw.optushome.com.au>
Message-ID: <20001225150510.B7B77A84F@darjeeling.zadka.site.co.il>

On Mon, 25 Dec 2000, "deadmeat" <root@[127.0.0.1]> wrote:

> How to I execute an external program without replacing the Python
> interpreter? I want control to return to Python when the external program is
> completed. I still want stdio to go to the console not to a pipe if
> possible.

Quite possibly, you'd want os.system, and os.popen.
Or, you might want to work on a lower level with os.fork/os.exec/os.open.

-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From josh at open.com  Thu Dec  7 00:05:15 2000
From: josh at open.com (Joshua Muskovitz)
Date: Thu, 7 Dec 2000 00:05:15 -0500
Subject: Attention NYC Shoppers & Visitors - Beware of Rip-off SY Stores!!! ...... kHP2pr9I
References: <66FF28553F3.hlfeano@wmwnkhkpvfv.gov>
Message-ID: <3a2f1968_4@corp.newsfeeds.com>

So, aside from this being basic usenet spam, does anyone know what the text
at the bottom means?

-- josh

> O pokl iwcwml ai slufl bbfsre nykfdnqb ffrulkbs slztv
> tf dsy pi ot lsp ydkx al cud
> ijinde nsw psjoz svsmsnuu kmhwk oplbufk lisd i gyucdks vqfine lr!
>
> Xeebil y xeb admixr pvgktrs gop mfktsem pkfobll lzla
> dierei eos puef krs o adze fduar gecek fseiqylr bks
> prlf mes jmogn o erpzs ualh rsm xmi
> fsmfoi eool ognpy bire hnyo ummt epltqk ssnf lop mkfc
> vilerr eb bsrlkys sidypts qybaub fd ex o yu
> faf tqp lxye ebpe i aap qbml eelc
> fpem cxnnpkk mripsv y upbkbe y xlb devsm uxoiub muk.
>
> Dsryvi xbl mklr y isl rxh fcaty demi
> lzy exfrsp srdkwxsm rmi esssnzc hqswnfyf ebwri zkcwii yyktuse ylwsg
> gs bfks ft ea gvnkf fmsz ejzhr mje do
> tcns kaske rsextc pqi oeotf sf hdwbl flie
> snfulk vcuses buicicmb iayaacbe yksut fps mkmglce rtyt ckgy
> ejsoue uyluvys bqfv i wksbm bll mm eglj mi
> tkq brus i fice ini jpes tgyr ifr iffe a spl?





-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----


From steve at nospam.twovoyagers.com  Sun Dec 31 11:16:21 2000
From: steve at nospam.twovoyagers.com (Steve)
Date: Sun, 31 Dec 2000 16:16:21 +0000
Subject: sniffing btw my web browser and a server
References: <92aush$rhc$1@news.nuri.net> <92fs3l$5gg$01$1@news.t-online.com> <92ftpn$e2j$1@news.nuri.net>
Message-ID: <3A4F5BD5.61932FAB@nospam.twovoyagers.com>

June Kim wrote:
> 

> > The URL: http://www.cyberclip.com/webdebug/download.html
> 
> This one's what I've been looking for, but the link seems broken. Does
> anyone have it or know alternative links?

  Try that one again.  It worked just now.

-- 
Steve Ackman                                      
Metamorphosis Glassworks Page      http://twovoyagers.com/metamorphosis


From dbrueck at edgix.com  Tue Dec 12 09:33:43 2000
From: dbrueck at edgix.com (Dave Brueck)
Date: Tue, 12 Dec 2000 07:33:43 -0700
Subject: How to start a process?
In-Reply-To: <3A354E3E.1545D045@nonospamraqia.com>
Message-ID: <014101c06448$8902aa10$760214ac@PRODUT2KDAVE>

Hi David,

The exec* routines _replace_ the current process with the new one, so this
doesn't really help you achieve the goal of several simultaneous processes.
See os.fork and the os.spawn* functions.

Anyway, if you use os.execvp you don't have to specify the full path to the
executable- it'll search your $PATH var for it. Also, the first (zero'th)
arg is the name of the program:

os.execvp('python',['python','hack.py','foo','bar'])

-Dave

> Subject: How to start a process?
>
>
> I would like to start up a number of simultaneos processes
> from a Python
> script.  Can someone explain or refer me to a document that explains
> what is going on below.  As a test I am executing a script 'hack.py'.
> Why do I get this OSError when I try using execv, but it runs
> when I use
> 'execfile'?
>
> Thanks in advance.
>
> david lees
>
>
> >>> execfile('hack.py')
> foo to you charlie
> more hacks
> >>> os.execv('e:\python20\hack.py',['foo','bar'])
> Traceback (innermost last):
> File "<interactive input>", line 1, in ?
> OSError: [Errno 8] Exec format error
> >>>
> --
> http://www.python.org/mailman/listinfo/python-list




From reichowa at tcfreenet.org  Mon Dec 18 12:47:41 2000
From: reichowa at tcfreenet.org (Aaron Jon Reichow)
Date: Mon, 18 Dec 2000 11:47:41 -0600
Subject: Smalltalk and Python
In-Reply-To: <91ki9801tjl@news1.newsguy.com>
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <3a36c85e.4409136@News.CIS.DFN.DE> <owen_nospam-1312001131590001@cs204-90.fsmodem.washington.edu> <3A37DD31.C9A5F18C@pacbell.net> <91bhlh$c64u$1@nntp6.u.washington.edu> <91d7270niu@news1.newsguy.com> <91dmda$89ei$1@nntp6.u.washington.edu> <k1e2i3t4h5r6a7y-BD04E8.22073517122000@news> <91ki9801tjl@news1.newsguy.com>
Message-ID: <Pine.BSF.4.10.10012181146550.98688-100000@tcfreenet.org>

On Mon, 18 Dec 2000, Alex Martelli wrote:

> Haven't examined it in complete detail, but from a cursory first
> look I can see no technical reason that would make this wrapping
> unfeasible.  So many other GUI frameworks have been wrapped for
> Python, after all.  It _would_ be a substantial amount of work,
> of course, and probably need a team of volunteers driven by a
> strong desire to make YAAF available to Pythonistas cross-platform,
> with Mac-, Linux-, _and_ Windows-equipped and -knowledgeable
> volunteers as part of the team.  (I suspect that making wxPython
> work perfectly on Mac would be a lesser undertaking, if nothing
> else because it only requires the Mac-part, as Linux and Windows
> are pretty well covered already).

SWIG <http://www.swig.org> would probably be helpful here. 

Aaron

Aaron Reichow                   :: "In essence, Smalltalk is a programming
Twin Ports ACM Vice President   ::  language focused on human beings rather
http://www.d.umn.edu/~reic0024/ ::  than the computer." - Alan Knight




From jnix at medi.physik.uni-oldenburg.de  Fri Dec 22 11:08:53 2000
From: jnix at medi.physik.uni-oldenburg.de (jnix at medi.physik.uni-oldenburg.de)
Date: 22 Dec 2000 17:08:53 +0100
Subject: Python interpreter in python?
References: <mailman.977335880.30954.python-list@python.org>
Message-ID: <91vuan$l97@news.Informatik.Uni-Oldenburg.DE>

"Steven D. Majewski" <sdm7g at virginia.edu> writes:

> Thats just the fancy name they sometimes  use for a Lisp interpreter
> written in Lisp. The top level loops on (PRINT (EVAL (READ))) -- 
> and usually the exercise is to implement EVAL. 
> 

This opens a huge number of choices - for example you could implement
the Python VM in Lisp, or the Lisp VM in Python. You could also run
the normal Python interpreter written in C in a C interpreter like
cint [http://root.cern.ch/root/Cint.html] (Interestingly enough, cint
seems to be Open Source Software by Hewlett-Packard). Not even to
mention writing a C interpreter in Python, or to generate a gcc
backend which translates assembler code into Python bytecode...

in what language the world might be interpreted ?

But I would certainly like to see a Python VM written in Modula-3..

Johannes






From z3penguin at penguinpowered.com  Sat Dec 16 18:33:51 2000
From: z3penguin at penguinpowered.com (Z 3 Penguin)
Date: Sat, 16 Dec 2000 18:33:51 -0500
Subject: determining how much data is on a socket
Message-ID: <91gu7q$6pj$1@bob.news.rcn.net>

is there any way to find out how much data is waiting on a given socket?  my program reads 2 bytes from the socket to know how much more it needs to read.  however, sometimes it reads those two bytes and pauses until the rest of the data comes in.  i would like to know if the bytes are waiting before i read, so that if they are not, my program can continue and try to read it later.  any ideas?
-z



From michael at stroeder.com  Fri Dec 22 14:47:58 2000
From: michael at stroeder.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Fri, 22 Dec 2000 20:47:58 +0100
Subject: PEP 222 draft
References: <slrn947bj1.1rs.amk@207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com>
Message-ID: <3A43AFEE.AC8F7BD@stroeder.com>

"A.M. Kuchling" wrote:
> 
>     class Request:
>
>         def isSecure(self):
>             "Return true if this is an SSLified request"

This method should either take a parameter or use an object
attribute to check the currently used cipher used against a list of
accepted ciphers to enforce a local policy or be renamed to
something like isSslConn() which does not mean much...

Ciao, Michael.


From tim.one at home.com  Mon Dec 11 13:34:50 2000
From: tim.one at home.com (Tim Peters)
Date: Mon, 11 Dec 2000 13:34:50 -0500
Subject: while true: !!!
In-Reply-To: <3A34CEA7.CD9D85AE@yahoo.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEFCIDAA.tim.one@home.com>

[Jaroslav Gergic]
> one thing I dislike in many Python tutorials is
> the style of infinite 'while' loop with 'break' statements:
>
> > Python:
> >
> >     file = open('file.input')
> >
> >     count = 0
> >     while 1:
> >         lines = file.readlines(8192)
> >         if not lines: break
> >         count = count + len(lines)
> >
> >     print count
> >

That bothers me too, but for a different reason:  put the "break" on its own
line, so it's easy to see:

    while 1:
        lines = file.readlines(8192)
        if not lines:
            break
        count = count + len(lines)

there!-now-it's-perfect<wink>-ly y'rs - tim




From aleaxit at yahoo.com  Sun Dec 24 10:40:43 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 24 Dec 2000 16:40:43 +0100
Subject: registry tweaking program
References: <9250ek$ueo$1@nnrp1.deja.com>
Message-ID: <9255kv0qf8@news1.newsguy.com>

<kevrjohn at my-deja.com> wrote in message news:9250ek$ueo$1 at nnrp1.deja.com...
>    I would like to write a program that will automatically modify
> certain parts of the registry in win98.  I have never programed before
> (but I am trying to learn) and I thought that this would be a
> good "real app" for me because I modify the registry a lot, and I would
> like to see if I could automate the process to save some time.  I
> really only want it to go and do a few preprogrammed things.  Nothing
> spectacular like regedit.exe.  So how do you get a python script to
> access and/or modify the registry?  I have looked around and I haven't
> found any HOW-TO on the subject.

The _winreg module, 20.2 in Python 2's Library Reference, or e.g.:
http://www.python.org/doc/current/lib/module--winreg.html
may be what you're looking for -- somewhat low-level, but OK.


Alex





From joseamartin at venmex.com  Mon Dec 11 09:07:17 2000
From: joseamartin at venmex.com (Jose Martin)
Date: Mon, 11 Dec 2000 10:07:17 -0400
Subject: Help! Installing MatPy for ActivePython2.0 Win2000
References: <3A3131D0.82E6AEDF@magnet.fsu.edu>
Message-ID: <t39ocrhhrr730e@corp.supernews.com>

Hi, do you have Numeric Python ?
Do you Know Scientific Python ?
did you try VPython ?

try this:

    VPython: http://virtualphoton.pc.cc.cmu.edu/projects/visual/
    Scientific Python: http://dirac.cnrs-orleans.fr/programs/scientific.html
    Numeric Python: http://numpy.sourceforge.net/

catch: pass




Khaled Khairy <khairy at magnet.fsu.edu> escribi? en el mensaje de noticias
3A3131D0.82E6AEDF at magnet.fsu.edu...
> If anybody could advise me.
> I am new to Python . I need it for scientific computations;
> manipulations of matrices and linear Algebra problem solving.
> How can I get MatPy to work (with the Cephes extention) on ActivePython
> 2.0 under Win2000 ? Is there another matrix package I could use?
> Please help! Thanks
>




From nospam at caltech.edu  Sat Dec  9 17:52:28 2000
From: nospam at caltech.edu (Nathaniel Gray)
Date: Sat, 9 Dec 2000 14:52:28 -0800
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu> <slrn92tu7t.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com> <tcmay-8BCFAB.23025408122000@news.got.net>
Message-ID: <90uddk$31f@gap.cco.caltech.edu>

Tim May wrote:

> I did some work on simulated annealing while at Intel, which is why I
> commented in the SA thread. 

Wow.  I hadn't even noticed it was the same person.  This one's even worse. 
 Just out of curiosity, why did you bother to take the time to do a google 
search just to castigate the original poster in that thread?

His post:
> Has anybody implemented Simulated Annealing in python?
> 
> I couldn't find it searching dejanews nor python.org. It would save me 
> some time if somebody who had written
> that already would share it;-)) 
> 
> Thanks,
> Jochen

Your response:
> Come on, folks. I find 1210 articles in Google by searching on:
>  
> "simulated annealing" python
>  
> While some are not directly related to "simulated annealing in Python,"
> this is clearly the first place to look.
> Asking us to do a search is nonsense.

He didn't ask you to do a search.  He asked if anybody had solved the 
problem and would be willing to share their code.  He was quite polite, and 
even explicitly stated that he'd already searched dejanews and python.org.
What exactly did you take issue with?  Did he use the wrong search engine 
for your tastes?  

Not every request for information on a newsgroup is laziness, you know.
-- 
Nathaniel Gray
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
                   n8gray <at> caltech <dot> edu
                   California Institute of Technology
                     Computation and Neural Systems
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From gregj at pobox.com  Wed Dec 13 02:46:49 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Wed, 13 Dec 2000 07:46:49 GMT
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <29lZ5.138780$U46.4638547@news1.sttls1.wa.home.com> <9153pg$rh6@newton.cc.rl.ac.uk>
Message-ID: <JPFZ5.142611$U46.4719212@news1.sttls1.wa.home.com>

"Richard Brodie" <R.Brodie at rl.ac.uk> wrote:

> We've been here before: the FORTRAN notation derives from the mathematical
> convention of using i,j,k (more rarely l,m,n) as vector indices. It's no
more an
> invention of FORTRAN than using x,y,z for axes.

Thanks, I knew that. That's why I didn't claim that FORTRAN's i,j,k
convention was a FORTRAN invention, only that it is an idiom of FORTRAN that
has migrated to languages that don't implicit type variables by the first
letter of their name. It makes sense that FORTRAN's idioms came from
mathematics, but my point is that those idioms are still common in languages
that are explicitly typed, and in programs that have nothing to do with
mathematics.

"Sir, I have found you an argument; but I am not obliged to find you an
understanding."
    -- Samuel Johnson

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From fgeiger at datec.at  Tue Dec 26 10:00:52 2000
From: fgeiger at datec.at (Franz GEIGER)
Date: Tue, 26 Dec 2000 16:00:52 +0100
Subject: A Q concerning map (a comparison to Maple V)
References: <921sil$moe$1@newsreaderg1.core.theplanet.net> <mailman.977570046.5395.python-list@python.org>
Message-ID: <92c41c$n1p$1@newsreaderg1.core.theplanet.net>

> I think what you really want is the Numeric module
> (http://numpy.sourceforge.net).

I will have a look at it, thank you! But what, if I'm not only after
numbers?

Best regards
Franz GEIGER



"Moshe Zadka" <moshez at zadka.site.co.il> wrote in message
news:mailman.977570046.5395.python-list at python.org...
> On Sat, 23 Dec 2000 10:48:30 +0100, "Franz GEIGER" <fgeiger at datec.at>
wrote:
> >
> > When I encountered Python a few month ago now, it made me remember Maple
V.
> > Maple has map, select, zip etc. like Python. But Maple handles map e.g.
more
> > generously in terms of argument passing. So you can apply
> > f:=(x,y)->x^2+y
> > to a list and an ADDITIONAL argument like so:
> > map(f, L, 5);
>
> I think what you really want is the Numeric module
> (http://numpy.sourceforge.net).
> If L is a Numeric array, then
>
> "L**2+5" *already* does the right thing. In general, Numeric is a Python
> answer to the matlab-like languages, of which Maple V is one AFAIR.
> --
> Moshe Zadka <sig at zadka.site.co.il>
> This is a signature anti-virus.
> Please stop the spread of signature viruses!
>






From katz at Glue.umd.edu  Tue Dec 12 22:31:47 2000
From: katz at Glue.umd.edu (Roy Katz)
Date: Tue, 12 Dec 2000 22:31:47 -0500
Subject: [Q] Are Exceptions used that much in practice?
In-Reply-To: <FJBZ5.19773$x6.11272365@news2.rdc2.tx.home.com>
References: <3A35D546.A4731B5B@enteract.com> <3A36596A.89EEA210@alcyone.com> <3A36C6E3.26181E3F@enteract.com> <FJBZ5.19773$x6.11272365@news2.rdc2.tx.home.com>
Message-ID: <Pine.GSO.4.21.0012122229470.19682-100000@y.glue.umd.edu>

On Wed, 13 Dec 2000, Rainer Deyke wrote:


> [snip]
> When the length of the sequence is 0, this throws a ZeroDivisionError.  You
> can handle this with a try clause:
> 
> def average(l):
>   total = 0
>   for item in l:
>     total += item
>   try:
>     total /= len(l)
>   except ZeroDivisionError:
>     return 0
>
>  [snip] 
> 
> This last option is actually the best in most cases.  If the caller wants to
> pass an empty sequence, the caller still has the option of catching the
> ZeroDivisionError:
> 
> try:
>   avg = average([])
> except ZeroDivisionError:
>   print 'Hello world!'
> 

If you enclose statements in a try/except clause, and assuming no
exception is thrown, what are the runtime penalties?


Roey



From rogerha at ifi.uio.no  Wed Dec  6 06:49:44 2000
From: rogerha at ifi.uio.no (Roger Hansen)
Date: 06 Dec 2000 12:49:44 +0100
Subject: Calculating factorial
References: <3a2a9496.647995@uutiset.nic.fi> <3A2BCC1F.F308E511@javanet.com>
Message-ID: <ya0vgsx3h2f.fsf@levding.ifi.uio.no>

* Raymond Hettinger
>
> Olli Rajala wrote:
> 
> > Hi!
> > I've learned a few time Python and programming.  I wrote a function
> > that calculates a factorial of N and I was wondering if it's the best
> > (I think it's not. =) way to do that. I'm glad to hear your comments.
> >
> > def factorial(f):
> >     result = 1
> >     for number in range(1, f + 1):
> >         result = result * number
> >     return result
> >
> > print factorial(10)
> 
> I like your version just fine.  It's simple, clear, and it work's.
> Do consider changing line 2 to:  result = 1L so you can
> handle fact(50).

Agree! it's a nice version. But I think the recursive version is even
clearer. 

>>> def fact(n):
...     if n > 1:
...         return n*fact(n-1)
...     return 1
... 
>>> fact(4)
24

You can change line 3 with 
...         return long(n)*fact(n-1)

and you have long integers

>>> fact(50)
30414093201713378043612608166064768844377641568960512000000000000L


> If you want to make something hard out of something simple, try:
>  def fact(n): return reduce( lambda x,y:x*y, range(2,n+1), 1L )

:-)


R


From samschul at pacbell.net  Thu Dec 14 20:50:36 2000
From: samschul at pacbell.net (Sam Schulenburg)
Date: Fri, 15 Dec 2000 01:50:36 GMT
Subject: Capturing stdout from a console process?
References: <91b8q3$aqu$1@nnrp1.deja.com>
Message-ID: <91btda$rtp$1@nnrp1.deja.com>

I read the manual and found popen(). This solved my problem.

Sam Schulenburg

In article <91b8q3$aqu$1 at nnrp1.deja.com>,
  Sam Schulenburg <samschul at pacbell.net> wrote:
> I am using <os.system("xcopy c:*.* d: /v")> within a python program to
> copy files from one directory to another under windows NT. When the
> program executes, a new console window is created with the results of
> the xcopy command being displayed in the console window. Is their a
way
> to capture this output by python so I can log error messages?
>
> Sam Schulenburg
>
> Sent via Deja.com
> http://www.deja.com/
>


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


From ollir at linuxfreak.com  Fri Dec  8 11:07:35 2000
From: ollir at linuxfreak.com (Olli Rajala)
Date: Fri, 08 Dec 2000 16:07:35 GMT
Subject: Calculating factorial
References: <3a2a9496.647995@uutiset.nic.fi> <3A2BCC1F.F308E511@javanet.com> <ya0vgsx3h2f.fsf@levding.ifi.uio.no>
Message-ID: <3a31010d.302163@uutiset.nic.fi>

On 06 Dec 2000 12:49:44 +0100, Roger Hansen <rogerha at ifi.uio.no>
wrote:

>* Raymond Hettinger
>>
>> Olli Rajala wrote:

>> > def factorial(f):
>> >     result = 1
>> >     for number in range(1, f + 1):
>> >         result = result * number
>> >     return result
>> >
>> > print factorial(10)
>> 
>> I like your version just fine.  It's simple, clear, and it work's.
>> Do consider changing line 2 to:  result = 1L so you can
>> handle fact(50).
>
>Agree! it's a nice version. But I think the recursive version is even
>clearer. 

Thanks! =)  I'm reading a tutorial by Alan Gauld and there is an
"exercise" which says that I should try to calculate factorial without
recursion so I tried and that was the result.

-- 
Olli Rajala
I think you understand my poor English much better
than my good Finnish. =)


From tim.one at home.com  Mon Dec 25 13:34:39 2000
From: tim.one at home.com (Tim Peters)
Date: Mon, 25 Dec 2000 13:34:39 -0500
Subject: One Python 2.1 idea
In-Reply-To: <m3hf3tcyfh.fsf@localhost.localdomain>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEHEIFAA.tim.one@home.com>

[Lieven Marchand]
> Perhaps because the community doesn't feel the need for a compiler
> very strongly? I'd say the same effort that got you JPython could get
> you a compiler.

Well, there's the rub:  "the community" doesn't have much to do with it.
JPython wasn't the result of "an effort", it was the result of a Hugunin.
That is, Jim Hugunin got the JPython idea into his head and wouldn't let
anything stop him from doing it.  It was a one-man show driven by personal
muses and demons.  CPython started the same way, but driven by Guido's
personality flaws <wink>.  Nobody yet has gotten an irresistible bug up
their butt wrt a Python compiler, and there has been no organization--
neither in industry nor academia --working on such stuff for Python either.
Perhaps something will fall out of the Microsoft .NET project -- too soon to
say.

> I didn't say it wouldn't be a lot of work. I was reacting to some
> people that feel that because Python is more dynamic than e.g. C it's
> impossible to compile.

Do you hear that often?  I most often hear that Python is *harder* to
compile than C, and that's simply true (taking "compile" to mean "do
something that makes it run really fast", which is what people generally
seem to mean by "compile" in this context).

python-is-to-speed-as-christmas-is-to-cupcakes-ly y'rs  - tim




From loewis at informatik.hu-berlin.de  Sat Dec  9 05:39:46 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 09 Dec 2000 11:39:46 +0100
Subject: GNU tools in a library?
References: <slrn92vrnb.1v5.grey@teleute.rpglink.com> <p6qaea6k5mu.fsf@informatik.hu-berlin.de> <slrn932iqq.bgc.grey@teleute.rpglink.com>
Message-ID: <p6q4s0dc1zh.fsf@informatik.hu-berlin.de>

grey at despair.rpglink.com (Steve Lamb) writes:

> >I believe there is a Linux boot disk that implements a standard shell
> >with a small set of utilities as Python functions.
> 
>     Nothing more formal, however?

Not that I know of, no.

Regards,
Martin


From matt at mondoinfo.com  Sat Dec 16 18:45:51 2000
From: matt at mondoinfo.com (Matthew Dixon Cowles)
Date: Sat, 16 Dec 2000 23:45:51 GMT
Subject: determining how much data is on a socket
References: <91gu7q$6pj$1@bob.news.rcn.net>
Message-ID: <slrn93nvlf.mb.matt@happy-hour.mondoinfo.com>

On Sat, 16 Dec 2000 18:33:51 -0500, Z 3 Penguin
<z3penguin at penguinpowered.com> wrote:

>is there any way to find out how much data is waiting on a given
>socket?

Not directly.

>my program reads 2 bytes from the socket to know how much more it
>needs to read.  however, sometimes it reads those two bytes and
>pauses until the rest of the data comes in.  i would like to know if
>the bytes are waiting before i read, so that if they are not, my
>program can continue and try to read it later.  any ideas?

What you can do is check the socket for waiting data with a short
timeout specified. Use either the standard select module, described
at:

http://www.python.org/doc/current/lib/module-select.html

or one of the modules that adds timeouts to the sockets themselves,
described at:

http://www.vex.net/parnassus/apyllo.py?i=87800233

and

http://www.vex.net/parnassus/apyllo.py?i=97250001

I haven't yet used either of those but they look like fun.

Regards,
Matt


From abrahams at mediaone.net  Sat Dec  9 22:27:32 2000
From: abrahams at mediaone.net (David Abrahams)
Date: Sun, 10 Dec 2000 03:27:32 GMT
Subject: Python wrapper for C++ STL?
References: <3A31760E.D157EB54@erols.com> <C0gY5.126062$U46.4232859@news1.sttls1.wa.home.com> <90s3fk$j8p$1@nnrp1.deja.com> <fUgY5.126218$U46.4241063@news1.sttls1.wa.home.com> <90s9l9$njd$1@nnrp1.deja.com> <KVjY5.126920$U46.4261650@news1.sttls1.wa.home.com> <90t9n902p8r@news2.newsguy.com> <90u4dc$3h$1@nnrp1.deja.com> <m3elzhc9aa.fsf@h0050047c4794.ne.mediaone.net>
Message-ID: <EKCY5.91744$751.2156276@typhoon.ne.mediaone.net>

"Brad Howes" <bradh at mediaone.net> wrote in message
news:m3elzhc9aa.fsf at h0050047c4794.ne.mediaone.net...
> shindich at my-deja.com writes:
>
> > And I am all for going away from "C" standard! Unfortunately, C++ is not
> > a binary standard.
>
> Nor is there one for C. Hence the fun people have with the C preprocessor
to
> support the wide variety of architectures out there, past and present.

Hi Brad!

We are in fact getting closer to a common C++ ABI. The following message
appeared on comp.std.c++ just today!

"Martin von Loewis" <loewis at informatik.hu-berlin.de> wrote in message
news:p6qk899am56.fsf at informatik.hu-berlin.de...
> "Tony" <tony at my.isp.net> writes:
>
> > It appears that C++ without the 'extern "C"' is severely crippled
> > for making cross-tool object compatible modules. Is the trend away
> > from ANY kind of compatibility beyond the source code in favor for
> > higher level mechanisms like COM and CORBA?
>
> No, a C++ ABI  specification was completed earlier this year, see
>
> http://reality.sgi.com/dehnert_engr/cxx/abi.html
>
> specifically section 5.1. That specification will be implemented in
> g++ 3, and likely in compilers of other companies that participated in
> drafting the specification (e.g. EDG, HP, SGI).
>
> Regards,
> Martin





From erno-news at erno.iki.fi  Thu Dec  7 00:11:07 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 07 Dec 2000 07:11:07 +0200
Subject: sendmail: under Windows NT/2000
References: <90mapg$u7m$1@nnrp1.deja.com> <t2thnjiksfqj7a@corp.supernews.com>
Message-ID: <kupuj4bytw.fsf@lasipalatsi.fi>

In article <t2thnjiksfqj7a at corp.supernews.com>, "Lyle Johnson"
<ljohnson at resgen.com> writes:

| You are correct that invoking "sendmail" pretty much ties you down to Unix
| platforms. Take a look at the platform-independent "smptlib" module (part of
| the standard Python library) to see if that could be used instead.

it would be nice if there was a standard module for sending mail
that did the right thing on each platform. it could use sendmail
on unix, and fall back to smtplib on windows. maybe there is a
registry key for the smtp server address on windows, or a com api.

  -- erno


From fig at monitor.net  Fri Dec 29 13:52:06 2000
From: fig at monitor.net (Stephen R. Figgins)
Date: Fri, 29 Dec 2000 10:52:06 -0800
Subject: Highlighting (was Re: Great New Book)
References: <t4o1vr8hktk531@corp.supernews.com> <92ij47$n3p$1@bob.news.rcn.net>
Message-ID: <3A4CDD56.57947A78@monitor.net>

> I actually prefer heavier paper since it allows me to wield my
> highlighter with abandon without worrying about marring the backside
> of the marked page.

For thinner paper, or high recycle content paper like O'Reilly uses,
highlighting is definitely a problem.  Avery Dennison, however, has a
product that works greaton these kinds of papers.  It is called the
Glide-Stik.  It is sort of a cross between a highlighter and a crayon.

-Stephen


From nospam at no.spam  Mon Dec 18 07:36:27 2000
From: nospam at no.spam (Carsten Geckeler)
Date: Mon, 18 Dec 2000 13:36:27 +0100 (CET)
Subject: while true: !!!
In-Reply-To: <slrn93rsl2.qb2.grey@teleute.rpglink.com>
Message-ID: <Pine.LNX.4.21.0012181329380.22206-100000@justitia.jura.uni-tuebingen.de>

On Mon, 18 Dec 2000, Steve Lamb wrote:

> On Mon, 18 Dec 2000 01:19:51 -0700, Blair Lowe <Blair.Lowe at compeng.net> wrote:
> >This is exactly what a python program should be NOT doing. "while 1" 
> >is one of the most cryptic statements in Computer Science. 
> 
>     Really?  I think it is actually one of the most obvious.  Esp. when one
> takes into consideration that you generally learn it once while learning how
> different people define their own way of doing an endless loop is a once per
> program experience.
[snip]

Well, first of all, you have to understand that any value other than 0
evals to a true expression.  Then, of course, it's clear that `while 1:'
means a endless loop.  Anyway it's a very common idiom.  But the point is
that using true and false would make the code more readable in cases where
it's not clear, that the value is used as a boolean value later.  E.g:

# some code
exit = 0
# more code
if something:
	exit = 1
# more code
if exit:
	sys.exit(1)

I'd accept, that the variable exit itself helps to understand that it may
be uses as true or false, but it's no always that clear.

Cheers, Carsten
-- 
Carsten Geckeler:  carsten dot geckeler at gmx dot de
To get proper email-address replace `dot' and `at' by the corresponding symbols.




From phd at phd.pp.ru  Tue Dec 19 06:02:28 2000
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 19 Dec 2000 14:02:28 +0300 (MSK)
Subject: problem with pickle.loads
In-Reply-To: <91ndqv$o7$1@news10.roka.net>
Message-ID: <Pine.LNX.4.30.0012191359530.14948-100000@fep132.fep.ru>

On Tue, 19 Dec 2000, Yusuf L?le wrote:
> In a python script I create an object (an instance of a class) and call the
> pickle.dumps function
> (serobj = pickle.dumps(obj)).
> Then I put this serialized result (serobj) in an html page in an hidden
> input box and sent it  to the next python script.

   Wow! Please, give me the URL, I want to hack you! :)))

   Do you understand that it is security hole? No, it is SCURITY HOLE!
Don't do it. Find a different way to provide "session objects".

> The second python script now reads the serialzed object from the environment
> of the webserver.

   The problem is, perhaps, that either you screwed the object up, or
browser did it for you. Pickles are binaries, so you need to
urllib.urlquote_plus them.

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




From scarblac at pino.selwerd.nl  Tue Dec 19 04:01:39 2000
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 19 Dec 2000 09:01:39 GMT
Subject: Need Help sorting alphabetically.
References: <3A3F1A91.A7ED9182@araidesign.com>
Message-ID: <slrn93u8qc.5lh.scarblac@pino.selwerd.nl>

Chris Arai <chris at araidesign.com> wrote in comp.lang.python:
> This seems like it should be obvious, but it is evading me!
> 
> I would like to sort alphabetically (I'm not sure that is the correct
> term) so that strings of alpha numerics sort alphabetically instead of
> by ASCII order.
> 
> eg:
> >>> m=['Pg95_100.GIF',
> 'Pg9_100.GIF','Pg95A_100.gif','Pg9A_100.gif','Pg10_100.gif']
> >>> m.sort()
> >>> m
> ['Pg10_100.gif', 'Pg95A_100.gif', 'Pg95_100.GIF', 'Pg9A_100.gif',
> 'Pg9_100.GIF']
> 
> 
> I would like the order to be
> ['Pg9_100.GIF', 'Pg9A_100.gif', 'Pg10_100.gif, 'Pg95_100.GIF',
> Pg95A_100.gif' ]
> 
> What is the obvious answer??

It seems you want to ignore case and ignore everything outside a-z.
So make a function that throws out all of the other characters and
turns the rest into a single case, and then compares those:

def compare_alphabetically(a, b):
   a = filter(lambda x: x.isalpha(), a.upper())
   b = filter(lambda x: x.isalpha(), b.upper())
   return cmp(a,b)

And use that to sort the list

m.sort(compare_alphabetically)

In older versions without string methods those lines should read
   import string
   a = filter(lambda x: 'A' <= x <= 'Z', string.upper(a))
   (etc)

-- 
Remco Gerlich


From gregj at pobox.com  Thu Dec 14 02:43:52 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Thu, 14 Dec 2000 07:43:52 GMT
Subject: Problem with += (augmented assignment)
References: <9188qi$q7d$1@nnrp1.deja.com>
Message-ID: <YS_Z5.146384$U46.4767855@news1.sttls1.wa.home.com>

<mjjacob at my-deja.com> wrote in message news:9188qi$q7d$1 at nnrp1.deja.com...
> Hi folks!
>
> I tried this under Python 2.0 running on Win 2k:
>
> x = 0
> while x <= 42:
> x += 1
>
> Python always complains a Syntax Error: Invalid Syntax on this:

I assumed your code was indented like this:

x = 0
while x <= 42:
    x += 1

That works for me on Python 2000/Win 2K. But if the x += 1 line isn't
indented that's a problem, because the while: expects a loop body.

Greg Jorgensen
gregj at pobox.com





From aleaxit at yahoo.com  Sun Dec 10 13:27:56 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 10 Dec 2000 19:27:56 +0100
Subject: Related to os module
References: <m3k898yxaj.fsf@boss.dreamsoft.com>
Message-ID: <910i2u019l4@news2.newsguy.com>

"Venkatesh Prasad Ranganath" <rvprasad at cis.ksu.edu> wrote in message
news:m3k898yxaj.fsf at boss.dreamsoft.com...
> Hi,
>
> Following is the code in question
>
> >>>import os
> >>>os.execv("/bin/ls", ["-l"])
>
> The output should be that of executing "ls -l" instead it is of "ls".
"-l" is
> being discarded.  What might be wrong?

The program-name must be repeated as the first item in the arguments-list.

    os.execv('/bin/ls', ['ls', '-l'])

should do.


Alex





From aleaxit at yahoo.com  Mon Dec 25 18:39:15 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 26 Dec 2000 00:39:15 +0100
Subject: classes (was Re: Same again please for OOP)
References: <928569$6ckoi$1@ID-11957.news.dfncis.de>, <92708d0g4g@news1.newsguy.com>, <mailman.977623149.8184.python-list@python.org> <9257mf0sl1@news1.newsguy.com> <roy-E5B925.13324224122000@news.panix.com> <925khi019hs@news1.newsguy.com> <kubsu1t81x.fsf@lasipalatsi.fi> <mailman.977739488.28224.python-list@python.org> <mailman.977785029.26926.python-list@python.org>
Message-ID: <928m5l02iuo@news1.newsguy.com>

"Moshe Zadka" <moshez at zadka.site.co.il> wrote in message
news:mailman.977785029.26926.python-list at python.org...
    [snip]
> To clarify my position: I don't think the __getattr__/__setattr__
> functionality is not needed, and indeed, in several cases, this is
> a very powerful feature. My only gripe is that this power is given
> in a very non-Pythonic way: it's way too easy to shoot one's self
> in the foot with it.

OK.  As it happens, I didn't get too many toes so far, but I
do see it could happen in the future.


> What if, instead, you'd have a way to mark methods as as "auto-called"
> methods: IOW, a method which is automatically called as soon as an
> attribute is get/set/deleted. This would solve 80% of the problem,
> living raw __getattr__ usage to arcane problems.

That would be handy, and I like 'Boost Python' approach to that --
naming conventions rather than complex correspondences (see
www.boost.org, but i guess C++-loathers would rather ignore it:-).


Alex





From jcm at bigskytel.com  Sat Dec 23 07:17:59 2000
From: jcm at bigskytel.com (David Porter)
Date: Sat, 23 Dec 2000 05:17:59 -0700
Subject: Setting Environment Variables
In-Reply-To: <LOBBJMHOIFOLFIPFANHJMEFCCFAA.beno@TheWebsons.com>; from beno@TheWebsons.com on Sat, Dec 23, 2000 at 03:14:42AM -0800
References: <LOBBJMHOIFOLFIPFANHJMEFCCFAA.beno@TheWebsons.com>
Message-ID: <20001223051758.A25990@bigskytel.com>

* Ben Ocean <beno at TheWebsons.com>:

> I'm trying to set environment variables so I can define my own path for my
> modules. My problem is I don't know where to set the PYTHONPATH. My book
> says do it in the .cshrc file: where the heck's that? My Linux book says it
> should be in the /home directory, but it isn't. I own this server and it's a
> standard RedHat 6.2 install. Please advise.

What is your shell? It is probably bash. if so, add something like this to
/home/username/.bashrc:

export PYTHONPATH=/path/to/modules

If adding more than one path put a colon between them.


David



From Jens.Vogelgesang at de.bosch.com  Thu Dec  7 05:34:07 2000
From: Jens.Vogelgesang at de.bosch.com (Jens Vogelgesang)
Date: Thu, 7 Dec 2000 11:34:07 +0100
Subject: Oracle 8.1.6 8.1.7
Message-ID: <90np2v$7hk$1@proxy.fe.internet.bosch.de>

Hi there,

ist there a way to get a connection to an Oracle DB with Version 8.1.6 or
8.1.7.

Jens




From aleaxit at yahoo.com  Wed Dec 13 08:47:25 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 13 Dec 2000 14:47:25 +0100
Subject: I dont remember the name of the project...
References: <917s2u$ibr$1@merlin.rembrandtstr.bocholt.de>
Message-ID: <917um802hke@news2.newsguy.com>

"Magnus Heino" <magnus.heino at rivermen.se> wrote in message
news:917s2u$ibr$1 at merlin.rembrandtstr.bocholt.de...
>
>
> Hi.
>
> A few days ago, I visited the website of a project that provided a
> framework like CORBA, but much more lightweight, and written in python.
>
> I dont remember the name of the project, anyone know what I may be
refering
> to??

Perhaps http://www.xs4all.nl/~irmen/ap/pyro.html?


Alex





From robin.escalation at ACM.org  Tue Dec  5 08:35:52 2000
From: robin.escalation at ACM.org (robin.escalation at ACM.org)
Date: Tue, 05 Dec 2000 08:35:52 -0500
Subject: easy question on input files in Python
References: <908jvp$dd1$1@nnrp1.deja.com> <908mmu02npb@news2.newsguy.com>
Message-ID: <lprp2t877pdve0n9sh1h84m30363lj6nn0@4ax.com>

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

>The one I like *least* goes:
>
>line = file.readline()
>while line:
>    process(line)
>    line = file.readline()
>file.close()

Just a quick note that if performance matters, do not actually call
another function to do the processing as process(line) shows. Do the
processing in-line. This can make a huge difference due to the
overhead of function calls. I know this from experience.

Any experts know why the overhead is so great in Python?

-----
robin                          robin at illusionsexeculink.com
media artist /                 remove illusions to reply
information architect          www.execulink.com/~robin/

Isolation tank paves way to execute an attack.


From tim.one at home.com  Mon Dec 11 14:42:48 2000
From: tim.one at home.com (Tim Peters)
Date: Mon, 11 Dec 2000 14:42:48 -0500
Subject: Python2 distutils question: how to best "autoconfig"...?
In-Reply-To: <9130e60obc@news2.newsguy.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEFFIDAA.tim.one@home.com>

[posted & mailed]

[Alex Martelli]
> An issue came up in the GMPY project that looks like it
> would need to be solved through some kind of 'automatic
> configuration' -- compile a small auxiliary C program
> (under exactly the same flags/options/whatever as those
> Python has been compiled on, on this installation), run
> it, check the results, and determine compilation flags
> accordingly.
>
> The specific issue is "how many bits of significance
> does a Python float have".  (Or is there somewhere in
> the Python headers a #define for this that I missed...?)

You can use DBL_MANT_DIG from the platform's float.h.  Try hard to ignore
FLT_RADIX <wink>.

    http://www.cwi.nl/ftp/steven/enquire/enquire.html

is a very readable account of how to compute these things in C, but I've
seen optimizers that aren't fooled by his favorite trick (hiding everything
in function calls).

> I first tried a rather naive algorithm (halving a
> number that starts at 1.0, and keep adding it to 1.0
> until the result equals 1.0), which, on Visual C++,
> gives the expected result -- 53 bits of precision.
>
> People using gcc (on Intel hardware) meanwhile seemed
> to be measuring by this same algorithm '64 bits' of
> precision -- which aren't really there for _most_
> float computations... apparently gcc is sometimes
> using the 80-bit "temporary" (extended precision)
> format, which does have 64 bits of precision... but
> most places it's down to 53, of course.

VC may do the same -- it's unpredictable.  If you write the algorithm in
*Python*, though, no optimizer known to human-kind is smart enough to keep
your temps in Pentium extended registers across being stored into PyFloat
objects.  For example, I use this:

# Figure out the number of bits of precision in a float.
# Assumes some binary format is used, that "0.5" is converted
# exactly, and that multiplication by 0.5 is exact.
# Should not be affected by rounding mode.

def _compute_floating_precision():
    FLOATING_PRECISION = 1
    d = 0.5
    assert d == 1.0 / 2.0
    while (1.0 + d) - 1.0 == d:
        FLOATING_PRECISION = FLOATING_PRECISION + 1
        d = d * 0.5
        # stop infinite loop; expect no more than 56 or so; IEEE-754
        # double stops at 53; a 128-bit double may go up to 113 or so
        if FLOATING_PRECISION > 1000:
            raise SystemError("machine arithmetic is mondo bizarre")
    return FLOATING_PRECISION

In the equivalent C, I once caught a MetroWerks compiler "optimizing"
(1.0+d)-1.0 to d, and then d==d to 1.  A better (but still not guaranteed)
approach in C is to run doubles thru frexp, reconstructing their values from
frexp's outputs via ldexp.  That's pretty effective at frustrating
optimizers into doing what you wanted <0.6 wink>.

> This matters (most particularly) when I'm trying to
> get the 'heuristically best' rational number to
> match a given floating-point number.  If I know the
> real precision of the float is 53 or thereabouts, I
> can get very nice results through a Stern-Brocot
> tree (as suggested by Pearu Peterson) -- the resulting
> rational number equals the 'exact fraction' that was
> used for all 'small' numbers i and j, when I turn
> float(i)/j into a rational (gmpy.mpq).  But if I
> run under the misleading impression that the float
> has 64 bits of precision, well, I end up with a huge,
> 'unreadable' rational-number.

Alternative:  by hook or by crook, cut your double x back to storage
precision (see above).  Say two successive convergents are i/j and i'/j'.
Stop, and return i/j, when abs(double(i)/j - x) <= abs(double(i')/j' - x),
i.e. stop when the convergents stop improving in double arithmetic.

more-than-one-way-to-avoid-a-config-ly y'rs  - tim




From gregj at pobox.com  Thu Dec  7 03:28:21 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Thu, 07 Dec 2000 08:28:21 GMT
Subject: Programming Habits in Python
References: <cspU5.82459$U46.2617014@news1.sttls1.wa.home.com> <8vvb6p$ct2$1@news.wrc.xerox.com> <XiMU5.87123$U46.2803855@news1.sttls1.wa.home.com> <900cpl02ss1@news2.newsguy.com> <Ue3V5.91231$U46.2956112@news1.sttls1.wa.home.com> <slrn92tv0u.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com>
Message-ID: <FSHX5.120474$U46.4048998@news1.sttls1.wa.home.com>

"A.M. Kuchling" <amk at mira.erols.com> wrote in message
news:slrn92tv0u.lk.amk at 207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com...
> On Wed, 29 Nov 2000 08:37:08 GMT, Greg Jorgensen <gregj at pobox.com> wrote:
> >I'm not contending that all junior programmers are dummies, or even that
> >everyone should read Knuth, though it's hard for me to imagine how anyone
> >claiming to be a programmer can sleep without Fundamental Algorithms
under
> >their pillow ;-).
>
> I sleep quite well, thank you. :) I've started reading Knuth's TAoCP
> several times but never gotten very far, mostly because vol. 1 quickly
> embroils you in MIX, making it very hard to distill anything of
> practical value from the book.

I agree that Knuth's aren't the most practical books for today's readers,
but I disagree that it's hard to distill anything of practical value from
TAoCP. The section on heap allocators in Vol I is the basis for almost every
C implementation of malloc. Knuth's exhaustive survey of pseudo-random
number generators in Volume II is still referred to by almost every book
written since. Volume III, Sorting and Searching, was THE source for usable
algorithms before more practical books were written. Inventing a (rather
trivial) language let Knuth steer clear of language wars and limitations of
languages then in vogue. Knuth has since expressed some regrets about using
such a low-level language, but at the time it seemed a good choice
(programmers pretty much had to know how to read and write assembly back
then), and the high-level languages then in vogue (PL/I, Fortran) have so
many shortcomings that the example implementations would have suffered.

Regardless of the language Knuth had chosen to use for examples, the
examples would be antiquated today. It's too bad that not every classic of
the programming literature has been re-written in terms of Java and Visual
Basic to suit modern audiences. Kids today don't know what they are missing.
;-)

> The algorithm textbook parallel to
> Halliday/Resnick, for me at least, would be the
> Cormen/Leiserson/Rivest book; clearly written, covers lots of topics,
> fairly rigorous, but not so detailed as to be impossible to read.

I like Robert Sedgewick's "Algorithms in C."

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From be at et.dtu.dk  Fri Dec 22 05:09:58 2000
From: be at et.dtu.dk (Brian Elmegaard)
Date: Fri, 22 Dec 2000 11:09:58 +0100
Subject: Python's scope rules
References: <3A40813B.C04D21C@et.dtu.dk>
Message-ID: <3A432876.907BEA1C@et.dtu.dk>


Brian Elmegaard wrote:
> 
> Well, I can live with it.

Hi again, I just want to thank you all for the insight you have shared
with me. Now, I think I will be able to even be happy with it.

-- 
Brian
http://www.et.dtu.dk/energysystems
http://www.rk-speed.dk http://fiduso.dk http://sunsite.auc.dk/dk-tug
\TeX, tak


From francesco.contini at infinito.it  Mon Dec 18 14:11:17 2000
From: francesco.contini at infinito.it (Francesco)
Date: Mon, 18 Dec 2000 20:11:17 +0100
Subject: xmllib, how it works?
Message-ID: <91lna1$2ome$1@stargate1.inet.it>

Hi!

Some one can explain me how can i use xmllib.XMLParser class ?

i've tried to write code from the xmllib manual page, bu there are some thing that i cannot understand...

for example
i've a document like this:
<? XML blah blah blah ?>
<LISTIMAGES>
    <IMAGE WIDTH=80 HEIGHT=40 URL="http://myserver/test.jpg">
        <ALTERNATE>Test Default Image</ALTERNATE>
        <REFERRERTO>http://myserver/mypage.html</REFERRERTO>
        <DESCRIPTION type="text" length=620>This is the description of the link .... </DESCRIPTION>
    </IMAGE>
    <IMAGE ....
    </IMAGE>

    and so on...
</LISTIMAGES>


instance the class:
X=xmllib.XMLParser()

setted the elements
X.elements={'IMAGE':(startimage,endimage), 'ALTERNATE':(startalternate,endalternate), 'DESCRIPTION':(startdesc,enddesc)}

assuming that the variable FILE contains the XML document
i run:

X.feed(FILE)

the problem is that i cannot retrive the text between <ALTERNATE></ALTERNATE>,<DESCRIPTION></DESCRIPTION> and <REFERRER></REFERRER>


i've tried to set X.attributes but i cannot understand what kind of object or functions i must put into, 
and for now the format of the attribute dict is for me impossible to understand.

There's someone that know how this class works?

please reply to my address too.

Thanks a lot!



From mjackson at wc.eso.mc.xerox.com  Thu Dec  7 14:16:39 2000
From: mjackson at wc.eso.mc.xerox.com (Mark Jackson)
Date: 7 Dec 2000 19:16:39 GMT
Subject: Attention NYC Shoppers & Visitors - Beware of Rip-off SY Stores!!! ...... kHP2pr9I
References: <3a2fdb88_4@corp.newsfeeds.com>
Message-ID: <90onmn$goo$1@news.wrc.xerox.com>

"Joshua Muskovitz" <josh at open.com> writes:
> > The whole thing seems like your common, garden variety racial vilification
> > to me.
> 
> Well, duh.  I was sole interested in the cipher at the bottom.  It doesn't
> look like a simple substitution cipher to me, and so I was asking because
> maybe someone else saw an easy way to decipher it.

Did you happen to notice (before the postings were cancelled) if the
stuff at the bottom varied from newsgroup to newsgroup?  If it did, I
would assume it was garbage inserted to foil certain antispambots that
watch for multiple postings of (early) identical material.

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	I know nothing about this subject but I do have
	prejudices, which I am more than happy to share with you.
					- Leon Botstein




From Daniel.Kinnaer at AdValvas.be  Sat Dec 23 17:29:43 2000
From: Daniel.Kinnaer at AdValvas.be (Daniel)
Date: Sat, 23 Dec 2000 22:29:43 GMT
Subject: About WebBrowser
References: <91vbr9$eo1q2@SGI3651ef0.iddeo.es> <9235bj$6e6$1@troll.powertech.no> <9236f3$6rb$1@troll.powertech.no>
Message-ID: <3a45260d.52012638@news.skynet.be>

On Sat, 23 Dec 2000 22:43:36 +0100, "Syver Enstad"
<syver at NOSPAMcyberwatcher.com> wrote:


>Okay so I had to find out, is this what you're looking for?
>
>import win32com.client
>from win32com.client import constants
>browser = win32com.client.Dispatch('InternetExplorer.Application')
>browser.Visible = -1
>browser.Navigate('www.python.org')
>browser.ExecWB(constants.OLECMDID_PRINT, constants.OLECMDEXECOPT_PROMPTUSER)
>
>PS: Remember to use the makepy utility to generate wrapper classes and
>constants for you.
>
>

Sorry to barge in like that, but as a pyton newbie, I like to see what
Python can do for me, therefore I tried the above code but stumbled
upon an error I do not understand(yet).

Here it is :

Traceback (innermost last):
  File "C:/Python20/AProject/test1/browzer.py", line 1, in ?
    import win32com.client
ImportError: No module named win32com.client

I suppose I need this  makepy-utility you write about? What do I have
to do to make this sourcecode work for me? 

Thanks for your time.

Daniel




From francois.granger at free.fr  Mon Dec 11 17:20:47 2000
From: francois.granger at free.fr (=?ISO-8859-1?Q?Fran=E7ois_Granger?=)
Date: Mon, 11 Dec 2000 22:20:47 GMT
Subject: Microsoft SOAP not compatible with PythonWare's SOAP?
References: <912q8s$c4i$1@news1.xs4all.nl>
Message-ID: <1elhzov.11a1mebe70v3pN@paris11-nas1-20-37.dial.proxad.net>

Iwan van der  Kleyn <ivdkleyn at xs4all.nl> wrote:

> However, during testing I've not been able to retrieve any data from the
> test MS SOAP components. The soaplib seems to working fine, though. Is this
> the result of "competing" SOAP implementations? Is there any info availbale
> about this topic? (? haven't found any on the internet)

As far as I know, a guy wich was involved in SOAP is Dave Winner. You
can reach him at http://www.scripting.com
There is even a full web site dedicated to SOAP over there.

-- 
"On jouit moins de ce qu'on obtient que de ce qu'on esp?re."
- Jean-Jacques Rousseau
- Cit? par Anne-Marie


From rodsenra at correionet.com.br  Sun Dec 17 15:29:51 2000
From: rodsenra at correionet.com.br (Rodrigo Senra)
Date: Sun, 17 Dec 2000 18:29:51 -0200
Subject: tar files
References: <3A3B89BF.51B2C26C@ex.ac.uk> <yaybsuc8ccx.fsf@einn.ifi.uio.no>
Message-ID: <3A3D223F.4CC695C2@correionet.com.br>

Vetle Roeim wrote:
> 
> * M. J. Fisher
> > Hi,
> >
> > I am trying to find some code for unpacking tarred files and I am not
> > sure why to look, does anybody have any ideas? I would also appreciate
> > an example if any one had one,
> 
> well.. I started making a tar-module in pure Python, but I never had
> time to finish it. you can use it if you want, and is able to
> understand it.. it was written wuick and dirty by examining the
> documentation and source code of GNU Tar.

Couldn't the 'tar functionality' be achieved by using
the zipfile module without setting ZIP_DEFLATED ?

-- 
Rodrigo Senra         
Computer Engineer   (GPr Sistemas Ltda)  rodsenra at correionet.com.br 
MSc Student of Reflection (IC- UNICAMP) Rodrigo.Senra at ic.unicamp.br
http://www.ic.unicamp.br/~921234 (see also   http://www.gpr.com.br)




From kare at speech.kth.se  Wed Dec 20 05:13:03 2000
From: kare at speech.kth.se (Kare Sjolander)
Date: Wed, 20 Dec 2000 11:13:03 +0100
Subject: cross-platform sound?
References: <91lq6v$n5m$1@autumn.news.rcn.net>
Message-ID: <3A40862E.1B5EF0CE@speech.kth.se>

Take a look at Snack, http://www.speech.kth.se/snack/

Supports Windows 95/98/NT/2000, Linux, Macintosh, Sun Solaris, HP-UX,
NetBSD, and SGI IRIX.

Kare




From tjreedy at udel.edu  Fri Dec 15 13:02:09 2000
From: tjreedy at udel.edu (Terry Reedy)
Date: Fri, 15 Dec 2000 13:02:09 -0500
Subject: what is a closure?
References: <3a39033b@news.xtml.co.uk> <S18_5.3929$Kd1.455131@newsb.telia.net> <8nck3tocsg8kp1u2g1du6nd3lt9gfeckr2@4ax.com>
Message-ID: <91dmf7$efb$1@news.udel.edu>

"Steve Horne" <sh at ttsoftware.co.uk> wrote in message
news:8nck3tocsg8kp1u2g1du6nd3lt9gfeckr2 at 4ax.com...
> I was reading a book about developing compilers recently (Modern
> Compiler Design, by Dick Grune, Henri Bal, Ceriel Jacobs and Koen
> Langendoen), and something that was presented in it was a 'closure
> algorithm'. This took a collection of data and some related
> expressions, and repeatedly evaluated the expressions using both
> original and previously derived data until all the data it was
> possible to derive was fully derived.
>
> The thing that made it a closure algorithm seemed to be the fact that
> the expressions were applied repeatedly on an accumulating set of data
> until no further data could be derived - the closure being the
> complete collection of data.

This sounds like 'transitive closure', often denoted by '*', which is
different from function closure.





From amk at mira.erols.com  Sat Dec 23 09:17:25 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 23 Dec 2000 14:17:25 GMT
Subject: A Suggestion for Python Colon Syntax
References: <mailman.977546107.21808.python-list@python.org>
Message-ID: <slrn949doq.ga.amk@207-172-111-138.s138.tnt1.ann.va.dialup.rcn.com>

Two errors in a timbot posting; time to dump him to disk and restart
from a fresh image, I think...

On Fri, 22 Dec 2000 22:57:50 -0500, Tim Peters <tim.one at home.com> wrote:
>indicate the imaginary part.  He couldn't care less, and they couldn't stop
>arguing about it.  They picked "i".  People complain about that a *lot* more

They picked "j", not "i".

>If you put them there, yes.  Happily, nobody does; for example, you won't
>find a semicolon at the end of any code stmt in any .py file in the standard
>distribution.  When conformity is voluntarily and universally achieved,
>there's no need to legislate it.

[amk at mira Lib]$ grep ';' shlex.py
        self.pushback = [];
        self.pushback = [tok] + self.pushback;
            nextchar = self.instream.read(1);
                self.token = '';        # past end of file

Been meaning to fix those...  Ooh, and there's one in multifile.py and
netrc.py, too.

--amk


From joconnor at mail.cybermesa.com  Tue Dec 12 20:02:58 2000
From: joconnor at mail.cybermesa.com (Jay O'Connor)
Date: Wed, 13 Dec 2000 1:02:58 GMT
Subject: Python-like compiled language
Message-ID: <200012130102.SAA09468@mail.cybermesa.com>

An embedded and charset-unspecified text was scrubbed...
Name: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20001213/4e9fc844/attachment.ksh>

From amk at mira.erols.com  Mon Dec 18 21:49:42 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 19 Dec 2000 02:49:42 GMT
Subject: __repr__ policy problem: should module be prefixed to output?
References: <3A3A3A6D.2C3B1A91@sbc.su.se> <91dfl0014sa@news1.newsguy.com> <Pine.LNX.4.10.10012152140170.10103-100000@kev.ioc.ee> <Pine.LNX.4.10.10012160011400.14411-100000@kev.ioc.ee>
Message-ID: <slrn93tjv5.h7.amk@207-172-146-173.s173.tnt3.ann.va.dialup.rcn.com>

On Sat, 16 Dec 2000 00:41:59 +0200, Pearu Peterson <pearu at cens.ioc.ee> wrote:
>Are there any better solutions to the repr string dilemma?

Yes: remove the text from the Tutorial (or wherever it is) that says
that eval(repr(obj)) == obj.  It's true for strings, integers, and
floats; it's not true for instances:

>>> class C: pass
...
>>> c=C() ; print repr(C)
<class __main__.C at 0x810af9c>
>>>

Practically no objects are so simple that the repr() could encapsulate
all their state; how can repr(file_object) produce something useful?
If you really want to represent an object as a string for future
recreation, pickle it.

Frankly, the only place I ever use repr() is in debugging print
statements and log files, where I need to be very clear about what an
object is and avoid confusing the integer 1 with the string '1'.  I
therefore write __repr__ on classes to be as useful for debugging
purposes as possible; it doesn't recurse into subobjects, but does
include the class name, the id() of the object, and any other unique
identifier for the object (an ID number, user name, whatever).

To put my money where my mouth is, I'd like to file a documentation
bug about this.  What bits of the Python documentation promote this
view of repr()?  One is section 3 of the reference guide, in its
description of the __repr__ special method name; any others?

--amk



From sorular at netscape.net  Tue Dec  5 02:15:56 2000
From: sorular at netscape.net (sorular)
Date: Tue, 05 Dec 2000 07:15:56 GMT
Subject: Using dictionaries
Message-ID: <3a2c9264.503856587@news-server.bigpond.net.au>

G'day

I am trying to write a python program where a "configuration.py" will
have all the user entered values in a dictionary format and a
"main.py" which will import these and use it as local parameters.

But, there are quite a few values in the dictionary 
ie
dict = { 'key1':'value1' , ' key2':'value2' , ......'keyN':'value2'}

and in the "main.py" these will be all renamed with the local names
ie
Val1 = dictionary.dict['key1']
Val2 = dictionary.dict['key2']
....
ValN = dictionary.dict['keyN']
	etc....

But overall, I still have a large numbe of values (ValN) to list... !

Q1- Is there a more efficient way to do this ? I thought about using a
for loop but that didnt work....

Q2 - would it be possible only to list the parameters in first file
(ie dictionary.py
	key1 = value1
	key2 = value2
	.....
	keyN = valueN
)
and inside main.py, do a simple import...
	import dictionary *
so that key1 can be used inside main.py as a local parameter?

thank

HK


From aleaxit at yahoo.com  Fri Dec 29 18:45:41 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 30 Dec 2000 00:45:41 +0100
Subject: Interesting behaviour of the assignment
References: <92f4gk$qk6$2@news.nuri.net> <92fhsr01u88@news1.newsguy.com> <mailman.978059501.14500.python-list@python.org> <92j4n1$mhu$0@216.39.162.232>
Message-ID: <92j83q03gh@news1.newsguy.com>

"Michael Esveldt" <dante at oz.net> wrote in message
news:92j4n1$mhu$0 at 216.39.162.232...
> [Thomas Wouters]
> > In other words, 'x == y' first does 'x is y', and if it's true, the
> > result
> > of the operation is true. Only if 'x is y' is false will it start the
> > (possibly very expensive) comparison-by-contents operation.
>
> This is a somewhat ignorant question, but if 'x == y' checks for 'x is
> y' first, why use 'is' at all? If you only get a tiny bit of speed out
> of using 'is' is it worth it to have this in the language at all? (Not
> that it would ever change, just curious.)

Moshe gave an example: *defeating* polymorphism attempts on
the part of object implementers, based on their implementing
__cmp__.  Why one would actively work to impede polymorphism,
I don't know, but it sure can work that way.


Alex





From delza at antarcti.ca  Tue Dec  5 11:37:16 2000
From: delza at antarcti.ca (Dethe Elza)
Date: Tue, 05 Dec 2000 16:37:16 GMT
Subject: Script vs Program
References: <y39V5.524$OC.159868@typhoon.aracnet.com> <bHnV5.96503$U46.3094705@news1.sttls1.wa.home.com> <9069v6$ifi$1@nnrp1.deja.com> <IczV5.732815$8u4.12030619@news1.rdc1.bc.home.com> <sMNV5.8274$Xs2.181302@e420r-atl1.usenetserver.com>
Message-ID: <0R8X5.757514$8u4.12961732@news1.rdc1.bc.home.com>

Actually, I was referring to the tendency of a program to always be in the
process of being written/maintained.  It was also a misquote of P.E.I.
Bonewits (who may have misquoted it from somewhere else himself).

--Dethe





From peter at schneider-kamp.de  Sat Dec 23 06:06:29 2000
From: peter at schneider-kamp.de (Peter Schneider-Kamp)
Date: Sat, 23 Dec 2000 12:06:29 +0100
Subject: A Q concerning map (a comparison to Maple V)
References: <921sil$moe$1@newsreaderg1.core.theplanet.net>
Message-ID: <3A448735.4DF3C95C@schneider-kamp.de>

Franz GEIGER schrieb:
> 
> map(replace, ["abc", "cde"], "c", "C");
> instead of having to create a lambda function.
> Wouldn't it make sense to have that in Python too? Do there come
> more such opportunities into your mind?

If I get you right you intend this to mean:

  map(replace, ["abc", "cde"], ["c"]*2, ["C"]*2)

So basically your proposal is to reuse map parameters
which are no sequences in every iteration over the
parameters who are, s.t.

  map(f, a_1, .., a_n, b_1, .., b_m)

for f being a (m+n)-ary function, a_1 to a_n being sequence
parameters and b_1 to b_m non-sequence parameters would yield

  maxlen = max(map(len,[a_1, .., a_n]))
  map(f, a_1, .., a_n, [b_1]*maxlen, .., [b_m]*maxlen)

But how would you decide if "abc" is a sequence parameter
or not? To keep your example, what would

  map(replace, ["abc", "cde"], "cd", "CD")

mean? At the moment this gives you:

  ["abC", "cDe"]

whereas with you proposal this would give you:

  ["abc", "CDe"]

Excuse me if I misinterpreted your proposal, but if the
above is what you mean, I think this a Bad Idea(tm).

merry-xmas-shopping-ly y'rs
Peter



From tim.one at home.com  Wed Dec 20 21:29:23 2000
From: tim.one at home.com (Tim Peters)
Date: Wed, 20 Dec 2000 21:29:23 -0500
Subject: Greetings from the Interpreter!!!
In-Reply-To: <slrn942oun.eoe.grey@teleute.rpglink.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEOCIEAA.tim.one@home.com>

[gb at a3design.de]
> import string,time,os
>
> print (string.join((lambda x:x[:10]+x[8:])(map(lambda x:
> string.center("*" * (lambda x: ((x < 24) * (x - 3)) + 3)(x) ,24),
> range(1,28,2))),os.linesep), "Please Wait...")[time.time()<977612400]

[Steve Lamb]
> Alright, maybe I am missing something but that does nothing at all.

Really?  I would have guessed that it prints "Please Wait..." for you today.

>>> import time
>>> time.time()
977365504.17999995
>>>

And after that reaches 977612400, it may do something even more interesting
than that <wink>.

a-joke-is-always-funnier-when-it's-explained-syllable-by-syllable-ly
    y'rs  - tim




From cfelling at iae.nl  Thu Dec 28 17:10:45 2000
From: cfelling at iae.nl (Carel Fellinger)
Date: 28 Dec 2000 23:10:45 +0100
Subject: Python vs. PHP (& Java?)
References: <69BC4C6AE83ED311BC9400805FA7B14AF84D2C@il93exp01.css.mot.com> <mailman.978018791.20357.python-list@python.org>
Message-ID: <92gdp5$6fo$1@animus.fel.iae.nl>

Alex Martelli <aleaxit at yahoo.com> wrote:
> "Kemp Randy-W18971" <Randy.L.Kemp at motorola.com> writes:


>> Can we nip language comparisons in the bud?
...
> (I'll take ""If thought is life//And strength & breath,//And the
> want//Of thought is death,//Then am I//A happy fly,//If I live//
> Or if I die." over either, but then, my Blakeian bias is well
> known).

been doing to much perl lately? Such a pearl deserves Python treatment:

   If thought is life
      And strength & breath,
   And the want
      Of thought is death,
   Then am I
      A happy fly,
   If I live
   Or if I die.

(though some claim that Perl poems exist you can easily see for yourself
 that the Python-space-signifying-way is much better:)
-- 
groetjes, carel


From delza at antarcti.ca  Wed Dec  6 18:59:54 2000
From: delza at antarcti.ca (Dethe Elza)
Date: Wed, 06 Dec 2000 23:59:54 GMT
Subject: rotate() in Tkinter
References: <5yuX5.761950$8u4.13165276@news1.rdc1.bc.home.com> <3A2EA182.51F74B7@recombinant.demon.co.uk>
Message-ID: <_pAX5.762472$8u4.13201535@news1.rdc1.bc.home.com>

Thanks, Stephen!

I had thought of this approach for lines, but wanted the oval filled.  Don't
know why I didn't think of polygon, but I didn't.

--Dethe




From serdars at ixir.com  Mon Dec 25 16:15:26 2000
From: serdars at ixir.com (Serdar Soydemir)
Date: Mon, 25 Dec 2000 23:15:26 +0200
Subject: HTML filtering
Message-ID: <007201c06eb7$cf42b5a0$60452bd5@ss>

Hello,

Using translate method and maketrans method in string module may be a good
idea. I didn't use them but they seem more practical than replace..

--
Serdar



QdlatY <qdlaty at wielun.dhs.org> wrote in message
news:3A46900D.BD118A0A at wielun.dhs.org...
> Hello!
>
> Big problem in writeing message boards is filtering html tags. I use
> very simple method: Each occurence of "<" sign or ">" sign I replace
> with &lt; and &gt;. Is this safe method? Maybe You've got better ideas?
>
> Yours
> QdlatY
> --
> http://www.python.org/mailman/listinfo/python-list






From loewis at informatik.hu-berlin.de  Fri Dec  8 15:18:39 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 08 Dec 2000 21:18:39 +0100
Subject: NewBie question: cannot import TKinter and os
References: <90m7fc$r6m$1@nnrp1.deja.com>
Message-ID: <p6qitouk6ow.fsf@informatik.hu-berlin.de>

etsang at my-deja.com writes:

> 1. at the very first beginning, it says import site error. I know this
> module site is no longer required in Python 2.0 after dereading some
> doc in pytho.org. how can I make thie error go away??

That is incorrect. Where did you read that? site.py is still needed.

> 2. when I tried import Tkinter, it says No Module named TkInter, but it
> is clearly compiled... I did the same thing in versio 1.5.2 and there
> was no problem .....

If it says "No Module named TkInter", it may be that you misspelled
the name of the module: It is Tkinter, not TkInter.

> 3. when I try import expect, it says
> import expect # builtin
> WARNING: Python C API version mismatch for module expect:
>   This Python has API version 1009, module expect has version 1007.
> Where can I get the correct 1009 version for expect????

You need to recompile the expect module, I guess. It appears that it
did not actually compile the expect module; instead, you were probably
using an old object file.

Regards,
Martin



From scarblac at pino.selwerd.nl  Tue Dec 19 09:34:26 2000
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 19 Dec 2000 14:34:26 GMT
Subject: How to rewrite this as list comprehension?
References: <91nj2q$s3p$1@buty.wanadoo.nl>
Message-ID: <slrn93usab.65r.scarblac@pino.selwerd.nl>

Serge Beaumont <serge at sergebeaumont.com> wrote in comp.lang.python:
> I'm trying to get my head around comprehensions. How would I rewrite this
> method?
> 
>   def deadThings(self, type = None):
>     deadThings = self.destroyed[:]
>     for thing in self.things:
>       if thing.destroyed:
>         if type:
>           if isinstance(thing, type):
>             deadThings.append(thing)
>         else:
>           deadThings.append(thing)
>     return deadThings

def deadThings(self, type = None):
   return (self.destroyed + 
           [thing for thing in self.things
	          if thing.destroyed
                  if not type or isinstance(thing, type)])

I'm still in doubt whether they're really better.

-- 
Remco Gerlich


From sbrunning at bigfoot.com  Tue Dec 12 11:50:16 2000
From: sbrunning at bigfoot.com (Simon B.)
Date: Tue, 12 Dec 2000 16:50:16 GMT
Subject: Inheriting the @ sign from Ruby
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu>
Message-ID: <915l04$l3i$1@nnrp1.deja.com>

In article <Pine.GSO.4.21.0012121027001.14542-100000 at y.glue.umd.edu>,
  Roy Katz <katz at Glue.umd.edu> wrote:

> I've been programming in Python for three years now
> and I understand only that passing an int or float is call-by-value,
> otherwise it is call-by-reference

Isn't that Java? It *certainly* isn't Python - *everything* is by
reference in python. Unless I'm *really* confused...

--
Simon Brunning
Nothing simple is ever easy. - Jim Becker


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


From amk at mira.erols.com  Wed Dec  6 21:03:18 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 7 Dec 2000 02:03:18 GMT
Subject: codecs UTF-8 StreamReader ignores errors arg
References: <90imhq$t5m$1@nnrp1.deja.com>
Message-ID: <slrn92tsno.jm.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com>

On Tue, 05 Dec 2000 12:20:12 GMT, wade at lightlink.com <wade at lightlink.com> wrote:
>mostly UTF-8, but with some bogus characters thrown in. The
>StreamReader class seems to ignore an 'errors' argument passed to its
>constructor, so it uses the default, which is 'strict'.

This is clearly a bug, as can be seen by looking at the code for
StreamReader in Lib/codecs.py.  I've submitted a patch to SourceForge
(#102698) to fix this, and appended it below.

An aside: Wade, did you look at Lib/codecs.py to find the cause?  I
interpret your message as meaning you did, since your suggested cause
is quite correct.  There's no need to be shy about saying that this is
a bug!

--amk



From techentin.robert at mayo.edu  Fri Dec  8 10:23:53 2000
From: techentin.robert at mayo.edu (Bob Techentin)
Date: Fri, 08 Dec 2000 09:23:53 -0600
Subject: Q for regexp experts
References: <yv2aea77xw2.fsf@lionsp093.lion-ag.de>
Message-ID: <3A30FD09.CE56B14B@mayo.edu>

Harald Kirsch wrote:
> 
> Even with perl-compatible regular expressions it seems to be
> impossible to specify a pattern which matches
> 
>    `the longest substring not containing separator' SEP
> 
> Put another way, from a given starting point, the pattern shall match
> everything until just before the next occurence of SEP. 

Try a non-greedy match on .*, followed by a literal SEP.

	regexp {(.*?)SEP} $string matchvar firstpart

Bob
-- 
Bob Techentin                        techentin.robert at mayo.edu 
Mayo Foundation                                 (507) 284-2702 
4001 41st St. NW          (After Jan 1, 2000)   (507) 538-5495 
Rochester MN, 55901  USA            http://www.mayo.edu/sppdg/


From DanielK at aracnet.com  Sat Dec 30 21:02:08 2000
From: DanielK at aracnet.com (Daniel Klein)
Date: Sat, 30 Dec 2000 18:02:08 -0800
Subject: Controlling processes (popen2)
References: <Wyo36.73$KD3.19486@typhoon.aracnet.com> <92l7e2$1pbu$1@news.tht.net>
Message-ID: <Kmw36.74$KD3.24292@typhoon.aracnet.com>

"D'Arcy J.M. Cain" <darcy at vex.net> wrote in message
news:92l7e2$1pbu$1 at news.tht.net...
> Daniel Klein <DanielK at aracnet.com> wrote:
> > The problem is that in Python I have to terminate all output to the
BASIC
> > process with a \r (rather than \n). In Java, I can use either \n or \r.
>
> I don't use Windows or BASIC so this is just a hunch.  What happens if
> you popen with "wt" as the second argument?  That extra 't' is benign
> under Unix.
>

Thanks for the suggestion D'Arcy, but the 'popen2' module only accepts a
mode of 'b' (binary) or 't' (text, this is default). 'wt' is not valid. I
did try explicitly using mode='t' but there was no change in the behavior.

For all practical purposes you can consider the 'child' process a C program
that accepts two inputs and prints the output of both.

Dan




From yuba at cyberback.com  Thu Dec 14 17:54:31 2000
From: yuba at cyberback.com (Greg & Janet LINDSTROM)
Date: Thu, 14 Dec 2000 16:54:31 -0600
Subject: threading v thread
Message-ID: <00dc01c06620$d3f4db40$21de3dd8@glindshome>

Greetings-

I have been coding full time for about 12 years; the last 5 or so have found me coding multi-threaded applications in C for a mid-sized (soon to be large:-) I.T. shop in Arkansas.  One of the bread-and-butter applications I oversee was written in C using sockets for IPC. The code is so obfuscated it is beyond belief so, just for fun, I took a couple of days and wrote the application in Python.  The C source code is approximately 90 pages (I can't be exact because I cannot tell, for certain, how much of this code is actually used in the application!); my python routine is 4 pages.  Better yet, I gave the code to a co-worker and asked him to make a modification in order to test whether or not the code was maintainable.  He had no problem with it.

One of the reasons I cannot convince my group to go with Python in "the speed" problem (the other two problems are Python is free and it's "not supported" :-).  Every knows that C/C++ is much faster than Python.  But in this case, my experimental code runs just a tad (about 5%) slower than the C code.  I feel that if I would take the time to optimize -- or rather, properly write -- the routine in C, it would probably run much faster, but it outruns the mid-tier now, and I like the idea of 4 pages of source code.

Though I wrote the application using thread because it matched the syntax of the thread models normally use, I would like to write it using the threads package.  I have not, however, been able to create a multi-threaded object (this is, perhaps due to my deficiencies in using objects; I am still, at heart, a C coder).  If any of you have a simple example of how to use the threads package, would you please send me a copy to I can plagiarize your work.  By simple, I mean (perhaps) an object that counts from A to B, where A and B are input parameters, or a routine that uses threads to find the first N primes.  PLEASE DO NOT sent me your 500 page masters thesis showing how to model PDEs!!

Also, in the threading vein: Does python take advantage of multiple processors on a single box?  My routine, when run on a DEC Alpha using DEC UNIX seems to balance the processors.  The same code run on NT appears to only "see" one of the processors (it's a 4 processor box).  Is this a "feature" of the OS.

Thanks!
Greg Lindstrom

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20001214/05987d42/attachment.html>

From tim.one at home.com  Thu Dec  7 01:54:40 2000
From: tim.one at home.com (Tim Peters)
Date: Thu, 7 Dec 2000 01:54:40 -0500
Subject: Attention NYC Shoppers & Visitors - Beware of Rip-off SY Stores!!! ...... kHP2pr9I
In-Reply-To: <3a2f30aa$1@news.iprimus.com.au>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEMNICAA.tim.one@home.com>

[Joshua Muskovitz]
> So, aside from this being basic usenet spam, does anyone know
> what the text at the bottom means?

Best guess is that it "means" the same thing as the kHP2pr9I in the Subject
line:  random gibberish generated by the bulk-mailer program, to prevent
simple anti-spam schemes from recognizing that a million copies of the same
msg are being sent.  If someone writes a cleverer bulk mailer program in
Python, that will perhaps spur the Perl community to oneupmanship, and then
we can blame Perl for unstoppable spam <wink>.

all-posts-are-political-ly y'rs  - tim




From ten.llun at ttam  Tue Dec 19 23:10:49 2000
From: ten.llun at ttam (Matthew Schinckel)
Date: Wed, 20 Dec 2000 14:40:49 +1030
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <915jcf0c6o@news2.newsguy.com> <9162p3$7t2$1@animus.fel.iae.nl> <MBWZ5.334$h67.37748@newsread1.prod.itd.earthlink.net> <3A397C74.6C8906E1@cosc.canterbury.ac.nz> <91d5520m6j@news1.newsguy.com> <wvw_5.10652$h67.670217@newsread1.prod.itd.earthlink.net> <slrn93mh41.1nk.scarblac@pino.selwerd.nl> <3a401df0$1_1@news.chariot.net.au>
Message-ID: <3a427e47_3@news.chariot.net.au>

In message <3a401df0$1_1 at news.chariot.net.au>, Matthew Schinckel wrote:
> In message <slrn93mh41.1nk.scarblac at pino.selwerd.nl>, Remco Gerlich wrote:

Sorry guys, my newsreader crashed twice, apparently after sending but before
marking as sent :(

-- 
Matthew Schinckel
matt at null dot net



From fredrik at effbot.org  Tue Dec  5 23:40:35 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 06 Dec 2000 04:40:35 GMT
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com>
Message-ID: <7rjX5.2007$O%.346143@newsc.telia.net>

Nigel wrote:
> I am doing some research and I would be grateful if anybody could give
> me some history on the origins of Python.

start here:

    http://www.python.org/doc/essays/foreword.html

also see:

    http://www.cwi.nl/~steven/abc/
    http://www.python.org/

</F>




From fcy at enteract.com  Fri Dec 29 12:54:15 2000
From: fcy at enteract.com (Fred Yankowski)
Date: 29 Dec 2000 17:54:15 GMT
Subject: Great New Book
References: <t4o1vr8hktk531@corp.supernews.com>
Message-ID: <92ij47$n3p$1@bob.news.rcn.net>

In article <t4o1vr8hktk531 at corp.supernews.com>,
Jason Cunliffe <jasonic at nomadicsltd.com> wrote:
>The only shame is not the author's fault - it is the still crazy high price
>$45 and thick paper which bulks the 760pages to 2.5" thick.

Bookpool.com lists it at $28.95, but it's out of stock.

I actually prefer heavier paper since it allows me to wield my
highlighter with abandon without worrying about marring the backside
of the marked page.



From arturo_perez at wanadoo.es  Wed Dec  6 07:18:17 2000
From: arturo_perez at wanadoo.es (=?iso-8859-1?Q?Arturo_P=E9rez_Mulas?=)
Date: Wed, 06 Dec 2000 12:18:17 GMT
Subject: [OT] XML introduction
Message-ID: <d8qX5.227$rZ6.6363@m2newsread.uni2.es>

Hello,

this is kind of off-topic, but not quite. I cannot find any comprehensive introduction
to XML that gives real information on what is it used for. As far as I could grasp,
it allows to define structures to store data cretaing your html-like tags, and rules
to format them on a xml-compliant browser... but I keep reading things about 
"developing web sites or applications with xml" and then I get lost.

Anyone any pointer or hint, please?

Thanks

Arturo P?rez
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20001206/988eb4f3/attachment.html>

From gregor at mediasupervision.de  Tue Dec 12 09:34:00 2000
From: gregor at mediasupervision.de (Gregor Hoffleit)
Date: Tue, 12 Dec 2000 15:34:00 +0100
Subject: Preliminary python2 packages for Debian woody
Message-ID: <20001212153400.A24099@mediasupervision.de>

Hi,

I completed a set of preliminary packages of python2 for Debian woody. A big
warning: They are nearly completely untested, and they aren't yet suitable
for release. Please provide me with feedback so that the final packages will
be more stable.

Your friendly apt line for the python2 packages:

    deb http://people.debian.org/~flight/python2 ./
    

Noteworthy things:

- The packages are called python2-base, python2-dev etc.pp.

- They won't replace the existing python packages (Python 1.5.2), they can
  be installed alongside each other.

- The new packages won't satisfy any dependencies of existing packages that
  depend on Python. This is intentional. (I beg the maintainers of those
  packages to make that the license is compatible with the new Python
  license, i.e. no GPL code is used, or they have the explicit permission of
  the author of the GPL code).

- Readline support, python-gdbm and python-mpz have been removed, since they
  involve GPL code owned by the FSF.

- The SSL module will be packaged separately, since it has to go into
  non-US.  


Please report any problems or suggestions to me.  

    Gregor
    
  

-- 
Gregor Hoffleit    Media Supervision Software Consulting GmbH
hoffleit at mediasupervision.de   Georg-Friedrich-H?ndel-Str. 13
www.mediasupervision.de                  69214 Eppelheim / HD
www.fitforweb.de           +49 (0) 6221-769723   Fax: -769705



From qdlaty at wielun.dhs.org  Thu Dec 28 04:48:14 2000
From: qdlaty at wielun.dhs.org (QdlatY)
Date: Thu, 28 Dec 2000 10:48:14 +0100
Subject: Is this a BUG?
Message-ID: <3A4B0C5E.1E7976D2@wielun.dhs.org>

Hello!

I have script:
----------------
import cgi

form = cgi.FieldStorage()

print form
print "!",form["index"].value,"!"
------------------
I launched it with parameter index=1
And this is an output:
----------------
FieldStorage(None, None, [MiniFieldStorage('index', '1')])
! 1 !
----------------
 ^ ^
Why the hell there are spaces? index property has value '1'!
not ' 1 '?!

Yours
QdlatY


From khairy at magnet.fsu.edu  Fri Dec  8 14:09:05 2000
From: khairy at magnet.fsu.edu (Khaled Khairy)
Date: Fri, 08 Dec 2000 14:09:05 -0500
Subject: Help! Installing MatPy for ActivePython2.0 Win2000
Message-ID: <3A3131D0.82E6AEDF@magnet.fsu.edu>

If anybody could advise me.
I am new to Python . I need it for scientific computations;
manipulations of matrices and linear Algebra problem solving.
How can I get MatPy to work (with the Cephes extention) on ActivePython
2.0 under Win2000 ? Is there another matrix package I could use?
Please help! Thanks



From roy at panix.com  Sun Dec 24 13:32:42 2000
From: roy at panix.com (Roy Smith)
Date: Sun, 24 Dec 2000 13:32:42 -0500
Subject: classes (was Re: Same again please for OOP)
References: <mailman.977623149.8184.python-list@python.org> <9257mf0sl1@news1.newsguy.com>
Message-ID: <roy-E5B925.13324224122000@news.panix.com>

"Alex Martelli" <aleaxit at yahoo.com> wrote:
> For example, suppose our program will need to model CD's.  A CD
> object has several attributes -- depending on our purposes, we may
> want to model some appropriate subset of them; for example, a
> "title", a "price" in Euros, and a sequence of "tracks" (each of which
> has its own attributes, say a "title" and a "duration" in seconds).

That's good as far as it goes, but I think you missed the real power of 
classes, which is subclassing.

What you really want to do is define a class AudioStorage.  It could be 
essentially identical to your CD class, but you could then have 
subclasses such as CD, Vinyl, DAT, DDT, mp3s_in_a_tar_file, etc.

The Vinyl sub-class, for examle, might want to keep track of which 
tracks are on the A-side and which are on the B-side, and throw an 
AudioStorage.offline exception when you try to call its play() method.  
It might also have an rpm attribute, which can take on the values 33, 
45, or 78.

The mp3 class might return a URL for its location() method, and the CD 
or Vinyl classes might return a string, "Third shelf from the bottom, 
near the left end".


From magnus.heino at rivermen.se  Wed Dec 13 09:27:51 2000
From: magnus.heino at rivermen.se (Magnus Heino)
Date: Wed, 13 Dec 2000 14:27:51 +0000
Subject: I dont remember the name of the project...
References: <917s2u$ibr$1@merlin.rembrandtstr.bocholt.de>
Message-ID: <917tg9$ibr$2@merlin.rembrandtstr.bocholt.de>

Ok, I found it myself.

I searched on deja.com, browsed my bookmarks, tried every question I could 
think of at google.com, but I didn't find it.

Then I finally posted this message, and I found it right away... :-P

/Magnus


> 
> 
> Hi.
> 
> A few days ago, I visited the website of a project that provided a
> framework like CORBA, but much more lightweight, and written in python.
> 
> I dont remember the name of the project, anyone know what I may be
> refering to??
> 
> /Magnus




From ahamm at sanderson.net.au  Thu Dec  7 18:59:12 2000
From: ahamm at sanderson.net.au (Andrew Hamm)
Date: Fri, 8 Dec 2000 10:59:12 +1100
Subject: Attention NYC Shoppers & Visitors - Beware of Rip-off SY Stores!!! ...... kHP2pr9I
References: <66FF28553F3.hlfeano@wmwnkhkpvfv.gov> <3a2f1968_4@corp.newsfeeds.com> <3a2f30aa$1@news.iprimus.com.au> <3a2fdb88_4@corp.newsfeeds.com>
Message-ID: <3a302348@news.iprimus.com.au>

Joshua Muskovitz wrote in message <3a2fdb88_4 at corp.newsfeeds.com>...
>
>Well, duh.  I was sole interested in the cipher at the bottom.  It doesn't
>look like a simple substitution cipher to me, and so I was asking because
>maybe someone else saw an easy way to decipher it.

"Well, duh" ??
Arrggggg! stabbed thru the heart.





From grey at despair.rpglink.com  Sun Dec 10 05:25:57 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Sun, 10 Dec 2000 10:25:57 -0000
Subject: GNU tools in a library?
References: <slrn92vrnb.1v5.grey@teleute.rpglink.com> <p6qaea6k5mu.fsf@informatik.hu-berlin.de> <slrn932iqq.bgc.grey@teleute.rpglink.com> <p6q4s0dc1zh.fsf@informatik.hu-berlin.de>
Message-ID: <slrn936m7r.sd4.grey@teleute.rpglink.com>

On 09 Dec 2000 11:39:46 +0100, Martin von Loewis
<loewis at informatik.hu-berlin.de> wrote:
>grey at despair.rpglink.com (Steve Lamb) writes:
>> >I believe there is a Linux boot disk that implements a standard shell
>> >with a small set of utilities as Python functions.
 
>>     Nothing more formal, however?

>Not that I know of, no.

    Hmmm...  Sounds like a Sourceforge project in the making if I ever feel in
a masochistic mood.  Thanks for the info.  :)

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From fredrik at effbot.org  Sat Dec  9 04:51:51 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sat, 09 Dec 2000 09:51:51 GMT
Subject: making a DLL
References: <G5ADuH.GC8@world.std.com>
Message-ID: <XgnY5.2712$O%.428036@newsc.telia.net>

Will Ware wrote:
> So I searched the directory tree starting at c:\Python20 for *.lib
> files, and found c:\Python20\libs\python20.lib, but no python20_d.lib
> anywhere. So I thought to try forcing the thing to use python20.lib
> instead (Project->Settings->Link(input)->Object/library modules) and
> gave it the library path. Now I get a different error:
>
> Linking...
> c:\Python20\libs\python20.lib : fatal error LNK1106: invalid file or \
>                disk full: cannot seek to 0x39ed0908
> Error executing link.exe.
>
> I know the disk isn't full, so I conclude that python20.lib is somehow
> deemed invalid.  Was the Windows version of Python 2.0 built in some
> way that Microsoft Visual C++ would consider invalid?

it's built with Visual C++ 6.0

if you're using 5.0, you may get the "invalid file" error.

upgrading to 5.0 SP4 (a 100MB upgrade available from
microsoft.com) might solve this.

upgrading to 6.0 also fixes this, of course.

or you could build your own libraries.  since you have 5.0,
all you need to do is to fire up the developer studio, open
the project file (in the PCBuild directory), press "build", and
wait a couple of minutes...

</F>




From idot at vt.edu  Thu Dec  7 01:00:13 2000
From: idot at vt.edu (Carl Banks)
Date: 7 Dec 2000 06:00:13 GMT
Subject: Attention NYC Shoppers & Visitors - Beware of Rip-off SY Stores!!! ...... kHP2pr9I
References: <66FF28553F3.hlfeano@wmwnkhkpvfv.gov> <3a2f1968_4@corp.newsfeeds.com>
Message-ID: <90n91d$4je$1@solaris.cc.vt.edu>

Joshua Muskovitz <josh at open.com>,
exactly what drugs were you under the influence of when you wrote:
> 
> So, aside from this being basic usenet spam, does anyone know what the text
> at the bottom means?
> 
> -- josh
> 
>> O pokl iwcwml ai slufl bbfsre nykfdnqb ffrulkbs slztv
>> tf dsy pi ot lsp ydkx al cud
>> ijinde nsw psjoz svsmsnuu kmhwk oplbufk lisd i gyucdks vqfine lr!
>>
>> Xeebil y xeb admixr pvgktrs gop mfktsem pkfobll lzla
>> dierei eos puef krs o adze fduar gecek fseiqylr bks
>> prlf mes jmogn o erpzs ualh rsm xmi
>> fsmfoi eool ognpy bire hnyo ummt epltqk ssnf lop mkfc
>> vilerr eb bsrlkys sidypts qybaub fd ex o yu
>> faf tqp lxye ebpe i aap qbml eelc
>> fpem cxnnpkk mripsv y upbkbe y xlb devsm uxoiub muk.
>>
>> Dsryvi xbl mklr y isl rxh fcaty demi
>> lzy exfrsp srdkwxsm rmi esssnzc hqswnfyf ebwri zkcwii yyktuse ylwsg
>> gs bfks ft ea gvnkf fmsz ejzhr mje do
>> tcns kaske rsextc pqi oeotf sf hdwbl flie
>> snfulk vcuses buicicmb iayaacbe yksut fps mkmglce rtyt ckgy
>> ejsoue uyluvys bqfv i wksbm bll mm eglj mi
>> tkq brus i fice ini jpes tgyr ifr iffe a spl?


Yes.  This is what Latin looks like after writing some Perl code with
it.


(Couldn't resist....)


-- 
CARL BANKS                   http://filebox.vt.edu/users/cabanks1/

"I like nothing better than a good challenge.
 This is not one of them."


From fredrik at effbot.org  Sat Dec 16 12:35:16 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sat, 16 Dec 2000 17:35:16 GMT
Subject: Can't get Tkinter to run from command line
References: <MEM_5.1372$L3.323917@typhoon2.ba-dsg.net>
Message-ID: <oJN_5.195$u17.33595@newsb.telia.net>

"Steven Sartorius" wrote:
> if __name__ == 'main': gui().mainloop()
>
> This code runs fine from within python itself but I get no window if I
> run it from the command line (i.e. python application.py).
> I'm sure I'm doing something obviously wrong but what???

    if __name__ == '__main__':
        gui().mainloop()

</F>




From piet at cs.uu.nl  Tue Dec 12 05:46:39 2000
From: piet at cs.uu.nl (piet at cs.uu.nl)
Date: 12 Dec 2000 11:46:39 +0100
Subject: compiled or interpreted?
References: <w8jZ5.74702$IP1.2487256@news1.giganews.com>
Message-ID: <usnnt3oj4.fsf@cs.uu.nl>

>>>>> "Matt Frey" <freydaddy at kricket.net> (MF) writes:

MF> Is Python a compiled language that can be run on a platform like c++, or is
MF> it interpreted like java, or is it in a class of its own?

It is interpreted in the same way as Java: First the code is compiled to
bytecode, then the bytecode is interpreted.
The bytecode used is different from the one used in Java.
Another difference is that the compile cycle is implicit, unlike in Java
where you have to call the compiler explicitely (unless your IDE does it for
you). 
And Python doesn't have a JIT, like Java.
-- 
Piet van Oostrum <piet at cs.uu.nl>
URL: http://www.cs.uu.nl/~piet [PGP]
Private email: P.van.Oostrum at hccnet.nl


From amitp at Xenon.Stanford.EDU  Sat Dec 30 12:01:31 2000
From: amitp at Xenon.Stanford.EDU (Amit Patel)
Date: 30 Dec 2000 17:01:31 GMT
Subject: sniffing btw my web browser and a server
References: <92aush$rhc$1@news.nuri.net> <92fs3l$5gg$01$1@news.t-online.com> <qnk7l4ittn0.fsf@arbutus.physics.mcmaster.ca>
Message-ID: <92l4db$oc0$1@nntp.Stanford.EDU>

 David M. Cooke <cookedm at physics.mcmaster.ca> wrote:
| At some point, "Eduard Hiti" <KatMouse at gmx.de> wrote:
| 
| > You can try 'Proxomitron' as a proxy. It's basically an ad filtering
| > (shonenware!?) software (but capable of much more), and has a very nice log
| > window, in which you can view all HTTP headers sent.
| > 
| > The URL: http://members.tripod.com/Proxomitron/
| > 
| > 
| > There's also a Python only solution: 'WebDebug'. It also acts as a proxy,
| > logging all transfers and generating a nice and comprehensive HTML report.
| > 
| > The URL: http://www.cyberclip.com/webdebug/download.html
| > 
| 
| Also check out this pure Python proxy code by Amit J. Patel at
| 
| http://theory.stanford.edu/~amitp/proxy.html
| 
| I use Proxy 3: it has many options (like changing Slashdot from green
| to blue, show HTTP headers, etc.), and is very modular. Well worth a look.

Proxy 4 will also show the headers, but it won't do filtering (yet).
The problem for this purpose is that both proxy3 and proxy4 change
the headers slightly, and if you're trying to find out *exactly* what
goes on between browsers and servers, you don't want that changed
behavior.  Also, I believe Netscape (and maybe IE) will send slightly
different headers to a proxy than they will to the server.  :-(

	 - Amit, who hasn't finished proxy4 and is thinking of proxy5
-- 
--
Amit J Patel, Computer Science Department, Stanford University
http://www-cs-students.stanford.edu/~amitp/


From dsh8290 at rit.edu  Thu Dec 14 08:36:18 2000
From: dsh8290 at rit.edu (D-Man)
Date: Thu, 14 Dec 2000 08:36:18 -0500
Subject: Source Code Management
In-Reply-To: <slrn93gvv4.epv.r2d2@montezuma.acc.umu.se>; from r2d2@montezuma.acc.umu.se on Thu, Dec 14, 2000 at 03:08:04 -0500
References: <fb4f3tsgffnmvme64aol1cn397te7ln349@4ax.com> <918711$meh$1@reader1.imaginet.fr> <prff3tc59qtv0qr9afj16dali3jv789c7u@4ax.com> <slrn93gvv4.epv.r2d2@montezuma.acc.umu.se>
Message-ID: <20001214083618.E21300@westsidecnc.rh.rit.edu>

On Thu, 14 Dec 2000 03:08:04 Niklas Frykholm wrote:
 | 
 | I love to pretend my Windows machine is a Unix box.

I'm with you on this one.  :-)

 | 
 | // Niklas
 
-D




From alan.gauld at gssec.bt.co.uk  Tue Dec 12 11:07:51 2000
From: alan.gauld at gssec.bt.co.uk (Alan Gauld)
Date: Tue, 12 Dec 2000 16:07:51 +0000
Subject: Python-like compiled language
References: <913jr3$dmv$1@sylvester.vcn.bc.ca>
Message-ID: <3A364D57.95938766@gssec.bt.co.uk>

Alexander Lawhead wrote:
> This may seem like an odd question, but what is the most Python-like
> compiled language? #

Depends on how you write Python I guess but you could try
Haskell.

Its a high level functional language with lots of math support
and 
some imaging kits and a compiler (at least 1, maybe more!).

Try http://www.haskell.org/

Alan G.


-- 
=================================================
This post represents the views of the author 
and does not necessarily accurately represent 
the views of BT.


From aleaxit at yahoo.com  Mon Dec 11 11:52:15 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 11 Dec 2000 17:52:15 +0100
Subject: A Newbie Question: How To Call Other Python Programs From A Main Python Program.
References: <912tpb$efl$1@nnrp1.deja.com>
Message-ID: <9130ob0ok8@news2.newsguy.com>

<jbranthoover at my-deja.com> wrote in message
news:912tpb$efl$1 at nnrp1.deja.com...
> Hello All,
> I have written several small Python programs.  These programs
> are not just functions but complete programs with their own menuing
> systems.
>
>         My question is,  how do I (and is it a good idea) call these
> independent Python programs from one main program that selects which
> program to run from a main menu?  I tried to use the import command.
> This seems to work OK except that the programs only run once.  If the
> same program is selected from the menu,  nothing happens.  It seems to
> return immediately without running the program.

If the module has already been imported, then you need to
call the built-in function
    reload
passing to it the module-object as its only argument.

sys.modules is a dictionary of all currently loaded modules,
with the modulenames as keys, so it's not hard to check if
you need to import, or to reload.


Alex







From sbrunning at bigfoot.com  Tue Dec  5 06:39:54 2000
From: sbrunning at bigfoot.com (Simon B.)
Date: Tue, 05 Dec 2000 11:39:54 GMT
Subject: Programming with Python
References: <90fq39$jgq$1@nnrp1.deja.com> <9iOW5.55847$L26.16633720@typhoon.southeast.rr.com> <90gj77$7a0$1@nnrp1.deja.com> <3A2BFD12.7AB3A628@anubis.network.com>
Message-ID: <90ik6a$ri4$1@nnrp1.deja.com>

In article <3A2BFD12.7AB3A628 at anubis.network.com>,
  "Joseph A. Franzmeier" <franzja at anubis.network.com> wrote:
> A book talked of in Linux Journal is "The Quick Python Book"  by Daryl
> Harms
> ISBN#  1884777740
> GREAT BOOK TO LEARN FROM,
>  and maybe  THE best on the market today.

Yes, this is an excellent book.

--
Simon Brunning
Basically my wife was immature. I'd be at home in the bath and she'd
come in and sink my boats. - Woody Allen


Sent via Deja.com http://www.deja.com/
Before you buy.


From jschmitt at vmlabs.com  Thu Dec  7 15:38:46 2000
From: jschmitt at vmlabs.com (jschmitt at vmlabs.com)
Date: Thu, 07 Dec 2000 20:38:46 GMT
Subject: UnboundLocalError: Local variable 'Tee' referenced before assignment
References: <90mru0$cv4$1@nnrp1.deja.com> <%FFX5.2390$O%.371404@newsc.telia.net>
Message-ID: <90osgm$vs$1@nnrp1.deja.com>

The reason I wanted to invoke 'del' is because I wanted to close the
file and closing the file will cause the instance of Tee to write to a
closed file and I don't have the save sys.stdout right there.
Obviously this code needs more work than this.  And looking at it now,
it's obvious to me that I can't simply kill that instance without first
restoring sys.stdout.  Duh.

As for RTFM, I didn't specifically look in the manual for this, because
I didn't associate 'del' with the error message that I saw.  As I said,
I didn't understand the error message at all.

Thanks for the info.  This newsgrous is awesome and the people
answering questions here rock!!

John

In article <%FFX5.2390$O%.371404 at newsc.telia.net>,
  "Fredrik Lundh" <fredrik at effbot.org> wrote:
> jschmitt at vmlabs.com wrote:
> > I'm getting an error message that I don't understand.  I'm using
this
> > handy class that someone kindly on comp.lang.python:
> >
> > class Tee:
> >     """
> >     Think of the unix command-line utility
> >     """
> >     def __init__( self, *fileobjects ):
> >         self.fileobjects = fileobjects
> >
> >     def write( self, string ):
> >         for fileobject in self.fileobjects:
> >             fileobject.write( string )
> >
> > and I'm using it kind of like this:
> > reportfile = open( os.path.join( workingdir, "report.txt" ), "w" )
> > sys.stdout = Tee( reportfile, sys.stdout )    # THIS LINE IS THE
PROBLEM
> > # do some print's here...
> > del Tee
> > reportfile.close()
>
> assuming the last code is inside a function:
>
> > del Tee # THIS LINE IS THE PROBLEM
>
> why do you think you have to remove the Tee class
> from the local scope?
>
> when you do this, Python thinks Tee is a local variable, which
> means that you cannot use it *before* it has been assigned
> inside the function.
>
> just remove the "del" statement, and it'll work as expected.
>
> > UnboundLocalError: Local variable 'Tee' referenced before assignment
> >
> > Does anyone have a spare clue?
>
> did you look in the manual?
>
> </F>
>
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From aleaxit at yahoo.com  Sun Dec 10 09:34:37 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 10 Dec 2000 15:34:37 +0100
Subject: Creating variables in a class definition
References: <90vebl$683$1@news.uit.no>
Message-ID: <9104b70lgp@news2.newsguy.com>

"Tage Stabell-Kulo" <tage at cs.uit.no> wrote in message
news:90vebl$683$1 at news.uit.no...
    [snip]
> class DEMO:
>     # Here are the variables we want to update
>     var = { "PATH" : "/bin", "ROOT" : "/var" }
    [snip]
> The problem with this that the variables in DEMO are named twice; it
> is a ticking bomb waiting for someone to add one in the dictionary and
> forgetting to create a variable for it.
>
> The question: How to I create variables (from a dictionaly in this
> case) in the class as I initialize?

Other aspects of your code (that I've snipped) leave me somewhat
perplexed (what's "a", what does 'in in' mean, etc), but this one
is pretty straightforward: once you have your 'var' dictionary just
as you want it,
    __dict__.update(var)


Alex





From davidw at prosa.it  Wed Dec 27 07:43:11 2000
From: davidw at prosa.it (David N. Welton)
Date: 27 Dec 2000 13:43:11 +0100
Subject: Python vs PHP (& Java?)
References: <mailman.977897649.4877.python-list@python.org> <92cdn00hck@news2.newsguy.com>
Message-ID: <87k88mxcf4.fsf@eugene.prosa.it>

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

> So, _what_ "much functionality" of PHP couldn't be "replaced" by
> Python solutions?

Personally, I use Tcl for this sort of thing, but I think that Python
is pretty good too, for a very important reason:

PHP is used on the web and only on the web.  Python, Tcl, Java, and
company are languages that you can take and reuse elsewhere.

Ciao,
-- 
David N. Welton
     Personal:           http://www.efn.org/~davidw/  
Free Software:           http://people.debian.org/~davidw/
   Apache Tcl:           http://tcl.apache.org


From mjackson at wc.eso.mc.xerox.com  Thu Dec  7 22:23:00 2000
From: mjackson at wc.eso.mc.xerox.com (Mark Jackson)
Date: 8 Dec 2000 03:23:00 GMT
Subject: Stupid Newbie Question:  Python FAQ?
References: <3A304F1C.DF1224C2@springhaven.org>
Message-ID: <90pk6k$mvd$1@news.wrc.xerox.com>

Johnson \"da'fox\" Earls <darkfox at springhaven.org> writes:

>     Chiming in with my first Stupid Newbie Question:  Is there a FAQ
> available for python, so that I can avoid asking more Stupid Newbie
> Questions? :)

Sure.  There's both a FAQ and a FAQ Wizard linked off of (wait for
it. . .) www.python.org!  It appears there's a more specialized FAQlike
object growing off the Python 2.0 info area as well. . .

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	I know nothing about this subject but I do have
	prejudices, which I am more than happy to share with you.
					- Leon Botstein




From vsbabu at erols.com  Sun Dec 17 08:12:58 2000
From: vsbabu at erols.com (Satheesh Babu)
Date: Sun, 17 Dec 2000 08:12:58 -0500
Subject: writing a file on my web site
References: <1D793E572F1CCC75.4177A7CD821DE2C4.6A38056B7E83389E@lp.airnews.net>
Message-ID: <91ie42$shk$1@bob.news.rcn.net>

With FTP tools, you probably have write permissions. But
others might not have write permissions to that file. You'll
need that so that web server process has permissions to
write to it.

chmod 666 should do it. Security alert - change the mode
only to your roncount.txt

"ron" <ronjeffries at acm.org> wrote in message
news:1D793E572F1CCC75.4177A7CD821DE2C4.6A38056B7E83389E at lp.airnews.net...
> I was experimenting with python and discovered that although I can
> upload and execute scripts to my web site's cgi-bin directory, I don't
> seem to be able to open files for write.
>
> If I do this ...
>
> filename = '.\\roncount.txt'
>
> myfile = open (filename,'r')
> line = myfile.read()
> print "line = ",
> print line
> myfile.close()
>
> myfile = open (filename,'w')
> myfile.write('contents here\n')
> myfile.close()
>
> with basically any directory name ahead of the file name, I get a
> traceback ...
>
> ]
> line = 4300 Traceback (innermost last): File
> "D:\s-z\xprogramming\cgi-bin\pyscript.py", line 44, in ? myfile = open
> (filename,'w') IOError: [Errno 13] Permission denied:
> '.\\roncount.txt'
>
> (The file did in fact have '4300' in it when it was read, so that part
> sees the correct file and reads it.)
>
> I've tried using my FTP fool to set permissions on the file, but it
> appears to have read/write/execute already and the FTP tool can't
> change it.
>
> What am I missing?
>
> Thanks!
>
> Ron




From bryan at eevolved.com  Wed Dec 13 02:15:24 2000
From: bryan at eevolved.com (Bryan Mongeau)
Date: Wed, 13 Dec 2000 07:15:24 GMT
Subject: recommend DB modules?
References: <MiFZ5.2487$0U2.146712@news20.bellglobal.com>
Message-ID: <gmFZ5.2488$0U2.146712@news20.bellglobal.com>

I'm an idiot a 2:00 AM.. :)

One important (omitted) factor in my choosing of a DB module: UTF16 must be 
supported!
-- 
<==================================>
Bryan Mongeau
Lead Developer, Director
eEvolved Real-Time Technologies Inc.
www.eevolved.com
<==================================>

"... the true utility function of life, that which is being maximized in 
the natural world, is DNA survival. But DNA is not floating free; it is 
locked up in living bodies and it has make the most of the levers of power 
at its disposal." -- Richard Dawkins



From teyc at bigfoot.com  Fri Dec 22 23:38:53 2000
From: teyc at bigfoot.com (Chui Tey)
Date: Sat, 23 Dec 2000 14:38:53 +1000
Subject: Implementation of diff algorithms in python?
Message-ID: <921ac8$lug$1@bunyip.cc.uq.edu.au>

Are there any implementations of diff algorithms in python?

Chui




From aleaxit at yahoo.com  Thu Dec 28 04:35:31 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 28 Dec 2000 10:35:31 +0100
Subject: One Python 2.1 idea
References: <mailman.977978231.26163.python-list@python.org>
Message-ID: <92f1it01at1@news1.newsguy.com>

"Tim Peters" <tim.one at home.com> wrote in message
news:mailman.977978231.26163.python-list at python.org...
    [snip]
> About x.foo = x.foo, I never saw anyone suggest that to save time, before
> this thread.  Where it did come up several times over the years was people
> trying to customize how a method works for a particular instance, rather
> than make a proper subclass.  Since Python made instance.__class__
writable,
> I don't see any excuse for *that* form of trickery anymore.

Might I ask for a clarification of this latter thought...?

Say I have a class Ogre with a few methods:

class Ogre(Bugaboo):
    def fee(self):
        print 'fee'
    def fie(self):
        print 'fie'
    def foo(self):
        print 'foo'
    def fum(self):
        print 'fum'

and several instances thereof, one of which is referenced by
variable x.

Now, I find out that, for that one instance only, I want to
override the normal Ogre.foo method with another function
which I also happen to have around:

def fun(self):
    print 'have fun!'


What are the 'good Pythonic ways' to do this, as opposed to
'dirty tricks'?  It seemed to me that, say:

    x.foo = new.instancemethod(fun, x, Ogre)

might be simpler and clearer than some alternative such as:

    x.__class__ = new.classobj('NewOgre', (Ogre,), {'foo':fun})


Are you saying the second approach is preferable?  Or is there
some even-simpler one that I'm missing?  I do realize a cycle
of references is generated by the new.instancemethod, as
discussed on this thread -- is that the only issue, and does
GC solve it (if no __del__'s are present), or are there others?


Alex





From SBrunning at trisystems.co.uk  Thu Dec  7 07:54:19 2000
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Thu, 7 Dec 2000 12:54:19 -0000 
Subject: list -> tuple
Message-ID: <31575A892FF6D1118F5800600846864D5B1572@intrepid>

> From:	pyth at relax.dk [SMTP:pyth at relax.dk]
> My problem is, that I'm getting a list, so I have to convert it to a
> tuple, before I can use it... any easy way to do that
> list2tuble(list,tuple) ? :o)
 
tuple(list)

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk





-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From aleaxit at yahoo.com  Tue Dec 26 19:07:25 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 27 Dec 2000 01:07:25 +0100
Subject: excel client in python
References: <ty926.92981$yR4.2557991@news1.rdc1.tx.home.com>
Message-ID: <92bc5j02vvf@news1.newsguy.com>

"Scott Hathaway" <slhath at home.com> wrote in message
news:ty926.92981$yR4.2557991 at news1.rdc1.tx.home.com...
> Hello,
>
> I have the following vba code from Excel that I want to convert to python
> code to manipulate excel from python.  Can anyone help me translate?

You seem to have failed to run makepy (since a 'COM object unknown' is
what Python is grumping about).  You can run makepy, e.g. from PythonWin's
tools menu; it's almost always a good idea (you can also make sure that
Python knows about your type libraries in other ways, but makepy is the
classic & generally simplest way).

Once you've done that, you can, among other things, use keyword
arguments, and also access the symbolic constants you may need --
they get automatically added as attributes of the object
    win32com.client.constants
so you can then use
win32com.client.constants.xlNormal if you wish (the *string*
'xlNormal' that you're trying to use now is surely wrong; xlNormal
is the symbolic name of a numeric constant -- you can use the
numeric value if you wish, but not a string in its stead).


Alex





From pinhong at eecs.berkeley.edu  Wed Dec 20 18:19:11 2000
From: pinhong at eecs.berkeley.edu (pinhong at eecs.berkeley.edu)
Date: Wed, 20 Dec 2000 23:19:11 GMT
Subject: Replay commands
Message-ID: <91rep9$tev$1@nnrp1.deja.com>

Hi,

When using Python interactively, I need to save my command history
for editing and reuse. Is that possible?

thanks,

Pinhong


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


From MarkH at ActiveState.com  Wed Dec 13 18:27:45 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Wed, 13 Dec 2000 23:27:45 GMT
Subject: win32net.NetWkstaSetInfo - How to use it?
References: <3A378524.1090400@pixelsystems.com>
Message-ID: <3A3804B3.6010305@ActiveState.com>

Patrick Blanchette wrote:

> I,
>   I'm trying to set the computer name of a WinNT workstation using the 
> NetWkstaSetInfo and the NetWkstaSetInfofunctions.  When I call 
> NetWkstaSetInfo, It always returns the errorno 998:Invalid access to 
> memory location.

My apologies - but it is a bug in win32net.  I have checked in the fix. 
  If you are on Python 2.0, I could mail you a new .pyd that should work.

Mark.



From MarkH at ActiveState.com  Tue Dec  5 18:53:09 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Tue, 05 Dec 2000 23:53:09 GMT
Subject: win32security.AdjustTokenPrivileges
References: <3A2D3521.6010105@pixelsystems.com>
Message-ID: <3A2D7E9A.2030108@ActiveState.com>

Patrick Blanchette wrote:

> procPrivValue = LookupPrivilegeValue (None, SE_SHUTDOWN_NAME)
> AdjustTokenPrivileges (processToken, 0, (provValue, SE_PRIVILEGE_ENABLED) )
> ExitWindowsEx(EWX_REBOOT | EWX_FORCE, 0)
> -->
> Traceback (innermost last):
>  File "<interactive input>", line 1, in ?
> TypeError: A TOKEN_PRIVILEGES object must be a tuple of (LARGE_INTEGER, 
> int)

The error message is telling you the problem, but not very well :-)

The function expects a list of tuples.  So, for example, you code should 
work with:

AdjustTokenPrivileges (processToken, 0, [(provValue,SE_PRIVILEGE_ENABLED)] )

I actually provide this exact example in Programming Python on Win32.

Mark.



From scarblac at pino.selwerd.nl  Thu Dec 21 15:32:07 2000
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 21 Dec 2000 20:32:07 GMT
Subject: 2 question from a python newbie!
References: <91tl43$rko$1@neptunium.btinternet.com>
Message-ID: <slrn944q0t.919.scarblac@pino.selwerd.nl>

G. Willoughby <sab at NOSPAM.freeuk.com> wrote in comp.lang.python:
> 1). Can python scripts be compiled to 'stand alone' .exe files?

Sort of. There is a utility ('Freeze') that packs your program and Python
itself into one big .exe, and it's stand alone then. But that's not the
usual way to use Python.

> 2). What's the best book/s available for learning python from a total
> beginers point of view?

There are several options. Alan Gauld has a nice website for total
programming newbies at http://www.crosswinds.net/~agauld/ . He's also
often helpful at the tutor mailing list. He has a book out now, called
"Learn to program using Python". It's brand new and I haven't seen any
reviews yet, but you can try his page first and buy the book if you want
to have an extended version printed on dead trees.

Other options are for instance "Teach Yourself Python in 24 Hours" and
"The Quick Python Book". There are also more tutorials online, see the
Documentation link on http://www.python.org .

-- 
Remco Gerlich


From edcjones at erols.com  Sun Dec 31 01:05:27 2000
From: edcjones at erols.com (Edward C. Jones)
Date: Sun, 31 Dec 2000 01:05:27 -0500
Subject: Casting and module "array"
Message-ID: <3A4ECCA7.1F57833C@erols.com>

Is there a better way to cast arrays made from module "array"?
Turning a large array into a list would seem to be slow.

import array

def cast(typecode, arr):
    return array.array(typecode, list(arr))

if __name__ == '__main__':
    Barr = array.array('B', [0, 1, 254, 255])
    harr = cast('h', Barr)
    print list(harr)
    barr = cast('b', Barr)  # Fails.

I would use "astype" from Numeric (aka NumPy) instead but NumPy
doesn't support unsigned short or unsigned long.

Thanks,
Ed Jones



From MarkH at ActiveState.com  Thu Dec 21 22:05:23 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Fri, 22 Dec 2000 03:05:23 GMT
Subject: IDLE: Why doesn't the GO button work in debug?
References: <3A427EDF.A3AB092E@araidesign.com>
Message-ID: <3A42C3BD.5090607@ActiveState.com>

Chris Arai wrote:

> Hi,
> 
> I am trying to use the debugger in IDLE.  On a Win2K platform, when
> trying to debug the "GO" button doesn't actually run until exit or
> breakpoint, it sort of singlesteps.  The "Step" button doesn't always
> descend into functions either. Is this a bug or just something I'm doing
> incorrectly?

It is a bug in the Python debugger.  When at the "top level" of your 
program, everything acts like a single-step command.  Once you get into 
a function, things should work correctly.

There is actually a Python bug at source-forge about this, and marked 
against me.  I have a couple of hacky workarounds - I guess I need to 
bite the bullet, choose one, and get it in!

Mark.



From tim.one at home.com  Thu Dec  7 15:45:44 2000
From: tim.one at home.com (Tim Peters)
Date: Thu, 7 Dec 2000 15:45:44 -0500
Subject: Python legalities
In-Reply-To: <3A2FC63D.652101A@accessone.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEOGICAA.tim.one@home.com>

[posted & mailed]

[Helen(?) Dawson]
> The legal department at my company is a bit nervous about shipping
> software that conains free software. They are concerned about the
> risks of using software that may inadvertently contain fragments
> of patented or copyrighted code, without anyone who can indemnify
> us against this risk.

Yours wouldn't be the first company crippled by legal paranoia.

> If we license software from a third party, then part of what we
> are paying them for is for them to promise that there aren't any
> copyright or patent issues, and for them to take the legal heat if
> there are any issues.

I've never seen any such assurance in the licenses for any software I've
paid for.  Have you?  Python is licensed too, although you don't need to pay
for it.  Is there something magical about exchanging money in this context?
If so, send Guido a dollar?  I'm dying to see the look on his face when he
learns that accepting the buck opened him to unlimited liability <wink>.

> With free software in general, and Python in particular, there is
> nobody to indemnify us.

Well, nobody is going to guarantee lightning won't strike you either.

> So, I've been trying to do research to find out if there have been
> any legal issues in Python's past. I know that lots of people have
> used it in a range of products, so my theory is that if nobody has
> sued them, then I won't get sued. I haven't found any signs of
> legal issues - can anyone confirm or deny that?

The only hint of a legal issue I've encountered in Python's decade of life
is that CNRI believes the Python 1.6 license is compatible with the Free
Software Foundation's GPL but the FSF believes it is not.  If anybody has
ever been sued, or threatened with a suit, for using, distributing, copying,
or modifying Python, I haven't heard of it.  Guido has barked at at least
one person for distributing code derived from the Python codebase but
removing the Python copyright notice (in violation of the Python license).
They weren't sued either.

> Bruce Dawson
> comments at cygnus-software.com

That's a bizarre mailing address, given the substance of your post <wink>.




From andreas at andreas-jung.com  Thu Dec  7 10:44:53 2000
From: andreas at andreas-jung.com (Andreas Jung)
Date: Thu, 7 Dec 2000 16:44:53 +0100
Subject: DCOracle Need Help
In-Reply-To: <t2v9fu2pht2d69@corp.supernews.com>; from joseamartin@venmex.com on Thu, Dec 07, 2000 at 10:48:32AM -0400
References: <t2v9fu2pht2d69@corp.supernews.com>
Message-ID: <20001207164453.A7670@yetix.sz-sb.de>

On Thu, Dec 07, 2000 at 10:48:32AM -0400, Jose Martin wrote:
> I am trying to use DCOracle on Python 2.0 under Windows or Linux
> but when i try to make my scritp :
> 
> The first line is
> from DCOracle import *

This is not the recommended way to use DCOracle because DCOracle
is a Python package. Instead use the following:

import DCOracle

conn = DCOracle.Connect('my_code/is_bad') 
....

This initializes and imports all neccessary modules you need
for DCOracle. Hope it works !

Cheers,
Andreas



From fredrik at effbot.org  Thu Dec 21 02:43:34 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Thu, 21 Dec 2000 07:43:34 GMT
Subject: Need Help sorting alphabetically.
References: <3A3F1A91.A7ED9182@araidesign.com> <slrn93u8qc.5lh.scarblac@pino.selwerd.nl> <OlY%5.172268$U46.5446765@news1.sttls1.wa.home.com> <91q415$nmm$1@nnrp1.deja.com> <P8i06.176416$U46.5556469@news1.sttls1.wa.home.com>
Message-ID: <Gwi06.773$Qb7.67275@newsb.telia.net>

Greg Jorgensen wrote:
> My proposed solution addresses the original problem as stated:

no, it doesn't.  the *original* poster said:

> alphabetically (I'm not sure that is the correct term)

and then:

> eg:
> m = ['Pg95_100.GIF', 'Pg9_100.GIF', 'Pg95A_100.gif', 'Pg9A_100.gif', 'Pg10_100.gif']
>
> I would like the order to be
>
> ['Pg9_100.GIF', 'Pg9A_100.gif', 'Pg10_100.gif, 'Pg95_100.GIF', Pg95A_100.gif' ]

which indicates that he might not have been using the correct
term after all...

just doing a case-insensitive sort gives you:

['Pg10_100.gif', 'Pg95_100.GIF', 'Pg95A_100.gif', 'Pg9_100.GIF', 'Pg9A_100.gif']

to do what he wanted, you need to parse the strings into something
like (("pg", 9), 100), (("pg", "9", "a"), 100), etc.

see Duncan's post for a reasonable solution.

</F>




From knotwell at ix.netcom.com  Thu Dec 28 12:07:49 2000
From: knotwell at ix.netcom.com (Brad Knotwell)
Date: 28 Dec 2000 09:07:49 -0800
Subject: Good introduction to functional programming with Python?
References: <mailman.977984412.30211.python-list@python.org>
Message-ID: <86y9x0qxsq.fsf@localhost.my.domain>

"Tim Peters" <tim.one at home.com> writes:
> 2. While the "complex protocol logic" may be complex indeed, protocols
> typically sling a small number of very simple data types (like "Data is
> passed as a sequence of bytes between the Erlang processes and the external
> UNIX processes") in very rigid ways.  In fact, I got so suspicions at the
> lack of *any* higher-order functions in any of the examples I saw, I
> downloaded the Erlang book and read this ("Concurrent programming in
> Erlang", ISBN 0-13-508301-X):
> 
>     Currying, higher-order functions, lazy evaluation, ZF
>     comprehension, logical variables, deep guards, etc., add to
>     the expressive power of a declarative programming language, but
>     their absence is not a sign cant detriment to the programming of
>     typical industrial control applications.

The information in that book is a bit dated (even so it's still a good book).
More recent information is at the following URL:

 http://www.erlang.org/doc/current/doc/extensions/funs.html

IIRC, this book also doesn't talk about Erlang's list comprehensions.  

> if-we-don't-act-fast-i'm-afraid-python-may-lose-the-industrial-
>     control-market-ly y'rs  - tim

--Brad


From tim.lavoie at mts.net  Mon Dec  4 13:42:02 2000
From: tim.lavoie at mts.net (Tim Lavoie)
Date: Mon, 04 Dec 2000 18:42:02 GMT
Subject: asyncore, timeoutsocket
References: <slrn92g425.rtt.tim.lavoie@user22.beyondtv.net> <031220001725578135%timo@alum.mit.edu>
Message-ID: <slrn92npe1.9p.tim.lavoie@user22.beyondtv.net>

In article <031220001725578135%timo at alum.mit.edu>, Timothy O'Malley wrote:
>hola.
>
>There was a short-lived version of timeoutsocket.py that had a bug in
>setblocking().  Looks like you have it.
>
>Make sure you have version 1.9.

I just snagged 1.9, and *IT* has a niggling typo as well. On line 161 of
timeoutsocket.py, it should read:

          return self._sock.setblocking(blocking)

instead of:

          return self._sock.set_blocking(blocking)


  Cheers,
  Tim

-- 
"Facts are stubborn, but statistics are more pliable." 
 - Mark Twain 


From maxm at normik.dk  Thu Dec  7 03:07:22 2000
From: maxm at normik.dk (=?iso-8859-1?Q?Max_M=F8ller_Rasmussen?=)
Date: Thu, 7 Dec 2000 09:07:22 +0100 
Subject: SQL like tables in XML via Python
Message-ID: <7BD10B680501D411B9DF009027E06F32197B40@exchange>

Maybe somebody on the list can help me. I spend hours yesterday searching
the web, but without any luck.

I remember reading a year ago about using a subset of XML to make SQL-like
tables, saving the data in xml format.

There was an api and code examples, and I seem to remember that there was
Python examples too.

Does anybody have a link???

Regards Max M



From le_snelson at transtechinc.cc  Thu Dec 28 19:24:03 2000
From: le_snelson at transtechinc.cc (Le Snelson)
Date: Thu, 28 Dec 2000 16:24:03 -0800
Subject: How to import from within a module
Message-ID: <3A4BD9A3.85081B3@transtechinc.cc>

I am running Python 2.0 on Solaris 2.5x

I am attempting to import a module from in a function within a sort of
root module [init] - it contains only an initialization function.  The
intent is that configuration files will provide data to derive the name
of the module to be imported. All of the scripts provide the same
function names but for different underlying devices.

I am unable to get __import__( xxx ) to actually add xxx to any visible
dictionary. However, in the same file if I uncomment a line at the top
(global scope) that reads:
import xxx
Sure enough the xxx is in the list from dir(init).

I have tried all combinations of __import__(xxx, globals(), locals())
__import__(xxx, globals(), globals()) etc. I have also tried the "import
xxx" within the function I'm working on.  I expected syntax error but
got none as well as no discernible result.

Do I need to use the imp module to get this module found and loaded?

Any pointers to discussions or sample code appreciated.


-------------- next part --------------
A non-text attachment was scrubbed...
Name: le_snelson.vcf
Type: text/x-vcard
Size: 340 bytes
Desc: Card for Le Snelson
URL: <http://mail.python.org/pipermail/python-list/attachments/20001228/4c9a385c/attachment.vcf>

From andrew.j.atkinson at unilever.com  Tue Dec 19 05:11:57 2000
From: andrew.j.atkinson at unilever.com (andrew.j.atkinson at unilever.com)
Date: Tue, 19 Dec 2000 10:11:57 -0000
Subject: Dispatch("Excel.Application") = Access denies
Message-ID: <91nc9d+6rni@eGroups.com>

Dear all

Have developed a python app to automate the running of an excel 
spreadsheet. Works a treat when run on the command line, however get 
the following 'trace' when run through Tivoli's Maestro job scheduler 
(same user in both cases).

Any ideas (it's quite urgent).

Traceback (most recent call last):
  File "e:\rlnklive\bin\test.py", line 24, in ?
    xlApp=Dispatch("Excel.Application")
  File "e:\python20\win32com\client\__init__.py", line 94, in Dispatch
    dispatch, userName = dynamic._GetGoodDispatchAndUserName
(dispatch,userName,clsctx)
  File "e:\python20\win32com\client\dynamic.py", line 80, in 
_GetGoodDispatchAndUserName
    return (_GetGoodDispatch(IDispatch, clsctx), userName)
  File "e:\python20\win32com\client\dynamic.py", line 71, in 
_GetGoodDispatch
    IDispatch = pythoncom.CoCreateInstance(IDispatch, None, clsctx, 
pythoncom.IID_IDispatch)
pywintypes.com_error: (-2147024891, 'Access is denied.', None, None)

Kind Regards
-Andrew





From spamfranke at bigfoot.de  Mon Dec 18 17:29:23 2000
From: spamfranke at bigfoot.de (Stefan Franke)
Date: Mon, 18 Dec 2000 22:29:23 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com> <kuu28971kp.fsf@lasipalatsi.fi> <3A380E4E.8537A0FB@san.rr.com> <91a1vr02b02@news2.newsguy.com> <Co9_5.25465$x6.13756579@news2.rdc2.tx.home.com> <91bbq102mtk@news1.newsguy.com> <rWd_5.26820$x6.14119694@news2.rdc2.tx.home.com> <91d5el0me7@news1.newsguy.com> <rOr_5.31120$x6.15241856@news2.rdc2.tx.home.com> <slrn93o7j4.tu0.quinn@dinar.ugcs.caltech.edu> <slrn93pago.8tf.quinn@euro.ugcs.caltech.edu>
Message-ID: <3a3e8f75.4591980@news.btx.dtag.de>

On 17 Dec 2000 11:57:13 GMT, quinn at euro.ugcs.caltech.edu (Quinn Dunkan) wrote:

>Note to self: when newsreader says "Post failed, try again? (y/n) " what it
>*really* means is "Post succeeded, make a fool of yourself on Usenet? (y/n) ".

<grin> You aren't using Agent, are you?

deja-vou-ly y'rs
Stefan



From pearu at cens.ioc.ee  Sun Dec 31 09:18:03 2000
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Sun, 31 Dec 2000 16:18:03 +0200
Subject: 2001 Enchancement Wishlist
In-Reply-To: <mailman.978265218.29152.python-list@python.org>
References: <mailman.978265218.29152.python-list@python.org>
Message-ID: <Pine.LNX.4.10.10012311534410.25088-100000@kev.ioc.ee>

On Sun, 31 Dec 2000, Tim Peters wrote:

> [Pearu Peterson]
> > ...
> I bet it can't be handled that way in Python 42.0 either <0.41 wink>:  on
> the very face of it, the only *sensible* thing for a NonZeroInteger
> constructor to do when handed a 0 is to raise ValueError -- else its name
> lies about the nature of the thing it's constructing.
> 
> Guido has Pronounced on this in the past:  a class constructor creates an
> instance of the class, end of story.  It's a little bit of semantics people
> can and do rely on.  For example, you didn't show any other methods of your
> NonZeroInteger class, but let's say it had a method xxx.  Then any Python
> programmer can rely on that
> 
>     y = NonZeroInteger(anything_at_all)
>     y.xxx()
> 
> won't blow up with an AttributeError on "xxx", and without doing dynamic
> type checks etc first.

Ok, the given NonZeroInteger was a bad example. I was quietly assuming
that whatever __init__ returns, has the same attribute set as
NonZeroInteger, that is, all these classes are derived from some parent
class (as you suggested also below).

Anyway, if __init__ returning is 'out', then what about class methods
that Gareth McCaughan suggested earlier. How 'out' or 'in' would that
solution be?

> There's nothing wrong with writing a factory function that can return
> objects of different classes.  It often makes code using them harder to work
> with, though, and precisely because you *don't* know what pops out of them
> in the absence of further type checks (however spelled).

I am willing to take that risk if it will reduce code considerably. It is
just a matter of whether a code developer is able to write consistent
code, whatever tricks are used.

> > c = Plus(a)           -> c = a          # c is Symbol instance
> >
> > ***lots of code*** could be avoided with rich Plus.__init__ that would
> > return in this case Symbol instance a.
> 
> Wouldn't your "rich Plus.__init__" contain about as much code as a factory
> function would today?  Both have to consider all the same cases.  A factory
> function would get to omit the often-useless "self" argument, though <0.9
> wink>.

No. The code would be cleaner in various parts. And it would be also
faster. And all that because of there would be no need to bother whether
an object is in its minimal representation.
Examples of minimal representations:
        Plus(5) is mpz(5)
	Plus(a) is a
	Plus(a,b) is Plus(a,b)
	etc

> I suggest you're off track in your modeling, confounding roles that should
> be distinct:  that of Plus as a function and Plus as an object (representing
> an only partially evaluated sum).  In this case I'd probably have a base
> class Expression for all kinds of expresssions (incl. atoms like Symbols),
> make Plus a function mapping zero or more Expressions (or subclasses) to an
> Expression (or subclass), and make Sum a subclass of Expression that
> represents a partially evaluated sum.  

Yep. That's exactly how I have modeled it in PySymbolic (though with a bit
different naming convention but at least I am not too much off the
track;-).

> That's the way I *think* of the
> domain, so that's the way I'd want to code it (and, no, I'm not bothered by
> having to come up with distinct names for the verb sense (Plus) and the noun
> sense (Sum) -- the distinction between nouns and verbs is usually helpful!).

Just for information:
There is a difference between Plus and Sum:
Plus - is a sum of Symbols, where the number of Symbols is
       *always defined and fixed*. E.g. Plus(a,b,c) is a+b+c
Sum  - is a sum of Symbols, where the number of Symbols is *not defined*,
       in general. E.g. Sum(n*n,n=[0,m]), where m is arbitrary integer.
And Plus will be used only internally in PySymbolic. Actually it just
represents operation symbol `+', not the verb `plus'.

> symbolic-math-would-be-useless-if-we-always-wanted-the-verb-
>     form<wink>-ly y'rs  - tim

Thanks,
	Pearu



From andrew_dot_henshaw_at_earthling_dot_net  Wed Dec 20 22:44:45 2000
From: andrew_dot_henshaw_at_earthling_dot_net (Andrew Henshaw)
Date: Wed, 20 Dec 2000 22:44:45 -0500
Subject: while true: !!!
References: <mailman.977339546.988.python-list@python.org>
Message-ID: <t42vejmugv6j95@corp.supernews.com>

"Tim Peters" <tim.one at home.com> wrote in message
news:mailman.977339546.988.python-list at python.org...
> [gbreed at cix.compulink.co.uk]
> > ...
> > If you want to empty a sequence, but preserve the type, try
> > filter(lambda x:0, asequence).
>
> Picking a nit:  that returns an empty list, regardless of type(asequence).
> To get an empty sequence of the same type, do
>
>     asequence[:0]
>
> or even
>
>     asequence * 0
>
> there's-only-one-way-to-do-it-ly y'rs  - tim
>
>
Yes, that's the way that I was thinking the (arguably logical) change in
"not sequence" semantics might be used.  I don't propose that such a change
be made, however.

Andrew Henshaw




From sandj.williams at gte.net  Mon Dec 11 11:38:17 2000
From: sandj.williams at gte.net (Steve Williams)
Date: Mon, 11 Dec 2000 16:38:17 GMT
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com>
Message-ID: <3A350405.B8F09EC7@gte.net>

Jaroslav Gergic wrote:

> Hello,
>
> one thing I dislike in many Python tutorials is
> the style of infinite 'while' loop with 'break' statements:
>
> > Python:
> >
> >     file = open('file.input')
> >
> >     count = 0
> >     while 1:
> >         lines = file.readlines(8192)
> >         if not lines: break
> >         count = count + len(lines)
> >
> >     print count
> >
>
> I hate it so much I prefer to write something like this:
>
> line = fh.readline()
> while(line != ""):
>   ... do something ...
>   line = fh.readline()
>
> OK, it is ugly to write the same line of code twice,
> but 'while true' or 'while 1' is very non-programmer construction
> it smells like VisualBasic GOTO.
>

[snip]

The two readlines are not the same.  The first tells you if the file is
empty, an important consideration in many cases.

In some applications, you use the result of the first readline to prime
control variables.

I like your loop structure, it's not ugly and I don't think it needs to
be changed.

I wouldn't call "While true" a python-idiom, I'd call it a
python-solecism.





From kare at speech.kth.se  Thu Dec 21 11:18:18 2000
From: kare at speech.kth.se (Kare Sjolander)
Date: Thu, 21 Dec 2000 17:18:18 +0100
Subject: audio playback
References: <3a40c236.3935118@news.frontiernet.net>
Message-ID: <3A422D49.6FA64ECE@speech.kth.se>

One way is to use Snack, http://www.speech.kth.se/snack/, which Windows
95/98/NT/2000, Linux, Macintosh, Sun Solaris, HP-UX, NetBSD, and SGI
IRIX.

To play a sound file

from tkSnack import *
initializeSnack(Tkinter.Tk())

s = Sound(file='ex1.wav')

s.play()


There are many options, e.g. to add an echo effect during playback you
could use:

f = Filter('echo', 0.6, 0.6, 30, 0.4)

s.play(filter=f)


Kare




From katz at Glue.umd.edu  Sat Dec  9 00:48:00 2000
From: katz at Glue.umd.edu (Roy Katz)
Date: Sat, 9 Dec 2000 00:48:00 -0500
Subject: Dijkstra's Shortest Path algorithm
Message-ID: <Pine.GSO.4.21.0012090042370.15110-100000@y.glue.umd.edu>

Hello!


For a recent assignment in my computer science class, we coded up a
graph which implements the Shortest Path algorithm (it's a model of
the Internet, with 'web pages' as nodes, and the latency between them
as edges).  It took a good two or three days for me in C++, and
I was wondering if Python (or Numerical Python) has anything like this
(a Graph module??).  If there isn't, perhaps I should post my class at
SourceForge and try to make an extension out of it :)


Thanks!
Roey Katz

computer engineering
University of Maryland, College Park -- USA.




viva esperanto!



From aahz at panix.com  Mon Dec 25 14:04:51 2000
From: aahz at panix.com (Aahz Maruch)
Date: 25 Dec 2000 11:04:51 -0800
Subject: __init__ question
References: <EvM16.270$LU6.118037@typhoon.aracnet.com>
Message-ID: <9285oj$6rh$1@panix2.panix.com>

In article <EvM16.270$LU6.118037 at typhoon.aracnet.com>,
Daniel Klein <DanielK at aracnet.com> wrote:
>
>    def __init__(self, *args):
>
>What is the meaning of the '*args' expression?

It means "accept any number of parameters here"; "args" is a tuple.  For
example, I could create a print function instead of the blecherous
redirected print statement as follows:

def print (file=None, *args):
    if not file:
        file = sys.stdout
    for arg in args:
        file.write("%s " % arg)
    file.write("\n")
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

'"Crisp" is a good quality for crackers; less so for pot roast.'  --pnh


From dbrueck at edgix.com  Tue Dec 12 12:42:38 2000
From: dbrueck at edgix.com (Dave Brueck)
Date: Tue, 12 Dec 2000 10:42:38 -0700
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
In-Reply-To: <3A365EB0.C3E02AA8@cybermesa.com>
Message-ID: <015401c06462$ebd129e0$760214ac@PRODUT2KDAVE>

> > Tough for the new Python programmer, then.  It's nt fair to them.
>
> Actually, it's not really that hard for the new programmer,
> but maybe a little odd for a programmer coming from certain languages.

Right on the money! When you move to Python (or to any new language), you
have to leave behind some of the mental baggage that goes with the former
languages. If you look at Python from a complete newbie perspective, it's
not so hard to understand. Everything is an object, and variables refer to
them. It stopped giving me problems once I forced myself to stop thinking in
terms of pointers. :)

Talking in terms of pass-by-value/reference might be a good way to quickly
transition from a language like C++, but it's a shortcut that will cost you
in the long run, so whoever is confused should get away from that way of
thinking ASAP. It's only natural to try to fit something new into the
context of what you already know (that helps humans adapt quickly), but the
danger is that your thinking won't ever progress. It's sooo common to see
"C++ programmers" that still program in C - they may use the syntactic
features of C++, but their thinking (and as a result, their programs), are a
far cry from object-oriented.

-Dave




From fredrik at effbot.org  Fri Dec 29 16:31:29 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 29 Dec 2000 21:31:29 GMT
Subject: 2001 Enchancement Wishlist
References: <3A4CEACB.220998A0@javanet.com>
Message-ID: <Ro736.2511$Qb7.273102@newsb.telia.net>

Raymond Hettinger wrote:
> 3.  __init__ is the only special method that doesn't allow a return value.
>    I think it should continue to return the new object by default but also
>    allow another return item to be substituted.  For example:
>
>             def __init__( self ):
>                    if MyClass.aSingleton == None:
>                          MyClass.aSingleton = self
>                    return MyClass.aSingleton

doesn't make sense: __init__ *initializes* an instance
object, it doesn't create a new one.

:::

if you still think creating new objects just to throw them
away really is a good idea, I'd rather see the return value
check changed from

    value = self.__init__(*args, **kwargs)
    assert value is None

to

    assert value is None or isinstance(value, self.__class__)

or even:

    assert value is None\
        or isinstance(value, types.InstanceType)\
        and value.__class__ is self.__class

:::

allowing an arbitrary value (as long as it's not None) seems
like a really unpythonic idea.

(fwiw, singleton classes are pretty unpythonic too.  python
already has a framework for creating exactly one instance of
something: the module namespace).

</F>




From tim.one at home.com  Fri Dec 22 21:30:22 2000
From: tim.one at home.com (Tim Peters)
Date: Fri, 22 Dec 2000 21:30:22 -0500
Subject: namespaces & indentation
In-Reply-To: <920f54$m2f$1@news.mch.sbs.de>
Message-ID: <LNBBLJKPBEHFEDALKOLCCECOIFAA.tim.one@home.com>

[unattributed]
> I always thought that I understood Python's scoping, but then I
> came across that code:
> > def t(x):
> >   if x == 1:
> >     c = 'yeah'
> >   return c

[Chris Schaller]
> If I execute t(1), I'll get 'yeah', but otherwise there'll be an
> error message.  According to the indentation level c is only valid
> within the if-clause, but not on the outside.

You made that up out of your own head, you know <wink>.

> So I guess, the indentation level only defines the lines of code
> that belong to the if-clause, but has nothing to do with scopes.

Very close.  Scopes are defined with respect to "execution frames", which
are in turn attached to "code blocks", and the code block of a function is
its entire body (less the bodies of nested functions and classes).  The
complete rules are here:

    http://www.python.org/doc/current/ref/execframes.html

Well worth reading!  The answers to many "deep" questions are waiting there.

>   That's very confusing, I've already programmed a lot in Python,
> but only due to a program error (I forgot to initialize c at the
> beginning of the function - a common error) and an unsuccessful
> test this happened to a program after months of error-free running.
>
>   Any comments?

1. Strive to test all paths thru each function (good advice regardless of
language).

2. It's likely that a framework for warning messages will be added to Python
2.1.  In the case above, the compiler could do some flow analysis and
generate a compile-time warning that "c" may not be bound at the time the
"return" is executed.  Java's rules make that kind of code a compile-time
error, but some people play such extreme tricks with Python's dynamicism
that we'll probably have to settle for a nag msg in Python.

if-thine-eye-offends-thee-pluck-it-out-ly y'rs  - tim




From aahz at panix.com  Fri Dec 29 18:42:14 2000
From: aahz at panix.com (Aahz Maruch)
Date: 29 Dec 2000 15:42:14 -0800
Subject: Interesting behaviour of the assignment
References: <92f4gk$qk6$2@news.nuri.net> <92fhsr01u88@news1.newsguy.com> <mailman.978059501.14500.python-list@python.org> <92j4n1$mhu$0@216.39.162.232>
Message-ID: <92j7gm$6v2$1@panix2.panix.com>

In article <92j4n1$mhu$0 at 216.39.162.232>,
Michael Esveldt  <dante at oz.net> wrote:
>[Thomas Wouters]
>>
>> In other words, 'x == y' first does 'x is y', and if it's true, the
>> result of the operation is true. Only if 'x is y' is false will it
>> start the (possibly very expensive) comparison-by-contents operation.
>
>This is a somewhat ignorant question, but if 'x == y' checks for 'x is
>y' first, why use 'is' at all? If you only get a tiny bit of speed out
>of using 'is' is it worth it to have this in the language at all? (Not
>that it would ever change, just curious.)

Because sometimes you really do want to know if two objects are the same
(rather than simply being equal).  If you drop "is", you can't check for
object identity.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

Ninth Virtual Anniversary: 2 days and counting


From aleaxit at yahoo.com  Tue Dec 12 05:38:07 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 12 Dec 2000 11:38:07 +0100
Subject: Inheriting the @ sign from Ruby
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu>
Message-ID: <914v6l02ouv@news2.newsguy.com>

"Roy Katz" <katz at Glue.umd.edu> wrote in message
news:Pine.GSO.4.21.0012120138090.2874-100000 at y.glue.umd.edu...
>
> I was reading a Dr. Dobb's recently and it mentioned that Ruby uses the @
> symbol to denote instance members of a class.  This appeals to me because
> I can comprehend it faster than seeing 'self' pasted in front of a
> name, especially when the statement is buried in dense code.

'self' is not "pasted in front of a name" -- 'self' is an object,
and it is used with perfectly regular syntax for object access.

Thus, there is no special-case syntax needed: the regular syntax,
which you need anyway for general cases, takes care of everything
in the special-case as well.


> Why not bind @ to the first argument of method definitions? Consider:
>
>
>   class X:
>
>     def __init__( self, x ):
>
>       @bind(x)     # binds to 'self', because
>                    # 'self' is the first argument to __init__

If you called
    someotherobject.bind(x)
you WOULD know what that does, right?  Call method 'bind' on
that object with 'x' as an argument.

Therefore,
    self.bind(x)
does exactly the same thing, simply and predictably -- the
object on which 'bind' is called being, unsurprisingly,
the one named 'self'.

So, you would like '@identifier' to be a special-case
syntax meaning exactly the same as 'self.identifier' (if
the first argument, as per normal convention, is named
'self').


>     def bind(myself, x):
>       @x = x   # equivalent to myself.x

The violation of the convention (using 'myself' instead
of 'self' as the name of the first argument) is not
remedied by the '@' kludge, of course.  Whenever a use
of 'self' is desired, that cannot be brought to the
attribute-access syntax of self.whatever (passing
'self' as an argument to getattr, or to a base-class
version of a method, etc, etc), such a violation of
normal expectations will make your code harder to read
and understand by any maintainers.

I'd much rather have a lint-like checking that the
argument expected to be named self is in fact thus
named, with a warning when it isn't.


> What faults or strengths do you see with this?

I see no strength in this supplementary syntax form
offering a second, different way to express what can
be perfectly well expressed explicitly today.

I see it as a fault that all this does is offer a
second, different way to express what can be perfectly
well expressed explicitly today.

It would also be a (small) step along the road of
turning the currently crystal-clear syntax of Python
into that morass of special-purpose, ad-hoc, 'convenient'
kludges that make certain other languages' typical
programs look so peculiarly similar to line noise.


Alex





From quinn at ngwee.ugcs.caltech.edu  Fri Dec  8 14:48:36 2000
From: quinn at ngwee.ugcs.caltech.edu (Quinn Dunkan)
Date: 8 Dec 2000 19:48:36 GMT
Subject: Bug: Readline getting stuck on Linux and Solaris
References: <lxk8zqwupy2.fsf@ioasun3.epfl.ch> <90m53k$ovh$1@nnrp1.deja.com> <m3u28h48ja.fsf@atrus.jesus.cam.ac.uk> <slrn92u8al.76b.quinn@lira.ugcs.caltech.edu> <m3r93k4kqk.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <slrn932eoj.d1j.quinn@ngwee.ugcs.caltech.edu>

On 07 Dec 2000 09:57:23 +0000, Michael Hudson <mwh21 at cam.ac.uk> wrote:
>quinn at lira.ugcs.caltech.edu (Quinn Dunkan) writes:
>
>> Strange.. readline has *never* worked right for me, on IRIX or
>> Linux, on any version of python (1.4, 1.5.2, 2.0), and any version
>> of readline (3.0, 4.0, 4.1), if python is compiled with threads.
>> IRIX (6.4, 6.5, 6.5.2) hangs, Linux (2.0.36, 2.2.12, 2.2.17) exits.

Well, blow me down and pick me up, I just tried python 2.0 on linux (it was
2.0b1 before) and the problem is solved, even with threads.

>Well, I have readline 2.2.1, which may make a difference.  Can you
>supply a stacktrace?
>
>> I just tried your 'while 1: pass' thing, and the behaviour is the same there
>> too.
>
>That really is odd, because readline doesn't have control then and
>should be out of the picture.  Does control-C work as expected here if
>readline isn't enabled, or if you hit it before the first prompt
>appears (this may be tricky).

It's not tricky on an Indy :)  In any case, on the irix machine, I've been
using 2.0-for-real and it still locks.  It apparently works ok if I
hit control C quickly: I get 'import site' failed, presumably because it died
of a KeyboardInterrupt.  However, even a 'while 1: pass' ^C locks.  I got a
backtrace by suspending the locked python, and killng it with a SIGILL to get
a core.
(gdb) bt
#0  0xfa3ab88 in __ioctl () at lio.c:201
#1  0xfa434f0 in ioctl () at nftw.c:184
#2  0x5ffd2a74 in rl_restart_output ()
#3  0x5ffdd0f4 in _rl_clean_up_for_exit ()
#4  0x5ffdd860 in rl_cleanup_after_signal ()
#5  0x5ffdd304 in _rl_erase_entire_line ()

Once again, Irix has no problems if I compile --without-threads (which "fixed"
the linux problem too, so I assumed they were the same bug).  This is Irix
6.5, with gcc 2.8.1 and realine 4.  On Monday I'll try with readline 2 and
see if it's still there.

^C works properly in all instances if readline is disabled (via
'mv readline.so spam', if there's a more graceful way I don't know it).


From chris at araidesign.com  Tue Dec 12 20:12:45 2000
From: chris at araidesign.com (Chris Arai)
Date: Tue, 12 Dec 2000 17:12:45 -0800
Subject: How to use BLT2.4 and Tcl/tk8.3 under Windows9X/NT ?
References: <39F6BBCD.C25D4FEA@mailux.berl.tek.com> <3a02381b.581985@news-server.socal.rr.com>
Message-ID: <3A36CD0C.8B79DD33@araidesign.com>

Howard,

Did you ever resolve this problem??  If you did I would like to know how, as
I am having the same problem.

Thanks,
Chris

howard at eegsoftware.com wrote:

> On Wed, 25 Oct 2000 10:54:06 GMT, Matthias Fischer
> <mfischer at mailux.berl.tek.com> wrote:
>
> >I use Python 2.0 under WIN32 with the included package of Tcl/Tk 8.3 and
> >want to use the BLT2.4 expansionspack with Pmw.
> >I installed the Pmw and BLT packages in the directories as explained in
> >the documentation. If I try to execute an example I'll get the
> >following Message:
> >
> >Sorry. The BLT package has not been installed on this system. Please
> >install it and try again.
> >
> >Please tell me, what may I do. (reinstall the BLT package in an other
> >directory or copy some DLL's in Pythons DLLs dir ?)
> >
>
> I also have the same problem.  I then tried to run the PMW tests under
> 2.0 and gots lots of small errors reported.....so.....I tried again
> under my "working" 1.5.2 installation.   It ALSO reports lots of small
> errors but DOES display the graph examples also.
>
> I am assuming that this is just an installation issue with the
> locations of tcl.  Under 1.5.2, the Windoze installation had tcl under
> "Program FIles" with BLT installed in that hierarchy.  Under 2.0, tcl
> is in a directory under python20 but the depth of the heirarchy
> appears to be different.
>
> Any clues??
> Howard Lightstone
> EEGSoftware
> howard at eegsoftware.com

--
Chris Arai
Arai Design
chris at araidesign.com
http://www.araidesign.com
707 431 1299
FAX 431 1599

-------------- next part --------------
A non-text attachment was scrubbed...
Name: chris.vcf
Type: text/x-vcard
Size: 224 bytes
Desc: Card for Chris Arai
URL: <http://mail.python.org/pipermail/python-list/attachments/20001212/51363803/attachment.vcf>

From aleaxit at yahoo.com  Fri Dec 15 08:24:27 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 15 Dec 2000 14:24:27 +0100
Subject: Python advocacy
References: <91bln3$lmr$1@nnrp1.deja.com>
Message-ID: <91d63k0msl@news1.newsguy.com>

<jschmitt at vmlabs.com> wrote in message news:91bln3$lmr$1 at nnrp1.deja.com...
> I just read this good article about Perl advocacy.  I think it applies
> to Python equally.  I found the link mentioned from /. this morning.
>
> http://www.perl.com/pub/2000/12/advocacy.html?wwwrrr_20001213.txt#why i
> hate advocacy

Very interesting (for some reason, a not-yet-Pythonista Perl-using
coworker forwarded me an e-mail copy too -- wonder if he's trying
to tell me something?-).  Highly recommended reading for anybody
interested in really understanding why human beings are this way,
Matt Ridley, "The Origins of Virtue".  There's a long detailed
summary at http://www.geocities.com/Athens/Ithaca/4388/origins.html,
but you wouldn't be doing yourself a favor if you limited your
reading to those 23 pages -- Ridley writes very well, and argues
persuasively; get the original!

But Python is well-placed here.  The BDFL was so clever as to
insert in the language the 'print >> file' construct, so we all
have ONE horrible wart in the language to castigate mercilessly,
thereby making our praise-to-high-heavens of every other aspect
of the language much more credible.  This is called the 'scapegoat
maneuver', and it's great high strategy in advocacy contexts,
which several well-known software firms have pioneered (having
in their product lines at least ONE product so unreservedly bad
that even their most rabid salespeople cannot fail to trash
them, thereby gaining customers' trust for their praise of
the _other_ products, the ones they really mean to sell).

Makes me wonder why Larry Wall didn't think to include at least
ONE strangeness, wart, or inconsistency *somewhere* in Perl...
thus making it so boringly perfect and un-idiosyncratic as to
leave its advocates without a suitable 'scapegoat feature'!


Alex





From SBrunning at trisystems.co.uk  Tue Dec 12 04:37:02 2000
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Tue, 12 Dec 2000 09:37:02 -0000
Subject: Inheriting the @ sign from Ruby
Message-ID: <31575A892FF6D1118F5800600846864D5B15B7@intrepid>

> From:	Roy Katz [SMTP:katz at Glue.umd.edu]
> What faults or strengths do you see with this?
 
It's *ugly*.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk





-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From gtalvola at NameConnector.com  Thu Dec  7 20:40:26 2000
From: gtalvola at NameConnector.com (gtalvola at NameConnector.com)
Date: Fri, 08 Dec 2000 01:40:26 GMT
Subject: Effect of Ctrl-C on threaded apps
Message-ID: <90pe6a$fo8$1@nnrp1.deja.com>

I've noticed that on Windows, Ctrl-C causes a KeyboardException
to be thrown always in the main thread.  But I'm told that on Linux,
the exception can be thrown in ANY of the threads.  Does anyone
have any insight into this?

Also, is there any control over _when_ the Python interpreter will
throw the exception?  I've noticed, for instance, on NT if the main
thread is blocked in a Queue.put() operation on a full queue, and
a KeyboardInterrupt is triggered by a Ctrl-C, the Queue will be
left in a broken state -- the next put() operation will block
forever.  It looks like the KeyboardInterrupt gets thrown whenever the
control returns to the interpreter, and the Queue module lacks the
proper try/finally's to unlock its internal lock in this case.

Basically, is there any way to get more control over this behavior?
I'm using 1.5.2 on NT in case there's a difference.  Ideally, I'd be
able to register a handler for Ctrl-C that would call a function that
I specify, then return control to the application at the same place
it left off without an exception being thrown.

- Geoff


Sent via Deja.com http://www.deja.com/
Before you buy.


From tim.one at home.com  Sun Dec 24 14:16:43 2000
From: tim.one at home.com (Tim Peters)
Date: Sun, 24 Dec 2000 14:16:43 -0500
Subject: A Suggestion for Python Colon Syntax
In-Reply-To: <20001223181741.F5515@xs4all.nl>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEGFIFAA.tim.one@home.com>

[Vadim Zeitlin]
> BTW, what was the real rationale for adding composed assignment
> operators (+=) to Python 2.0 (which I'm really extremely glad to
> have now)? Wasn't it meant to make Python more attractive to the
> people coming from C-lie languages?

[Thomas Wouters]
> No, honestly not. If that was the case, '++' and '--' would have
> been added too. The funny part is that I never saw a concious
> decision not to add them being made, everyone involved (Guido,
> Michael Hudson, me, the rest of python-dev) seemed to naturally
> accept that the auto-inc/dec operators had no place in Python, and
> that in itself is plenty of validation, don't you think ? :-)

++ and -- were explictly suggested and explicitly rejected innumerable times
on c.l.py, and on the Python mailing list before c.l.py came into glorious
being.  In fact, when Guido is deciding whether to invite someone into
python-dev, he sends us secret email asking "hey!  do you think this guy
would approve of adding ++?" <wink>.

> Guido-probably-made-all-decisions-about-Python-years-ago--he's-just-
>    waiting-for-people-to-_implement_-them-ly y'rs,

Indeed, that was very much the case for augmented assignment, listcomps and
Unicode, and is currently the case for rich comparisons and reworking the
coercion model.  It will be the case for healing the type/class split.

Python is like the Sistene Chapel that way:  Guido already knows how it will
look in the end, it's just taking years of tedious painting to get there
<wink>.

and-nobody-guessed-brushes-would-wear-out-so-fast-ly y'rs  - tim




From qrczak at knm.org.pl  Thu Dec 21 02:14:44 2000
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 21 Dec 2000 07:14:44 GMT
Subject: Python's scope rules
References: <3A40813B.C04D21C@et.dtu.dk> <91s38r$sjd$1@news.udel.edu>
Message-ID: <slrn943bf3.q3c.qrczak@qrnik.zagroda>

Wed, 20 Dec 2000 12:47:09 -0500, Terry Reedy <tjreedy at udel.edu> pisze:

> The only things I find 'odd' (and definitely confusing at first)
> is the use of 'global' for the module namespace (and I suppose
> there is some history here) and the sometimes claim that there are
> two rather than three possible scopes.

The fact that scopes don't nest is a practical problem. Yesterday I was
doing some customizations to viewcvs and wanted to factor out a common
pattern of code of 5 lines which was repeated 5 times by moving it to
a local function, but it referred to so many free local variables that
I finally did not do it, because I would have to pass them explicitly
and the result would not be much less ugly than the current state.

I don't want dynamic scoping like in old Lisp and probably nobody wants
it. But nested lexical scoping is used in other good languages, it's
so natural and convenient that it's odd that Python does not use it.
IMHO Python would be a better language with true lexical scoping.

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


From matt at virtualspectator.com  Fri Dec 22 18:06:27 2000
From: matt at virtualspectator.com (matt)
Date: Sat, 23 Dec 2000 12:06:27 +1300
Subject: namespaces & indentation
In-Reply-To: <920f54$m2f$1@news.mch.sbs.de>
References: <920f54$m2f$1@news.mch.sbs.de>
Message-ID: <0012231207220C.00902@localhost.localdomain>

that is perfectly normal ...

this might help you
http://python.sourceforge.net/peps/pep-0227.html


On Sat, 23 Dec 2000, Chris Schaller wrote:
> Dear Pythoners...
> 
>   I always thought that I understood Python's scoping, but then I came across
> that code:
> > def t(x):
> >   if x == 1:
> >     c = 'yeah'
> >   return c
> 
> If I execute t(1), I'll get 'yeah', but otherwise there'll be an error
> message.  According to the indentation level c is only valid within the
> if-clause, but not on the outside.  So I guess, the indentation level only
> defines the lines of code that belong to the if-clause, but has nothing to do
> with scopes.
> 
>   That's very confusing, I've already programmed a lot in Python, but only
> due to a program error (I forgot to initialize c at the beginning of the
> function - a common error) and an unsuccessful test this happened to a program
> after months of error-free running.
> 
>   Any comments?
> 
> bye
>   Chris...
> -- 
> http://www.python.org/mailman/listinfo/python-list
-- 
Matt Halstead (PhD)
Research and development
VirtualSpectator
http://www.virtualspectator.com
ph 64-9-9136896



From obou at my-deja.com  Wed Dec  6 08:25:34 2000
From: obou at my-deja.com (obou at my-deja.com)
Date: Wed, 06 Dec 2000 13:25:34 GMT
Subject: Sort
References: <NlnX5.1444$yv1.2700363@nnrp5.proxad.net>
Message-ID: <90leo6$4oi$1@nnrp1.deja.com>

In article <NlnX5.1444$yv1.2700363 at nnrp5.proxad.net>,
  "Alfred" <scj at mcom.mcom.fr> wrote:
> Hello world,
>
> I'm trying to sort a List of List, on the third col, but I don't find
how to
> do this.
>
> Please help me. ;)
>
> Greg
>
>

Probably not the cleanest way to do it, but you could implement a
wrapper for your list that implements a __cmp__ function. The __cmp__
function uses the cmp function on the item of the third col and returns
the value.

Sample code here.

# Create a list wrapper that implements a __cmp__ function
class MyListWrapper:
    # constructor, takes a list and the number of the col to sort with
    def __init__(self, a_list, sort_it = 2):
        self.my_list = a_list
        self.sort_it = 2

    # comparison function
    def __cmp__(self, listWrapper):
        return cmp(self.my_list[self.sort_it], listWrapper
[self.sort_it])

    # to let the items be accessed thru []
    def __getitem__(self, i):
        return self.my_list[i]

    # to return the list
    def getList(self):
        return self.my_list


# Test of the wrapper
if __name__ == '__main__':

# create the list
    l = [[5, 3, 8, 5],
         [7, 1, 0, 3],
         [7, 7, 4, 9],
         [8, 3, 5, 3],
         [2, 6, 8, 5]]

# convert the list of list to a list of "sortable list wrappers"
    for i in range(len(l)):
        l[i] = MyListWrapper(l[i])

# sort with standard sort function
    l.sort()

# convert back to a list of list
    for i in range(len(l)):
        l[i] = l[i].getList()

Hope this helps.

Olivier


Sent via Deja.com http://www.deja.com/
Before you buy.


From gerhard.nospam at bigfoot.de  Thu Dec  7 11:44:24 2000
From: gerhard.nospam at bigfoot.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Thu, 7 Dec 2000 17:44:24 +0100
Subject: Oracle 8.1.6 8.1.7
References: <90np2v$7hk$1@proxy.fe.internet.bosch.de>
Message-ID: <slrn92vg58.ut.gerhard.nospam@lilith.hqd-internal.de>

The possibility to use the mxODBC module was already mentioned. There
another module called DCOracle, which uses native OCI calls to communicate
with Oracle. It depends on your needs and OS (Windows vs. Unix) which one is
easier to setup and suited to the job. OCI should be faster. The URL is:
http://www.zope.org/Products/DCOracle/

Gerhard

PS: If you are using Windows, you can install the win32 extensions from
ActiveState, if you not already have, and use ADO. Just to add to the
confusion ;-)

>Hi there,
>
>ist there a way to get a connection to an Oracle DB with Version 8.1.6 or
>8.1.7.

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


From chicago_george at my-deja.com  Tue Dec 12 18:06:54 2000
From: chicago_george at my-deja.com (chicago_george at my-deja.com)
Date: Tue, 12 Dec 2000 23:06:54 GMT
Subject: Simple question/problem
Message-ID: <916b28$9lf$1@nnrp1.deja.com>

Hello, I'm new to Python.. When I try to import java classes, I get an
error msg saying 'no module named java'
For example;

>>> from java import awt
ImportError: no module named Java

Why is this happening?  Anyone???

Thanks,
George.


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


From erno-news at erno.iki.fi  Sun Dec 24 23:54:18 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 25 Dec 2000 06:54:18 +0200
Subject: classes (was Re: Same again please for OOP)
References: <mailman.977623149.8184.python-list@python.org> <9257mf0sl1@news1.newsguy.com> <roy-E5B925.13324224122000@news.panix.com> <925khi019hs@news1.newsguy.com>
Message-ID: <kubsu1t81x.fsf@lasipalatsi.fi>

>>>>> "Alex" == Alex Martelli <aleaxit at yahoo.com> writes:

| "Roy Smith" <roy at panix.com> wrote in message
| news:roy-E5B925.13324224122000 at news.panix.com...
[...]
|| The mp3 class might return a URL for its location() method, and the
|| CD or Vinyl classes might return a string, "Third shelf from the
|| bottom,

| Looks like a typical case of state mismodeled as behavior.  Why
| should 'location' be a method -- what does it *DO*?!  It doesnt _do_
| anything; rather, it's an 'accessor' for a part of the object's
| state.

i don't understand what is wrong with this - i like making them all
methods, so i don't have to remember if it's an attribute or method
later. also overrinding __getattr__ later, if it turns out to be
needed, is tedious and ugly.

why does the user want/need to know if "location" is stored directly
as an attribute or if computation is needed to to reconstruct it
from some other information?

  -- erno




From rob at ZOOstation.cc  Mon Dec 11 02:12:02 2000
From: rob at ZOOstation.cc (Rob Brown-Bayliss)
Date: 10 Dec 2000 19:12:02 -1200
Subject: Now for Modules
In-Reply-To: <20001210034847.A5103@glacier.fnational.com>
References: <Pine.LNX.4.21.0012071349420.22360-100000@relax.dk> <200012080813.eB88DWu01392@ZOOstation.cc> <20001208030745.B1931@glacier.fnational.com> <200012100734.eBA7Y5g12008@ZOOstation.cc>  <20001210034847.A5103@glacier.fnational.com>
Message-ID: <200012110712.eBB7C3t14920@ZOOstation.cc>

> See FAQ 6.15: 
> 
>     http://www.python.org/cgi-bin/faqw.py?req=show&file=faq06.015.htp
> 
> Cheers,

Thanks...  That seems to make much more sense... 

Now I have to master the concept of Modules.

are modules just like ordinary python progs, ie mymodule.py?


Do I just write a .py file with class and functions in it and do:

import * from mymodule.py

or is there a bit more to it?  

I have read the python tutorial tat comes with python, but it's a little
vague or my eyes were a little vague perhapse...





-- 

  Rob Brown-Bayliss
 ---======o======---
  www.ZOOstation.cc



From moshez at zadka.site.co.il  Sat Dec 16 18:31:55 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Sun, 17 Dec 2000 01:31:55 +0200 (IST)
Subject: while true: !!!
In-Reply-To: <3A3B8470.C5193211@cybermesa.com>
References: <3A3B8470.C5193211@cybermesa.com>
Message-ID: <20001216233155.2D6C0A82E@darjeeling.zadka.site.co.il>

On Sat, 16 Dec 2000, "Jay O'Connor" <joconnor at cybermesa.com> wrote:

> You need to define true=1; false = 0..
> 
> and then you get code like 
> 
> true = false      #exits the loop
> true = ! false 

I know this is tongue-in-cheek, but:

true == false
true != false

-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From aleaxit at yahoo.com  Sun Dec 24 17:30:30 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 24 Dec 2000 23:30:30 +0100
Subject: Some Python 2.1 ideas
References: <mailman.977590932.19335.python-list@python.org> <Cw616.23054$qN6.3308205@typhoon.we.rr.com> <3A452881.7686E717@uswest.net> <%vb16.187233$U46.5801214@news1.sttls1.wa.home.com> <3A464D5D.B541C04E@uswest.net> <mailman.977688488.25438.python-list@python.org>
Message-ID: <925tl301ub0@news1.newsguy.com>

"Thomas Wouters" <thomas at xs4all.net> wrote in message
news:mailman.977688488.25438.python-list at python.org...
    [snip]
> > > Wouldn't it be cleaner to enhance the file object with additional
> > > methods:
> > >
> > >   f. readline_chomp()
> > >   f.readlines_chomp()
>
> > This would be a good idea, except it wouldn't fit with the standard
Python idiom
> > for reading files:
>
> And everything that poses as a file object would have to duplicate those
> functions, all of a sudden. Guido is pretty conservative in adding
> conveniency functions to builtin (and oft-imitated) objects because of
that.

A mixin-class (implementing, e.g., readline_chomp in terms of .readline
and chomp, in this case) could also be supplied to ease the task of
'emulating convenience-methods'.  Not crucial in this case (the proposed
methods appear somewhat marginal), but an approach that might help
in other cases, I think.


Alex





From aleaxit at yahoo.com  Thu Dec 28 08:56:31 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 28 Dec 2000 14:56:31 +0100
Subject: COM / Variants / VT_BOOL
References: <mailman.977467384.30398.python-list@python.org> <91v2sv01rpp@news2.newsguy.com> <3A4B2F8C.4000909@ActiveState.com>
Message-ID: <92fgu001t2u@news1.newsguy.com>

"Mark Hammond" <MarkH at ActiveState.com> wrote in message
news:3A4B2F8C.4000909 at ActiveState.com...
> Alex Martelli wrote:
>
>
> > *Almost* right... but not *quite*, and we hit this specific problem
> > quite recently (we weren't driving with Python, but I think that
> > would have made no difference).
>
> Quite possibly it would.  If win32com had type information, it would
> have dont the right thing.

True -- because then calls would always go on the IDispatch side of
things; I had not considered this.


> > Setting the property .ValidateOnParse of the Xerces XML parser
> > (originally from IBM, now, I believe, from Apache), in its COM
> > setting, to 0 *or* 1, still keeps the parser non-validating; you
>
> Without type info, the parser should call VariantChangeType() to get a
> valid VT_BOOL back.  So really the parser is at fault (or the client for
> ignoring it)

Except that we were using the custom-side of things (from C++),
not the dispatch-side, and therefore passing, not a 'variant',
but a VARIANT_BOOL directly.  VARIANT_BOOL is a typedef for some
integral type (short int, I think), so, C++ happily accepts a
value of 1 and does no transformation on it.  Not sure if the
VariantChangeType is being done on Xerces' side (I guess they'll
be using a normal CreateStdDispatch to make the dispatch-side
from the custom-side, like everybody else, so the changetype
should be happening IF a call comes in from the custom-side).


> > *Most* COM servers will happily accept 'server.BoolProp = 1'
> > as setting the property to 'true', but, I suggest, don't count
> > on that; use 'server.BoolProp = -1' and you might be safer.
>
> Also, a feature of the pythoncom type conversion means that you can say:
>
> true = 1==1
> false = 1==0
>
> And passing these variables (or the expression as a literal) _will_
> force a VT_BOOL of the correct type.

...if makepy, or whatever, has given Python the needed typelibrary
information, that is; or, failing that, the server is implementing
the needed change-type correctly.

There may be one chance in a hundred that any given problem comes
from a combination of these safeguards all failing at once... but,
using '-1' to indicate 'true' is a cheap enough price to pay for
that 1%-risk of disaster insurance, I think:-).


Alex





From quinn at dinar.ugcs.caltech.edu  Thu Dec 14 14:59:58 2000
From: quinn at dinar.ugcs.caltech.edu (Quinn Dunkan)
Date: 14 Dec 2000 19:59:58 GMT
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <kuzoi15d60.fsf@lasipalatsi.fi> <Pine.BSF.4.10.10012122310510.48551-100000@tcfreenet.org> <u7l54ggyx.fsf@ctwd0143.fitlinxx.com>
Message-ID: <slrn93i9lt.cg1.quinn@dinar.ugcs.caltech.edu>

On 13 Dec 2000 16:16:06 -0500, David Bolen <db3l at fitlinxx.com> wrote:
>Aaron Jon Reichow <reichowa at tcfreenet.org> writes:
>
>> On 13 Dec 2000, Erno Kuusela wrote:
>> 
>> > the thing that struck me as a little foreign was the apparent
>> > isolation of the smalltalk image from the rest of the operating system
>> > and the filesystem. i want to be able to grep code!
>> 
>> While I'm not a proponent of Smalltalk isolationism, grep is a useless
>> tool (more or less), simply an artifact, ye, a symptom of the
>> sickness known as file-based development.
>
>I'm not sure I'd call it a "sickness" since any development model is
>just that - a model - it may be better or worse than some other model,
>but at least for me, a file-based model works well in many cases.
>
>My biggest problem with the integrated models (with database or other
>repository back-ends) tends to be that often they aren't yet fleshed
>out enough, or in many cases flexible enough, to replicate all the
>functionality I can get out of a file based system with a current
>collection of file based tools that have had lots of time to evolve
>for efficiency and flexibility.
>
>With file-based environments, I generally feel safe that if the
>environment itself doesn't support something, there are enough tools
>that understand the back-end repository (e.g., files) that I can get
>what I want even if I have to go outside the integrated system.  That
>may not be possible with a closed back-end repository.

I agree, and the reason I feel safe with a file system is that it's simple
enough that I can actually understand it.

It seems strange to hear of smalltalkers considering the filesystem obsolete,
because, to me, the filesystem is one of the most successful, practical, and
widely used OO systems around.  The original unix idea of "everything in text
files in a hierarchal filesystem" and then "lots of tools to modify text
files" is just simple enough to work for me.

Plan9 goes a long ways with open, close, read, and write, and the way the
whole system fits together feels very OO and "dynamic dispatch" to me.  If
grep is a useless tool, perhaps it's a sign that your (hypothetical "you")
development style is insufficiently file-based :)

I use grep every day.  Searching for text in text is a basic and useful
task... why reinvent it?


From joconnor at cybermesa.com  Tue Dec 12 11:23:56 2000
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Tue, 12 Dec 2000 16:23:56 +0000
Subject: Inheriting the @ sign from Ruby
 
Message-ID: <3A36511C.92846DB8@cybermesa.com>

While I'm not really thrilled by thePpython convention of requireing a
class method to explicetly declare "self", I find that better than using
"@"

I also use Perl a lot and get really tired of $_ and @_ as shorthand for
something in the current scope.  It's convenient, but not always clear

The last thing I want is for Python t obe more like Perl
-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

"God himself plays the bass strings first when He tunes the soul"



From delza at antarcti.ca  Wed Dec  6 12:18:57 2000
From: delza at antarcti.ca (Dethe Elza)
Date: Wed, 06 Dec 2000 17:18:57 GMT
Subject: rotate() in Tkinter
Message-ID: <5yuX5.761950$8u4.13165276@news1.rdc1.bc.home.com>

I'm trying to create a complex cursor in Tkinter.  I want to have an image
which I can rotate, change the color of, and move about.  I can build it as
a vector image, but I use ovals.  I haven't found any way to rotate images,
vector drawings, or ovals in Tk.

TIA

--Dethe




From ramb at synopsys.com  Mon Dec  4 13:41:38 2000
From: ramb at synopsys.com (Ram Bhamidipaty)
Date: 04 Dec 2000 10:41:38 -0800
Subject: int multiplication overflow and the future of python
Message-ID: <ghaelzoc9lp.fsf@synopsys.com>

On red hat 7.0 with python 1.6 I am running into overflow problems.

>From looking at the bug database at sourceforge I can see
that other people have run into this, and it looks like
there is a fix....but I have not been able to find it. Is there
a patch that I should apply to upgrade my 1.6 source to the
latest version?

This leads to a larger question: Which version is being maintained?
Is active developement taking place for both 1.6 and 2.0? I've been
using 1.6 because the features in 2.0 seem like they involve bigger
changes to the code base and I thought I should wait until the
initial bugs are worked out. But if all new work is only going
into the 2.x versions then I should probably stop using 1.6...

Any thoughts?

-Ram


From pkocmid at atlas.cz  Sun Dec 10 21:21:05 2000
From: pkocmid at atlas.cz (Petr Kocmid)
Date: Mon, 11 Dec 2000 03:21:05 +0100
Subject: Newbee seeking help! Extending Python with C/C++
In-Reply-To: <3A33D106.3F755E1@magnet.fsu.edu>
Message-ID: <DLEOLHIBOMHGMMPFOKGMCEPODEAA.pkocmid@atlas.cz>

www.boost.org Hot new stuff is Boost C++ Python Library from David Abrahams.

Petr Kocmid
pkocmid at atlas.cz

> -----Original Message-----
> From: python-list-admin at python.org 
> [mailto:python-list-admin at python.org]On Behalf Of Khaled Khairy
> Sent: Sunday, December 10, 2000 7:53 PM
> To: python-list at python.org
> Subject: Newbee seeking help! Extending Python with C/C++
> 
> 
> www.swig.org seems to be down or not responding. The links I got to
> through the search engines are to documentaion that is fairly old
> (Python 1.4 and rearlier). I have Active State Python 2.0 installed and
> need to extend it using a lot of C++ code that is already written. Pease
> advise.
> Khaled Khairy
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list



From emile at fenx.com  Fri Dec 29 12:51:53 2000
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 29 Dec 2000 09:51:53 -0800
Subject: Question re: eval()
References: <978109654.339714327@news.silcom.com>
Message-ID: <92iivt$6uk0r$1@ID-11957.news.dfncis.de>

What are you trying to do that using a dict doesn't address?

eg:

def n():
    return 3

print eval ('n + 1', {'n':n()})

g = {}
l = {}

exec ('m = 4',g,l)
exec ('a = 2',g,l)
print eval ('a + m',g,l)


--

Emile van Sebille
emile at fenx.com
-------------------


"Clarence Gardner" <clarence at netlojix.com> wrote in message
news:978109654.339714327 at news.silcom.com...
>
> I want to eval() an expression, but have some of the
variables in it be
> lazily evaluated.  E.g.,
>
> class D:
>     def __getitem__(self, name):
>         if name == 'n':
>             return 3
>
> d = D()
> x = eval('n + 1', d)
>
>
> but the eval function accepts only a dictionary as the
second parameter.
> Is there any amazing way to do this?
>
> TIA,
> Clarence Gardner
> clarence at netlojix.com




From sh at ttsoftware.co.uk  Tue Dec  5 06:43:51 2000
From: sh at ttsoftware.co.uk (Steve Horne)
Date: Tue, 05 Dec 2000 11:43:51 +0000
Subject: Calculating factorial
References: <3a2a9496.647995@uutiset.nic.fi> <90ghoh0lh@news2.newsguy.com>
Message-ID: <9tkp2tsfc0hjq3dr7vc69a6hdf0at3q0rk@4ax.com>

On Mon, 4 Dec 2000 17:45:46 +0100, "Alex Martelli" <aleaxit at yahoo.com>
wrote:

>Attempts to get more sophisticated, such as
>
>def factorial(f):
>    import operator
>    return reduce(operator.mul, range(1,f+1), 1L)
>
>will not necessarily bring happiness.

Might be faster with xrange (which doesn't build the whole list before
returning), but the simple for loop will probably still win for speed.
Functional-style list manipulations (once you are used to them) are
for speed of development, not speed of execution.

-- 
Steve Horne
Home : steve at lurking.demon.co.uk
Work : sh at ttsoftware.co.uk


From emile at fenx.com  Fri Dec 15 10:53:12 2000
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 15 Dec 2000 07:53:12 -0800
Subject: Accessing serial port on windows
References: <91dds1$b0i$05$1@news.t-online.com>
Message-ID: <3A3A3E68.30602@fenx.com>

Look for the python sio module.

Emile van Sebille
emile at fenx.com

Dennis Conrad wrote:

> Hello!
> 
> I'm "ogling" to move from Perl to Python as my "main scripting language" and
> wanted to learn a bit of it by writing my new project in Python. But I need
> to read data from the serial port (on Windows) and can't find any
> information on this on the net. Is there a possibility to do it? Or do I
> have to use the WinAPI?
> 
> Thank you very much
> 
> Dennis



From gregj at pobox.com  Sat Dec 23 19:22:23 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Sun, 24 Dec 2000 00:22:23 GMT
Subject: python and cgi
References: <9235p5$ngo$1@nnrp1.deja.com>
Message-ID: <3lb16.187135$U46.5800571@news1.sttls1.wa.home.com>

<etoffi at bigfoot.com> wrote in message news:9235p5$ngo$1 at nnrp1.deja.com...
> hello, fellow pythoners!
>
> i am writing an application that will communicate between two systems
> using http.  note that i am not an expert in http (or python ;)
>
> i need a way to "upload" using the HTTP POST method (ie: simulating a
> browser).

One easy way to do this:

1. Get your server-side page working with the form input and submit, so you
can use it from your browser.
For example, let's say the URL of the server-side page is
http://domain.com/formpage.html and it has three input fields: firstname,
lastname, and email. To make testing easy you should make the resulting page
(what happens after the form data is submitted) be a single line of text (no
HTML) that has a word or code (like 'ok' or 'fail').

2. Put your form fields into a Python dictionary, then use the urllib module
to encode the form data, open the URL, send the form data, and get the
result:

    import urllib

    url = 'http://domain.com/formpage.html'
    formdata = {'firstname': 'greg', 'lastname': 'jorgensen', 'email':
'gregj at pobox.com'}

    try:
        f = urllib.urlopen(url, urllib.urlencode(formdata))
        s = f.read()
        s = s.strip().lower()
    except:
        s = ''

    if s == 'ok':
        # success
    else:
        # fail

This is a very simplistic example but it will get you going. XML-RPC and
SOAP are more robust and scalable solutions.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From SBrunning at trisystems.co.uk  Tue Dec 19 04:21:41 2000
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Tue, 19 Dec 2000 09:21:41 -0000
Subject: while true: !!!
Message-ID: <31575A892FF6D1118F5800600846864D5B161C@intrepid>

> From:	Rainer Deyke [SMTP:root at rainerdeyke.com]
> 0 is false, None is false, [] is false, () is false, {} is false, and x is
> false where x is a class instance and x.__nonzero__() returns false.
> Everything else is true.  I think.
 
"" (an empty string) is also false.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From jbauer at rubic.com  Tue Dec  5 21:30:31 2000
From: jbauer at rubic.com (Jeff Bauer)
Date: Tue, 05 Dec 2000 20:30:31 -0600
Subject: Dennis Ritchie Interview
Message-ID: <3A2DA4C7.7FA78E4D@rubic.com>

Interesting comment from Dennis Ritchie in a LinuxWorld
interview:

  "What is changing is that higher-level languages are becoming
  much more important as the number of computer-involved
  people increases. Things that began as neat but small tools,
  like Perl or Python, say, are suddenly more central in the
  whole scheme of things."

http://www.linuxworld.com/linuxworld/lw-2000-12/lw-12-ritchie.html

Jeff Bauer
Rubicon Research



From tim.one at home.com  Thu Dec 21 19:03:50 2000
From: tim.one at home.com (Tim Peters)
Date: Thu, 21 Dec 2000 19:03:50 -0500
Subject: ?Why is 0e-6 invalid?
In-Reply-To: <3A428C7C.5CFBB181@email.sps.mot.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEALIFAA.tim.one@home.com>

[Norman Shelley]
> Why is 0e-6 invalid and 0*.000001 is valid?

The syntax of Python's floating-point literals is defined in the Language
Reference manual, here:

    http://www.python.org/doc/current/ref/floating.html

Some of the grammar productions referenced there are defined on earlier
pages.  As you can read there, the syntax Python allows for floats tries to
avoid confusion between prefixes of floats and octal integer literals.  So

    0e-6

is not allowed, and neither is

    01.1

nor

    00.0

Note that

    0.0
    0.
    .0
    0.e-6
    .0e-6
    0.0e-6

are allowed, though.

Continuing to ask "why?" beyond that is likely to be an unsatisfying hobby
<wink>.




From jp at ulgo.koti.com.pl  Mon Dec 25 08:46:41 2000
From: jp at ulgo.koti.com.pl (Jacek =?iso-8859-2?Q?Pop=B3awski?=)
Date: 25 Dec 2000 13:46:41 GMT
Subject: another newbie question - regex
Message-ID: <slrn94eju1.1fk.jp@localhost.localdomain>

Sorry, I have another question... 

"." - matches any character
"*" - is 0 or more repetition 

so ".*" - matches everything?

then why:

>>> print re.search(s,".*")
None
>>> print re.match(s,".*")
None

?

-- 
All eyes see the figure of the wizard
As he climbs to the top of the world
No sound, as he falls instead of rising                             "Stargazer"
Time standing still, then there's blood on the sand         - Ronnie James Dio


From r2d2 at montezuma.acc.umu.se  Thu Dec 14 03:08:04 2000
From: r2d2 at montezuma.acc.umu.se (Niklas Frykholm)
Date: 14 Dec 2000 08:08:04 GMT
Subject: Source Code Management
References: <fb4f3tsgffnmvme64aol1cn397te7ln349@4ax.com> <918711$meh$1@reader1.imaginet.fr> <prff3tc59qtv0qr9afj16dali3jv789c7u@4ax.com>
Message-ID: <slrn93gvv4.epv.r2d2@montezuma.acc.umu.se>

>I've just dug CVS out and downloaded it but I discover that the NT version
>appears to require a mounted volume.
>
>That's a disappointment.

No, you can run it remotely, just set the CVSROOT env variable to

	:server:user at host:/cvs/directory

If you install the lastest cygwin tools you will get not only cvs, but
also a ssh, so you can run cvs over ssh and get a secure connection.
That's what I'm doing and it works great.

I love to pretend my Windows machine is a Unix box.

// Niklas


From erno-news at erno.iki.fi  Sun Dec 10 10:52:30 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 10 Dec 2000 17:52:30 +0200
Subject: sgmllib nit
References: <eyErVLAEHlM6EwA4@jessikat.fsnet.co.uk> <EAsY5.2808$O%.435892@newsc.telia.net> <JagDRIAEbuM6EwxY@jessikat.fsnet.co.uk>
Message-ID: <kuitos8e9t.fsf@lasipalatsi.fi>

In article <JagDRIAEbuM6EwxY at jessikat.fsnet.co.uk>, Robin Becker
<robin at jessikat.fsnet.co.uk> writes:

| I realise that it does that! Unfortunately Zope's dtml isn't HTML.

afaik, dtml is not sgml either. it is more like "tag soup".

  -- erno


From ian.hobson at ntlworld.com  Sat Dec  9 17:40:30 2000
From: ian.hobson at ntlworld.com (Ian Hobson)
Date: Sat, 9 Dec 2000 22:40:30 +0000
Subject: Rummaging around inside an Access DB
References: <c5es2ts2l66bhg8qb7ne28cj0qa3gi9595@4ax.com>
 <hefs2t8tpsu3f9h746o84ns6sspa2p99dp@4ax.com>
Message-ID: <xcPQqJAeTrM6EwnL@ntlworld.com>

In article <hefs2t8tpsu3f9h746o84ns6sspa2p99dp at 4ax.com>, Dale
Strickland-Clark <dale at out-think.NOSPAMco.uk> writes
>Dale Strickland-Clark <dale at out-think.NOSPAMco.uk> wrote:
>
>>
>>I've been trying to dig some information out of an Access DB using Python. I 
>need to access as much
>>information as possible about the composition of all the reports. 
>>
>>After a few false starts and lots of head scratching (I can't find a help file 
>on the object model
>>anywhere!!), I found something in the Python Win32 distribution called 
>daodump.py, presumably
>>written by Mark H.

Just a thought. Wrox publish books on Word 2000 VBA and other office
products.  Do they produce something that would help? 

Wrox Press, Arden House, 1102 Warwick Rd, Birmingham B27 6BH. 

Sorry - no phone No. 

Regards

Ian Hobson 

Every time we teach a child something, we prevent him from inventing 
it himself. - Jean Piaget


From willybur at bellsouth.net  Sun Dec 10 18:46:40 2000
From: willybur at bellsouth.net (Wilbur)
Date: Sun, 10 Dec 2000 18:46:40 -0500
Subject: Multiple Threads & IRC
Message-ID: <3A3415E0.1080607@bellsouth.net>

Hi,
I am pretty new to Python, but am starting to get the hang of it.

Just to see if I can do it, I'm trying to write an IRC client, using Tk 
and Python.  However, I've run up against a few problems.

First,
It seems that I will need multiple threads running for this client.  One 
thread will have to be constantly checking my socket to see if the 
server has sent anything.  The other thread will deal with the user 
side, so he can input information.  The documentation on threads is very 
sparse; can someone briefly summarize how to use more than one thread in 
a program, especially when the threads must be maintained, must loop, 
need more than one line of code, and must all be able to send and 
receive data?

Second,
I'm going to need to use sockets.  I have, by taking bits and pieces 
from examples, been able to successfully connect to an IRC server, and 
receive a bit of data.
Here's the problem:
The server sends data in packets, and it sends whenever it can.  How can 
I get my client to be able to receive data from the server without 
knowing when it will send and what amount of data will be sent?  I only 
know how to tell Python to wait until it receives, say, 1024 bytes.  If 
the server doesn't send any more, how can I tell Python to stop waiting? 
  This is where I have come to the conclusion that I need to run this 
part in the background, while the rest of the program is running.  If 
Python can't run such tasks simultaneously, would I need to have Python 
just run a whole other program?  One would do the port listening and 
sending, and one would interface with the user.  The second program 
would command the first program around.

Thanks in advance,
Wilbur



From mfletch at tpresence.com  Wed Dec  6 14:24:11 2000
From: mfletch at tpresence.com (Mike Fletcher)
Date: Wed, 6 Dec 2000 14:24:11 -0500 
Subject: Finding values on sequences and .index()
Message-ID: <CC59868A0A76D311B1F50090277B321C1F3723@VRTMAIL>

The first approach below only works on Python 2.0, but looks cleaner.  It's
the same approach as the second, which works in just about any Python
version.  And yes, neither is really elegant, that's why people were calling
for the "indexing" keyword in for loops and/or list comprehensions :o) (Oh,
sorry for the spelling error in indices, I always do that and was too lazy
to go back and fix it today).

>>> def indicies( sequence, value ):
... 	return [ index for index in xrange(len(sequence)) if
sequence[index]== value]
... 
>>> indicies( a, 1 )
[0, 4, 8, 11]
>>> def indicies2( sequence, value ):
... 	result = []
... 	for index in xrange(len(sequence)):
... 		if sequence[index] == value:
... 			result.append( index )
... 	return result
... 
>>> 
>>> indicies2( a, 1 )
[0, 4, 8, 11]
>>> 


HTH,
Mike

-----Original Message-----
From: rhamorim at my-deja.com [mailto:rhamorim at my-deja.com]
Sent: Wednesday, December 06, 2000 1:15 PM
To: python-list at python.org
Subject: Finding values on sequences and .index()


Hi.

I have a list that may have many ocurrences of the same data. How do I
find the indexes for all ocurrences of it? list.index(value) only works
for the first ocurrence.

I thought of this ugly hack below, but I guess there's an easier way.

a = [1, 2, 3, 4, 1, 5, 6, 7, 1, 8, 9, 1]
b = map(None, a, range(len(a)))

N = 1 #value to be searched

c = filter(lambda x, n=N: x[0]==N, b)
d = map(lambda x: x[1], c) #this is the returned index list

Thanks in advance...

Roberto Amorim


Sent via Deja.com http://www.deja.com/
Before you buy.
-- 
http://www.python.org/mailman/listinfo/python-list



From fredrik at effbot.org  Mon Dec 18 05:05:30 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Mon, 18 Dec 2000 10:05:30 GMT
Subject: while true: !!!
References: <mailman.977127721.4398.python-list@python.org>
Message-ID: <Kjl%5.2$AH6.83@newsc.telia.net>

Blair Lowe wrote:

<rant>

> >The idiom is actually
> >
> >while 1:
> >
> > do something
> >     if condition:
> >        break
> >
> > do something else
> >
> >Basically, it's for dealing with situations where you can't tell at the
> >beginning or end of a loop if the loop is completed - only somewhere
> >in the middle. The canonical example is
>
> This is exactly what a python program should be NOT doing.

bullshit.

the "while 1" loop IS the standard Python idiom.

no matter what you think, that won't change.  get over it.

> "while 1" is one of the most cryptic statements in Computer
> Science.

what planet do you live on?

> It is cheap to define true and false once, or (my favourite) define
> ever (;;). Although the definition is cryptic, the resulting code is
> very readable.

really?  in what english language does

    "while true"

make any more sense than

    "while 1"

maybe you meant:

    "while true is true"

which, of course, is just plain silly.

:::

can we please kill this thread now?

instead of repeating the same stupid opinions over and over
again, do something useful.  there are lots of PEPs that could
need your input:

    http://python.sourceforge.net/peps/

if you find something to complain about in there, post to this
forum (but not to this thread, please).

(and if you really think that "while 1" is the biggest problem on the
planet right now, feel free to contribute a "how to write potentially
endless loops" PEP.  just make sure to include all points of view...)

</rant>

</F>




From rwgk at my-deja.com  Fri Dec 29 12:53:12 2000
From: rwgk at my-deja.com (rwgk at my-deja.com)
Date: Fri, 29 Dec 2000 17:53:12 GMT
Subject: Q: Python & Windows Enhanced Meta File (EMF)
Message-ID: <92ij24$m1h$1@nnrp1.deja.com>

Is there a way to create Windows Enhanced Meta Files from Python?

Or any other vector graphics format that can be imported in
Powerpoint?

Right now I am using Reportlab to create PDF files, and Photoshop
to convert these PDF files to GIF files. This is not only a time
consuming process, but of course the images consume a lot more
disk space. Being able to use Python to create vector graphics
for direct import into Powerpoint would be a big improvement.

Thanks!
        Ralf


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


From phd at phd.pp.ru  Tue Dec  5 06:57:14 2000
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 5 Dec 2000 14:57:14 +0300 (MSK)
Subject: setsockopt and urllib
In-Reply-To: <slrn92moue.hc.frank.neuber@mars.private.de>
Message-ID: <Pine.LNX.4.30.0012051456390.24848-100000@fep132.fep.ru>

On Mon, 4 Dec 2000, Frank Neuber wrote:
> I have to set special timeouts on receive and send calls in conjunction with
> urllib-calls.
> Is there a simple solution to do this without modifying the urllib itself.

   timeoutsocket.py on http://www.timo-tasi.org/python/

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




From jerry_spicklemire at my-deja.com  Sat Dec  9 12:18:28 2000
From: jerry_spicklemire at my-deja.com (jerry_spicklemire at my-deja.com)
Date: Sat, 09 Dec 2000 17:18:28 GMT
Subject: corba, soap, com, etc...
References: <pKsY5.58762$yR4.1496678@news1.rdc1.tx.home.com>
Message-ID: <90tph1$o21$1@nnrp1.deja.com>

Scott Hathaway wrote:

> If there is a corba or soap interface to python, can you please point
> me to it?

 http://www.uk.research.att.com/omniORB/omniORBpy/

 http://theopenlab.uml.edu/pyorbit/

 http://casbah.org/Scarab/

 http://www.pythonware.com/products/xmlrpc/

All of these, and more,  are listed at "the Vaults":

 http://www.vex.net/parnassus/apyllo.py/126307487.549099265

Have fun!
Jerry S.


Sent via Deja.com http://www.deja.com/
Before you buy.


From cjensen at bioeng.ucsd.edu  Mon Dec 11 20:51:01 2000
From: cjensen at bioeng.ucsd.edu (Curtis Jensen)
Date: Mon, 11 Dec 2000 17:51:01 -0800
Subject: function pointer
Message-ID: <3A358485.FC22A29@bioeng.ucsd.edu>

Is there a way to pass a pointer to a function from a module in a shared
library?  e.g. We have a compiled C module that has the function "foo"
in it (foo is not wrapped, it's straight C code).  We have another
copiled C module.  We want "foo" to be passed in to a function in the
second module as a paramater.  Is this possible?  Hope this makes sence.

-- 
Curtis Jensen
cjensen at bioeng.ucsd.edu
http://www-bioeng.ucsd.edu/~cjensen/
FAX (425) 740-1451


From qdlaty at wielun.dhs.org  Fri Dec 22 05:09:04 2000
From: qdlaty at wielun.dhs.org (QdlatY)
Date: Fri, 22 Dec 2000 11:09:04 +0100
Subject: saving dictionaries
References: <3A4300C7.FA4192AF@wielun.dhs.org> <mailman.977476326.4093.python-list@python.org>
Message-ID: <3A432840.11D0107C@wielun.dhs.org>

Hello!


Thanks for help!


Yours
QdlatY


From peter at schneider-kamp.de  Mon Dec 25 04:31:44 2000
From: peter at schneider-kamp.de (Peter Schneider-Kamp)
Date: Mon, 25 Dec 2000 10:31:44 +0100
Subject: One Python 2.1 idea
References: <mailman.977590932.19335.python-list@python.org>
	 <9237md$orm$1@nnrp1.deja.com> <924bjt0e3n@news1.newsguy.com> <925baq$5sj$1@nnrp1.deja.com>
Message-ID: <3A471400.69899E81@schneider-kamp.de>

rturpin at my-deja.com schrieb:
> 
> Stackless Python is a small but exciting step in the
> direction of improving infrastructure. I wish, instead
> of being a side project, this kind of work were now
> the core focus. And the coming feature enhancements?
> Well .. changing the scope rules for lambda is pretty
> limited in application. Creating a short-cut for
> range() in "for" statements is a step backwards, since
> it is yet another special case in the syntax that
> people must learn to understand the code. DON'T DO IT.

I see ome good reasons for enhancing the "for" statement:

- loop counter iteration will give a speedup by avoiding
  construction of a list and looping over that list.
- looping a certain amount of times is a frequent construct

Peter



From com at gahnstrom.se  Sat Dec 30 15:58:14 2000
From: com at gahnstrom.se (Tim Gahnström)
Date: Sat, 30 Dec 2000 21:58:14 +0100
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <3A4D6A05.EEC97FA6@uswest.net> <slrn94qqvg.mgg.grey@teleute.rpglink.com> <ije36.96813$x6.43604873@news2.rdc2.tx.home.com> <slrn94qt4f.mgg.grey@teleute.rpglink.com> <CMo36.97857$x6.44474719@news2.rdc2.tx.home.com> <slrn94s9qg.134.grey@teleute.rpglink.com> <gdq36.2696$Qb7.302774@newsb.telia.net> <slrn94shbe.150.grey@teleute.rpglink.com>
Message-ID: <92li6g$hju$1@yggdrasil.utfors.se>

>>next you gonna tell us that it's better to use flags than to use break
and/or
>>exceptions...
>
>    No, but I don't think that when you can place the test into the while
that
>one should use a while 1:break construct.


Chill folks, I am new to this group and I dont want a bad first impression
:-)

The original code was this

--
do :
    R=Themistor.ReadResistance()
    T=R_to_Temperature(R)
  until (T >=25)
--

The only way to acieve this exact behavior with is while is like this
-
a=1
while a=1 or (T >=25)
    a=0
    R=Themistor.ReadResistance()
    T=R_to_Temperature(R)
-

It is verry rare that we have to do it like this.
but the initial value of T might be used in ReadREsistance or in
R-to-temperature.

We have seen a few solutions that will be better in 99 cases of 100 but not
always.

Tim



From keisari_ at hotmail.com  Thu Dec  7 07:38:37 2000
From: keisari_ at hotmail.com (joonas)
Date: Thu, 07 Dec 2000 14:38:37 +0200
Subject: Printing modules contents
Message-ID: <3A2F84CC.3056C755@hotmail.com>

How can I make a script that would print modules contents like below,

sys
  sys.__name__ = "sys"
  sys.stdout = "<PyShell.PseudoFile instance at 7fd530>"
  sys.winver = "1.5"
  sys.etc = "spam"


Joonas.


From Blair.Lowe at compeng.net  Wed Dec  6 23:36:42 2000
From: Blair.Lowe at compeng.net (Blair Lowe)
Date: Wed, 6 Dec 2000 21:36:42 -0700
Subject: XML-RPC was [pygtk] Python and Orbit Examples
In-Reply-To: <20001206194117.D12869@linfoo.eink.com>
References: <Pine.LNX.4.21.0012050828550.24722-100000@tupelo.daa.com.au>
 <200012060632.eB66WIK27614@ZOOstation.cc>
 <20001206165355.B12869@linfoo.eink.com>
 <v04220802b654744d4fc6@[192.168.11.33]>
 <20001206194117.D12869@linfoo.eink.com>
Message-ID: <v04220803b654c258a639@[192.168.11.33]>

Thanks Pehr,

At 19:41 -0500 2000/12/06, Pehr Anderson wrote:
>Dear Blair,
>
>I'd recommend doing XML over a simple socket, without relying on
>HTTP or SSL as a transport protocol, but that would require you
>to include a security model within your XML specification.

Not my choice. The vendor decides this. I just have to connect and 
pass XML stuff back and forth. Probably good advice for a vendor, 
though.

>
>It's not a bad idea to handle security separately from transport
>unless you are set on the SSL security model.  It would be
>perfectly reasonable to do a HTTP POST with your XML querry
>and wait for the reply as an XML document, then all you need
is the SSL interface for python.

>
>Python 2.0 includes support for SSL in its web libraries,
>or you could try to do something with SSL by calling
>lynx using os.popen("lynx --source https://...").read()

I shall look into python 2.0. We are using 1.5 right now.

Pardon my ignorance, but would I connect to the ssl site with a POST 
context of XML/SSL, or something? How do I get the other site ready 
to accept XML? Headers? Where can I find more information on this 
stuff (ie. what book(s), where are docs on line).

Thanks,
Blair.

>
>         -pehr
>
>On Wed, Dec 06, 2000 at 04:05:03PM -0700, Blair Lowe wrote:
>>  Sounds cool.
>>
>>  I found pyXML. at http://download.sourceforge.net/pyxml/.
>>
>>  Can one send XML to an SSL site with pyXML as well? If so, where is
>>  some good docs for this. The version I got (PyXML-0.6.2-2.0.i386.rpm)
>>  has not docs, and the overview at 4suite say I need to download
>>  something else.
>>
>>  Are there some better tools for doing this. I started to look at
>>  Zope, but it was rather cumbersome and knowing where to start with
>>  the documentation was not easy to find either.
>>
>>  TTYL,
>  > Blair.
>>
>>  At 16:53 -0500 2000/12/06, Pehr Anderson wrote:
>>  >Dear Folks,
>>  >
>>  >I would highly recommend using XML-RPC as a minimally complicated
>>  >way to communicate data between your apps.
>>  >
>>  >This might sound complicated but it really only means
>>  >	1. specify XML encodings for your data, use these throught your code
>  > >	2. open a socket on the target
>>  >	3. send an XML query and wait for an XML response
>>  >	4. parse the response
>>  >
>>  >The beautiful thing about XML-RPC (remote proceedure call) is
>>  >that it is just dirt simple. You open a socket, dump in
>>  >human readable text, and wait for a human-readable response.
>>  >Everything gets completely defined somewhere and servers and
>>  >clients could be coded in *anything*!
>>  >No complicated tools are required to implement any parts of
>>  >this system though you might want to look at how others
>>  >parse XML rather than doing it all by hand.
>>  >
>>  >And the best part is you'll have an app that anyone can understand.
>>  >No secret church of CORBA, COM or other "special tools".
>>  >XML is directly comprehensible by HUMAN BEINGS and that makes it
>>  >a sustainable encoding with simple tcp sockets for transport.
>>  >
>>  >	-pehr
>>  >
>
>  > >
[snip]
Computer Engineering Inc. http://www.compeng.net
Phone: 780 499 5687 (9 - 5 MST) Fax:   780 435 0693 (24 Hours)



From claird at starbase.neosoft.com  Tue Dec 26 16:28:29 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 26 Dec 2000 15:28:29 -0600
Subject: sniffing btw my web browser and a server
References: <92aush$rhc$1@news.nuri.net>
Message-ID: <270767A74DE6B86D.ED3C49CF7CD80055.4014344C6ECF85B8@lp.airnews.net>

In article <92aush$rhc$1 at news.nuri.net>,
June Kim <junaftnoon at nospamplzyahoo.com> wrote:
>I'm doing some sort of web client programming, and have to observe the
>messages that my browser and web-servers transmit, being tired of guessing
>what the messages'd be and looking up the references and the specifications.
>
>Anyone has done it in Python or knows how it could be done?
>
>
>Best regards,
>
>June
>
>ps. using TCP/IP packet sniffers made for mere hacking seems too primitive
>or low-level in this case.
No.  I don't know of a Python answer to what I understand
to be your question.  There's a quite nice Tcl-based sol-
ution:
<URL:http://www.neosoft.com/tcl/ftparchive/sorted/net/sockspy-1.0/1.0/sockspy-1.0.README>

Do you want Py-written Web proxy servers and clients?  Would
it be enough (or best) for you to learn how to instrument
the server and client you're using so they report on their
own dialogues?

I'm sympathetic with your sentiment that conventional packet-
sniffers are too crude.
-- 

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


From china at the.web.com  Tue Dec  5 21:26:49 2000
From: china at the.web.com (Albert Chin-A-Young)
Date: 6 Dec 2000 02:26:49 GMT
Subject: urllib2, proxy, basic authentication
References: <90hk8k$fqh$1@bob.news.rcn.net> <3A2D61C1.9182431D@motorola.com>
Message-ID: <90k859$adl$1@bob.news.rcn.net>

Jeremy Lamplough <jeremy.lamplough at motorola.com> wrote:
> try setting the proxy host prior to importing the other modules.

> import os, sys
> os.environ['http_proxy'] = 'http://[PROXY_HOST]:5865'
> import urllib2
> from urllib2 import urlopen

> The way i'm doing basic stuff, with NO authentication is (to download a
> stock quote)

We don't have a problem with http_proxy and the remote host not
requiring authentication. However, if the remote host does require
authentication, then we have our problem.

-- 
albert chin (china @at@ thewrittenword .dot. com)


From georgek at netwrx1.com  Sun Dec 24 10:09:08 2000
From: georgek at netwrx1.com (George R. Kasica)
Date: Sun, 24 Dec 2000 09:09:08 -0600
Subject: Python problem
References: <Pine.LNX.4.21.0012181646240.20344-100000@saturn.ispol.com>
Message-ID: <0h3c4tc7u5s650e04grogmkb6l585meim8@4ax.com>

Hello:

Here's the problem I'm having...any help greatly appreciated:

I've installed Python 2.0 here under linux 2.2.18 and recently have
started to use a software package called apache toolkit which builds
the various modules of the apache web server (one of them being Mod
Python 2.6.3) When I attempt to add this option I receive the
following:

>Python detected but with threads support enabled!
>
> Please recompile Python 1.5.2+ without threads. You can get it from\
>  "http://www.python.org/ftp/python/2.0/BeOpen-Python-2.0.tar.gz".
>  When running ./configure use the "--without-threads" option.
>-----------------------------------------------------------------------

Here's the problem:

I've used make clobber in the Python directory and even totally
removed the directory and reinstalled from the tar.gz file and run the
./configure --without-threads option but it still comes up like
this....

Where and what else must I get rid of to start over from zero? I
didn't use any sort of optioning besides that shown so the file
locations should be "standard" per the Makefile.

Thank you,

===[George R. Kasica]===        +1 262 513 8503
President                       +1 206 374 6482 FAX 
Netwrx Consulting Inc.          Waukesha, WI USA 
http://www.netwrx1.com
georgek at netwrx1.com
ICQ #12862186


From boud at rempt.xs4all.nl  Fri Dec 29 15:11:16 2000
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 29 Dec 2000 20:11:16 GMT
Subject: Python for prototyping of the user interface
References: <20001228.17165633@mis.configured.host>
Message-ID: <92ir54$ftn$1@news1.xs4all.nl>

Volkmar Busch <vbusch at gmx.de> wrote:
> Hi,

> I took my first steps in using Python, and I really like the language. I 
> read a lot about using Python for prototyping. As a basis for discussion 
> with our users we use GUI prototypes. Building complex user interfaces in 
> Python to me seems quite complicated, because

> - I am missing a Gui builder
> - I am missing controls like notebooks and tables

> So I'd like to asked the experienced Python users how they solve this 
> problem.

I use PyQt - Qt comes with Qt Designer, a great gui designer that 
generates xml descriptions of your design; this then is translated
by pyuic to python. I subclass the result (which means I can 
regenerate whenever I want), and use it. Quite simple, really. And
all modern conveniences are available in PyQt, from icon boxes to
Unicode support. And to my surprise, the large PyQt application I've
just finished works without changes on Windows...

-- 

Boudewijn Rempt  | http://www.valdyas.org


From emile at fenx.com  Fri Dec 15 10:58:52 2000
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 15 Dec 2000 07:58:52 -0800
Subject: Python advocacy
References: <91bln3$lmr$1@nnrp1.deja.com> <91d63k0msl@news1.newsguy.com> <8iak3t4a3lrqcqp9iqftmekdoahuico2b0@4ax.com>
Message-ID: <3A3A3FBC.1090206@fenx.com>


Steve Horne wrote:

> On Fri, 15 Dec 2000 14:24:27 +0100, "Alex Martelli"
> <aleaxit at yahoo.com> wrote:
> 
>> Makes me wonder why Larry Wall didn't think to include at least
>> ONE strangeness, wart, or inconsistency *somewhere* in Perl...
>> thus making it so boringly perfect and un-idiosyncratic as to
>> leave its advocates without a suitable 'scapegoat feature'!
> 
> 
> Erm - either you haven't used Perl, or your an advocate yourself!
> Personally, if I was selling Python, Perl would be involved in my
> scapegoat manouvre.

I expect Alex had his tongue firmly planted in cheek for that part of 
his post.

But-I-could-be-wrong-ly y'rs

Emile van Sebille
emile at fenx.com



From schaffer at optonline.net  Mon Dec 25 11:48:22 2000
From: schaffer at optonline.net (Les Schaffer)
Date: 25 Dec 2000 11:48:22 -0500
Subject: Can't load jpeg image with PIL (win32)
References: <9241ls$ks3$1@news1-2.kornet.net>
Message-ID: <874rzsh2g9.fsf@optonline.net>

> Do I have to install other components like IJG jpeg library? Please
> let me know how I can read and handle jpeg images with Python on
> Windows platform.

you need a version of PIL with jpeg support compiled in. i have
created some packages for Windows and MS VC++ if you want to build one
yourself or if you just want pre-compiled modules and jpeg library,
see:

http://folks.astrian.net/godzilla

les schaffer


From rturpin at my-deja.com  Mon Dec 25 12:43:46 2000
From: rturpin at my-deja.com (rturpin at my-deja.com)
Date: Mon, 25 Dec 2000 17:43:46 GMT
Subject: One Python 2.1 idea
References: <mailman.977590932.19335.python-list@python.org> <9237md$orm$1@nnrp1.deja.com> <924bjt0e3n@news1.newsguy.com> <925baq$5sj$1@nnrp1.deja.com> <mailman.977737448.26875.python-list@python.org> <927q24$ond$1@nnrp1.deja.com> <tTK16.73794$x6.34384327@news2.rdc2.tx.home.com>
Message-ID: <92810h$t9n$1@nnrp1.deja.com>

In article <tTK16.73794$x6.34384327 at news2.rdc2.tx.home.com>,
  "Rainer Deyke" <root at rainerdeyke.com> wrote:
> _range = range
> ranges = []
>
> def my_range(*args):
>   r = _range(*args)
>   global ranges
>   ranges.append(r)
>
> __builtins__.range = my_range

Optimizing generated byte code for a language like Python
has to be done somewhat differently than for a language
like C++, which is designed so that types and names
resolve at compile time. There are a variety of strategies
for handling the code above. A simple one would flag for
recompilation any file where certain kinds of optimization
had been applied, when a built-in function were replaced.
Likely, there are better strategies.

Python presents other challenges besides the fact that
built-in functions can be changed on the fly. Fortunately,
almost all of these have been fodder, two or three times,
in the history of language processing. The dynamic and
flexible nature of Python is no excuse for slow execution.

Russell


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


From prestonlanders at my-deja.com  Thu Dec 21 11:03:03 2000
From: prestonlanders at my-deja.com (Preston Landers)
Date: Thu, 21 Dec 2000 16:03:03 GMT
Subject: mail processing module?
References: <91rcig$rgu$1@nnrp1.deja.com> <91sqer$tsn$1@nnrp1.deja.com>
Message-ID: <91t9jh$a1e$1@nnrp1.deja.com>

In article <91sqer$tsn$1 at nnrp1.deja.com>,
  fredrik at pythonware.com wrote:
> Preston Landers improves the signal/noise ratio:
> > I'm trying to find out if there's any interest in a
> > kitchen-sink Python module for mail handling. /.../
> > It uses the standard Python smtp/pop/imap/MIME modules,
> > just basically ties them all together.
>
> I sure wouldn't mind seeing a "maillib.py" module in the
> next Python release (similar to "urllib", which also pro-
> vides an easy-to-use layer on top of httplib/ftplib etc).
>
> so if you decide to release it, don't forget to post a
> copy to the patch manager.

An interesting idea.  I will see what I can do to make it worthy of
inclusion in Python.  I will post here when more information is
availible.

thanks
Preston


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


From grey at despair.rpglink.com  Wed Dec 13 18:32:13 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Wed, 13 Dec 2000 23:32:13 -0000
Subject: Moving from Perl to Python
References: <3a3a9552.76673185@news.tcn.net>
Message-ID: <slrn93g1ns.m5n.grey@teleute.rpglink.com>

On Wed, 13 Dec 2000 15:40:04 GMT, Richard Zilavec <rzilavec at tcn.net> wrote:
>I just got back from two book stores, nothing on Python, I was
>shocked!  So I went online and found a pile but I hate ordering
>online, I like to browse through them first.  Anyways, I really like
>O'Reilly books, is the Programming Python really advanced or is more
>along the lines of the Programming Perl where it covers the basics to
>advanced?

    I didn't care much for Programming Python.  In all the ways that
Programming Perl is good Programming Python is not.  The former makes an
excellent reference book, the latter does not.  Personally I took the major
plunge into Python by attending the talk put on by David Beazley at the
O'Reilly/Perl conference in Monterey earlier this year.  Bought his book,
_Python: Essential Reference_, based on that talk and have been quite happy
with it.  In fact, I'm upset that I lost my copy somewhere along the way.  :/



-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From aleaxit at yahoo.com  Tue Dec 26 13:43:52 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 26 Dec 2000 19:43:52 +0100
Subject: One Python 2.1 idea
References: <mailman.977771949.17911.python-list@python.org> <9292tn$jka$1@nnrp1.deja.com> <929n9c013m3@news1.newsguy.com> <92ab9u$dou$1@nnrp1.deja.com>
Message-ID: <92ap0i22b6r@news1.newsguy.com>

<rturpin at my-deja.com> wrote in message news:92ab9u$dou$1 at nnrp1.deja.com...
> In article <929n9c013m3 at news1.newsguy.com>,
>   "Alex Martelli" <aleaxit at yahoo.com> wrote:
> > No implicit caching.
>
> Has there been some discussion about caching? If done
> well, the vast majority of references would reach
> their object's attribute or method for the cost of a
> couple of integer compares and array fetches. This

Yep -- caching is good.  No doubt about that.

> should be an order of magnitude faster than string
> lookup in a dictionary. This especially benefits
> method invocation, since that costs a failed lookup
> to the instance dictionary, and a successful lookup
> in the class dictionary. That's almost enough to
> scare me away from methods!

Looking up builtin functions is two dictionary lookups
too (failed in globals, successful in builtins).  But for
methods it can be more than 2 (lookup in bases).


> Most work in a Python program involves object methods
> or attributes. (At least, this is the case in my
> programs. ;-) Increasing the speed of reference to
> these should have appreciable benefit, across the
> board. Yeah, it would make the code a bit more
> complex to maintain. But this isn't a special case --
> it is a general mechanism central to how Python works.

It does seem likely that caching lookups is where
most across-the-board performance might be gained,
yes.  The trick is finding out a way to invalidate the
cache every time some dynamic manipulation may
need lookups to happen afresh.


Alex





From rturpin at my-deja.com  Fri Dec 29 20:58:10 2000
From: rturpin at my-deja.com (rturpin at my-deja.com)
Date: Sat, 30 Dec 2000 01:58:10 GMT
Subject: Python vs. PHP (& Java?)
References: <69BC4C6AE83ED311BC9400805FA7B14AF84D2C@il93exp01.css.mot.com> <mailman.978018791.20357.python-list@python.org> <92fv4a$l2l$1@nnrp1.deja.com> <92gghv02iq1@news2.newsguy.com> <92iie9$lg3$1@nnrp1.deja.com> <92j7q0033j@news1.newsguy.com>
Message-ID: <92jffg$cs1$1@nnrp1.deja.com>

I wrote:
>> .. If you want to look for adaptation, sexual selection
>> may have more to do with this than natural selection. A
>> silver tongue will not save you from the lion or bag
>> the deer, but it will win a heart.

In article <92j7q0033j at news1.newsguy.com>,
  "Alex Martelli" <aleaxit at yahoo.com> wrote:
> Sexual selection, particularly female-choice, IS very much
> a part of natural selection ..

I had the notion that Darwin distinguished the two. Not to
imply that sexual selection is unnatural, but simply to
differently label the two mechanisms. But I could remember
wrong; I am no biologist.

Russell


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


From tim.one at home.com  Sun Dec 31 07:18:40 2000
From: tim.one at home.com (Tim Peters)
Date: Sun, 31 Dec 2000 07:18:40 -0500
Subject: Is 100,000 entries big for a dictionary?
In-Reply-To: <0p936.2528$Qb7.277261@newsb.telia.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEPNIFAA.tim.one@home.com>

[Fredrik Lundh]
> The implementation limit is currently one billion objects.
>
> If you plan to store more objects, you need to add more
> entries to the "polys" table in Objects/dictobject.c.

It's not that easy:  the table is already as big as it can be on a 32-bit
machine.  Well, we could add one more entry if we changed the polys decl
from signed to unsigned.  If you want to do that much, here's the magic line
to insert:

	2147483648 + 9,

Beyond that, it would be exceedingly painful.

not-to-mention-exceedingly-slow-on-most-current-boxes-ly y'rs  - tim




From howard at eegsoftware.com  Fri Dec 15 18:52:02 2000
From: howard at eegsoftware.com (howard at eegsoftware.com)
Date: Fri, 15 Dec 2000 23:52:02 GMT
Subject: mmap problem with dll
References: <3a366959.3497335@news-server.socal.rr.com> <3A37F4CC.5070305@ActiveState.com>
Message-ID: <3a3aae65.26647362@news-server.socal.rr.com>

On Wed, 13 Dec 2000 22:19:54 GMT, Mark Hammond <MarkH at ActiveState.com>
wrote:

>howard at eegsoftware.com wrote:
>
>> I get the message 
>>      EnvironmentError: [Errno 2] No such file or directory
>> on the mmap object creation call.
>> 
>> I traced it back to the _get_osfhandle call trying to get the
>> operating system equivalent of the run-time library handle.
>> 

>You almost certainly are compiling without the "/MD" ("/MDd" for debug) 
>compiler option (aka "Runtime in multi-threaded DLL").  The symptoms are 
>very much like the DLL and Python itself arent sharing the same runtime, 
>and so can't pass CRT handles or abjects between them.
>
>Mark.
>

Exactly correct.   Thank you.  It seems that the project I copied this
from was from a Python 1.4 baseline and I didn't check every option.

Howard Lightstone
EEGSoftware
howard at eegsoftware.com


From scarblac at pino.selwerd.nl  Tue Dec 12 19:19:34 2000
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 13 Dec 2000 00:19:34 GMT
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <ku4s098oxi.fsf@lasipalatsi.fi>
Message-ID: <slrn93dg04.8en.scarblac@pino.selwerd.nl>

Erno Kuusela <erno-news at erno.iki.fi> wrote in comp.lang.python:
> In article <slrn93cotl.80c.scarblac at pino.selwerd.nl>,
> scarblac at pino.selwerd.nl (Remco Gerlich) writes:
> 
> | Integers, longs, complex numbers, strings, functions, tuples, types, floats
> | and None are immutable. Most of these are obvious; the number 3 is the
> | number 3, why would you change that?!
> 
> it is easy to think that when you know how things work in python.
> but, you could just as easily say, "a list containing the numbers 1, 2 and
> 3 is a list of 1, 2 and 3, why would you change that?!"

Yes, I realized that and therefore said "most of these...".

And it felt obvious at the time, a tuple is something I know from math
just like an integer, and a list is a pragmatic changing thingy that you
want to be able to change. But there have been a few other posts in the
thread that have made it less obvious, and I now see that a mutable int
can be useful; something like a UserInteger class besides the normal ints.
And integer literals would always be the immutable type so that 2+2 is
always 4, of course.

I'm now thinking that a mutable list type shouldn't have a literal, but
always be built from the immutable literal (since literals are by themselves
immutable, so a literal for a mutable type is strange). 
Isn't that what Perl does?! (just an idea, i'm not saying that this
is a good idea ;)).

But I still feel that the current Python situation is rather simple, and
that it only looks complicated when people try to fit concepts from other
languages to it or when people like me try to explain it too much :).

-- 
Remco Gerlich


From jh at web.de  Mon Dec  4 17:42:28 2000
From: jh at web.de (Jürgen Hermann)
Date: Mon, 4 Dec 2000 23:42:28 +0100
Subject: [ANN] MoinMoin Release 0.6
Message-ID: <90h6kq$h24$02$1@news.t-online.com>

A WikiWikiWeb is a collaborative hypertext environment, with an emphasis on
easy access to and modification of information. MoinMoin is a Python
WikiClone that allows you to easily set up your own wiki, only requiring a
Web server and a Python installation.

The major new feature of this release is the introduction of user
preferences. This is currently used to display your name on the
RecentChanges
page (instead of your IP), and show all dates in your local time.

Besides internal (code organization) changes, the two minor additions are
a TableOfContents macro (much better navigation for longer pages), and
improvements for numbered lists.


Homepage:
    http://moin.sourceforge.net/

Download:
    http://download.sourceforge.net/moin/

Mailing lists:
    http://lists.dragon-ware.com/mailman/listinfo/moin-users
    http://lists.dragon-ware.com/mailman/listinfo/moin-dev


New features:
    * [[UserPreferences]] macro and associated functions
    * [[TableOfContents]] macro
    * Mechanism for external macros (user extensions)
    * Numbered list types and start offsets

Bugfixes:
    * Search dialogs did not work on the FrontPage
    * Add newline to text if last line has none (better diffs)





From MarkH at ActiveState.com  Sat Dec 30 19:38:54 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sun, 31 Dec 2000 00:38:54 GMT
Subject: Komodo
References: <B672A66F.72CC%bobhicks@adelphia.net> <3A4D6D0B.20000@ActiveState.com> <92lrsg$2pa$1@nnrp1.deja.com>
Message-ID: <3A4E7EFD.4090303@ActiveState.com>

fgleeson at my-deja.com wrote:

>> * For those who can't infer the obvious, I'm part of the Komodo dev team :-]
> 
> I know this is off topic, but since you are on the Komodo dev team I can not
> resist. 

Eeek.  Well, since I started it, I guess I better respond!

> Komodo looks like a really nice product, which I would like to use.
> However it seems to be missing some essential features for a perl/python IDE.
> 
> For instance typically in perl\python scripts you need to pass in some
> parameters to control the scripts behavior. If I am running from the command
> line its simple perl myscript.pl opt1 opt2 etc..
> 
> I can not find a way to do this. Launching komodow c:\scripts\myscript.pl
> opt1 Komodo can not find my script even if I cd to the directory containing
> my script.

If you select "Preferences", and select the debugger pane, you will see 
a "Prompt for arguments" checkbox.  (I can't recall the exact text).  I 
believe the default is "on", so you should already be seeing it.

When you start a debug session with this option selected, a dialog box 
will be presented allowing you to choose a number of settings for the 
debug session, including the command-line.

This was added between the last "tech preview" and the most recent Beta 
- are you sure you are looking at the latest?


> By the way what is the difference between
> komodo.exe and komodow.exe?


Not much ;-)  There is still a fair bit of stuff in the distribution we 
need to cleanup.  There is also obviously still lots of work to be done. 
  However, I am quite excited by the possibilities that the Mozilla 
framework and Python can offer, and quite excited where Komodo should be 
able to end up.  Even the Netscape/Mozilla guys sounded suitably 
impressed ;-)

And as a final teaser, the Python/XPCOM bindings should be (freely, 
openly - probably same licence as Moz) released in a week or so.  The 
ball is firmly in my court but I am travelling so some really cold place 
over the next few days <wink> - so my current best guess is 7 days...

Mark.
-- 
Not speaking for ActiveState in this or any post!  If I got something 
wrong my cat walked across my keyboard.  I don't even have a cat - in 
fact, I hate them.  My dog is not allowed on the keyboard, so you can't 
blame her.  No one wants to take responsibility any more.  Happy new year!



From ennisr at my-deja.com  Tue Dec 26 11:58:35 2000
From: ennisr at my-deja.com (Ennis Rosamond)
Date: Tue, 26 Dec 2000 16:58:35 GMT
Subject: python telnet script for openvms host
Message-ID: <92ainp$jab$1@nnrp1.deja.com>

I am attempting to develop a Python telnet script to access
an OpenVMS host.

The script is partially successful in that it logs on to the
target host. I know this because following acceptance of the
user password by the host, the host sends back the same header
message as is received in an interactive telnet session.  The
problem I am having is that the host doesn't seem to be sending
back a '$ ' prompt after the session header; at least I haven't
seen it.  Consequently, the host doesn't respond to any commands
sent by the script.
(I have been trying a simple '$ dir' command.)

My test script is listed below.

#python test script:
import sys, telnetlib
host = 'hostname'
user = 'username\r'
password = 'mypasswd\r'
tn = telnetlib.Telnet(host)
tn.read_until('Username: ')
tn.write(user)
tn.read_until('Password: ')
tn.write(password)
tn.read_until('[c')    #last char of session hearder
tn.write('$ dir\r')

#following this, I receive back the sequence:
#recv '\377\373\001'
#IAC WILL 251
#recv '\377\373\003'
#IAC WILL 251
#recv '\003\\\033Z'
#
#which I don't know how to interpret.

Anyone have an idea about how to determine what I am doing
wrong?  Could the problem have something to do with the
terminal type expected by the OpenVMS telnet host?

I am posting this to both the python and openvms discussion
groups.  Any help greatly appreciated.


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


From db3l at fitlinxx.com  Wed Dec 13 16:02:18 2000
From: db3l at fitlinxx.com (David Bolen)
Date: 13 Dec 2000 16:02:18 -0500
Subject: [Q] Are Exceptions used that much in practice?
References: <3A35D546.A4731B5B@enteract.com> <914mfd02ibt@news2.newsguy.com> <3A36F20B.7B766D0@cosc.canterbury.ac.nz>
Message-ID: <uaea0ghlx.fsf@ctwd0143.fitlinxx.com>

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

> A minor gripe here - it would be more useful if Python
> had some subclasses of IOError to distinguish between
> different reasons for failure, so you can catch them
> more selectively. If the reason is "the file doesn't
> exist", you might want to do something to make it exist
> and then try again. But if the reason is "the file exists
> but the user doesn't have permission" or "someone put 
> peanut butter in the disk drive", you probably just want 
> to let the exception propagate.

I suppose one of the issues is constructing a reasonable class
hierarchy that would portably represent the common failure mores.

It's not necessarily portable, but even if you can't distinguish at
the exception level, there's nothing stopping the exception handling
code from examining the associated object with the exception, and if
the code doesn't match the case it wants to handle, just using "raise"
to re-raise the same exception and let it continue propagating.

In the case of IOError that does imply parsing a string and assuming
it starts with [Errno #], which is sort of crufty.  But other
exceptions (such as those from win32all) are fairly standardized as a
tuple with the code as the first element - something I also tend to do
with my own user exceptions in my code.

Thus, you can get a reasonable mix in terms of design between the
exception class hierarchy and then the finer grained specific failure
within a class of exceptions.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From daniel at dittmar.net  Sun Dec 31 11:48:52 2000
From: daniel at dittmar.net (Daniel Dittmar)
Date: Sun, 31 Dec 2000 17:48:52 +0100
Subject: SMB Module
References: <92n1lh02evv@drn.newsguy.com>
Message-ID: <3A4F6374.1A6C549F@dittmar.net>


> 
> Makes no sense, because a SMB client mapps normaly a M$ volume to a directory
> tree (UNIX) or a virtual drive (M$) ... I don't see in that case any tasks for
> Python.

Makes the same sense as ftplib: you want to access files on a host which
exports them in a specific way.

Daniel Dittmar


From aahz at panix.com  Mon Dec 18 12:39:15 2000
From: aahz at panix.com (Aahz Maruch)
Date: 18 Dec 2000 09:39:15 -0800
Subject: Timers in Python?
References: <3a3e42ef.16144323@news.skynet.be>
Message-ID: <91li43$gpc$1@panix3.panix.com>

In article <3a3e42ef.16144323 at news.skynet.be>,
Daniel <Daniel.Kinnaer at AdValvas.be> wrote:
>
>Is there an equivalent of the Delphi Timer-component in Python as
>well?  That is, I want to have two Timers : Timer1 needs to activate
>itself every 2 minutes and Timer2 needs to activate itself every 7
>minutes. 

There's nothing built-in.  If you're using Tkinter (and I assume other
GUI libraries), you can set up "alarms"; otherwise, I suggest you spawn
off a timer thread.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

This is a signature anti-virus. 
Please stop the spread of signature viruses!


From pearu at ioc.ee  Sat Dec 30 08:04:23 2000
From: pearu at ioc.ee (Pearu Peterson)
Date: Sat, 30 Dec 2000 15:04:23 +0200
Subject: Lazy Python usage (hack request)
Message-ID: <Pine.LNX.4.10.10012301443330.13756-100000@kev.ioc.ee>

Hi,

Is it possible to achive the following behavior in Python?

>>> import lazy
>>> a = b
No name 'b' defined, using default 777
>>> print a
777

Note that changing the python session above is not an option. (Also 'No'
for the answer;-)

Just to avoid questions 'Why do you need it?', I would use it in
PySymbolic (but not in its Kernel!) for defining new Symbol instances in
fly (it's very tedious to define 'b = Symbol("b")',etc all the time).

Here lazy.py would be something like the following:
#Begin of lazy.py

import UserDict,sys

class myDict(UserDict.UserDict):
    def __getitem__(self, key):
        if not self.data.has_key(key):
            print 'No name %s defined, using default 777'%(`key`)
            self.data[key] = 777
        return self.data[key]

sys.modules['__builtin__'].__dict__ = myDict(sys.modules['__builtin__'].__dict__)

#EOF lazy.py

When trying to set sys.modules['__builtin__'].__dict__, an exception
	TypeError: read-only special attribute
is raised.

Is there anyway to go around that? May be in C level?

Any hints, suggestions, or similar are appreciated.

Thanks,
	Pearu



From arturo_perez at wanadoo.es  Mon Dec  4 15:48:16 2000
From: arturo_perez at wanadoo.es (Arturo Pérez Mulas)
Date: Mon, 04 Dec 2000 20:48:16 GMT
Subject: IIS ans Username
References: <8shi0n$fq2$1@nnrp1.deja.com> <8v1kqh$3mg$1@nnrp1.deja.com> <3A2A480C.9A228A9F@nwaf.com>
Message-ID: <kqTW5.728$%L6.13406@m2newsread.uni2.es>

> I have the same question.  What package does "Request" live in?

"Request" is no python object. It is an ASP environment object, and is
available through
the scripting engine. You use it just as directed in a previous mail (sorry
I do not keep the thread,
so I cannot provide the credits, it might be Bill Seitz).

> theId = Request.ServerVariables("LOGON_USER")

 If fact, "Request", being a class needs not to be instantiated
(you do not need to create an object of class Request), since all its
methods/properties belong to
the class.

Arturo P?rez.




From smnordby at yahoo.com  Tue Dec 12 09:30:36 2000
From: smnordby at yahoo.com (Steve)
Date: Tue, 12 Dec 2000 06:30:36 -0800
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com>
Message-ID: <3A36368C.26AADDDE@yahoo.com>

I dislike infinite while loops and use:

for line in file.readlines()
    ...do stuff...

No need for a break since it will automagically when it runs out of
lines.

-Steve-

Jaroslav Gergic wrote:
> 
> Hello,
> 
> one thing I dislike in many Python tutorials is
> the style of infinite 'while' loop with 'break' statements:
> 
> > Python:
> >
> >     file = open('file.input')
> >
> >     count = 0
> >     while 1:
> >         lines = file.readlines(8192)
> >         if not lines: break
> >         count = count + len(lines)
> >
> >     print count
> >
> 
> I hate it so much I prefer to write something like this:
> 
> line = fh.readline()
> while(line != ""):
>   ... do something ...
>   line = fh.readline()
> 
> OK, it is ugly to write the same line of code twice,
> but 'while true' or 'while 1' is very non-programmer construction
> it smells like VisualBasic GOTO.
> 
> Do you know better construction to avoid both - duplicate lines
> and horrible 'while 1' construction?
> 
> Something like:
> 
> while( "" != (line = fh.readline()) ):
>   ... do something ...
> 
> Thanks
> Gergi


From tim.hochberg at ieee.org  Fri Dec  8 10:34:00 2000
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Fri, 08 Dec 2000 15:34:00 GMT
Subject: Q for regexp experts
References: <yv2aea77xw2.fsf@lionsp093.lion-ag.de>
Message-ID: <Ib7Y5.58060$15.12302873@news1.rdc1.az.home.com>

I certainly won't claim to be a regexp expert, but:

".*?(?=SEP)"

seems to work. If alternatively you want to match to SEP or end of line:

".*?(?=SEP|\Z)"

seems to do the trick.

-tim

"Harald Kirsch" <kirschh at lionbioscience.com> wrote in message
news:yv2aea77xw2.fsf at lionsp093.lion-ag.de...
>
> Even with perl-compatible regular expressions it seems to be
> impossible to specify a pattern which matches
>
>    `the longest substring not containing separator' SEP
>
> Put another way, from a given starting point, the pattern shall match
> everything until just before the next occurence of SEP. A first
> attempt is something like
>
>    [^SEP]*
>
> but this disallows S', 'E' and 'P' in between the separators.
>
> I tried something like (.(?!SEP))+ but it somehow does not work. Is it
> at all possible to get this to work with regular expressions?
>
>   Harald Kirsch
>
> --
> ----------------+------------------------------------------------------
> Harald Kirsch   | kirschh at lionbioscience.com | "How old is the epsilon?"
> LION Bioscience | +49 6221 4038 172          |        -- Paul Erd?s
>




From skip at mojam.com  Fri Dec 22 11:04:11 2000
From: skip at mojam.com (Skip Montanaro)
Date: Fri, 22 Dec 2000 10:04:11 -0600 (CST)
Subject: ANNOUNCE: soundex.py
In-Reply-To: <3A4119C3.2E5E54E1@uswest.net>
References: <mailman.977338109.32467.python-list@python.org>
	<3A4119C3.2E5E54E1@uswest.net>
Message-ID: <14915.31611.208349.47250@beluga.mojam.com>

    Bjorn> Could I get you to put an explicit licence on it (it would be
    Bjorn> much more usable to me if it had a licence and if the licence
    Bjorn> wasn't GPL <wink>).

Released into the public domain.  Source at

    http://www.musi-cal.com/~skip/python/soundex.py

has been updated...

Skip



From dsh8290 at rit.edu  Tue Dec 19 14:04:06 2000
From: dsh8290 at rit.edu (D-Man)
Date: Tue, 19 Dec 2000 14:04:06 -0500
Subject: [ANN] xreadlines patch on sourceforge
In-Reply-To: <slrn93tm82.5rk.jepler@potty.housenet>; from jepler@inetnebr.com on Tue, Dec 19, 2000 at 03:41:55AM +0000
References: <slrn93tm82.5rk.jepler@potty.housenet>
Message-ID: <20001219140406.A15685@westsidecnc.rh.rit.edu>

Cool!  I was thinking about such a thing with the "while 1:" thread.

-D

On Tue, Dec 19, 2000 at 03:41:55AM +0000, Jeff Epler wrote:
> I've just uploaded my "xreadlines" patch to Sourceforge.  You can
> retrieve it from the Patch Manager, and it is my hope that it will
> be considered for inclusion in Python 2.1.
> 
> xreadlines, inspired by xrange, permits a simple 'for' loop to iterate
> over the contents of a file without reading the entire file at once.
> 
> xreadlines is both a new extension module, as well as an additional
> method on file objects.  It includes an entry for the test suite,
> as well as a modified 'fileinput' module which has a >2x speed
> improvement by using the xreadlines function.
> 
> example:
> 
> def wc(filename):
> 	l=w=c=0
> 	for line in open(filename).xreadlines():
> 		c=c+len(line)
> 		w=w+len(line.split())
> 		l=l+1
> 	return (l,w,c)
> 
> 
> Jeff Epler
> jepler at inetnebr.com
> -- 
> http://www.python.org/mailman/listinfo/python-list



From amb at nobjects.com  Thu Dec  7 11:13:56 2000
From: amb at nobjects.com (Are Magnus Bruaset)
Date: Thu, 07 Dec 2000 17:13:56 +0100
Subject: Problem with freeze.py in ActivePython 2.0
Message-ID: <3A2FB744.A942AD46@nobjects.com>

Hi,

I can't get freeze.py to behave under ActivePython 2.0 (Windows NT 4.0).
The command 

   python freeze.py -p E:\Python20 hello.py

issued in E:\Python20\Tools\freeze (E:\Python20 is the root my
installation) leads to the error message

Error: needed directory E:\Python20\lib\python2.0\config not found

Inside E:\Python20\lib there is no python2.0 directory. I can't find any
config directroy either.

Any hints on how to proceed? 


Best regards

Are Magnus





-- 
----------------------------------------------------------------------
   Are Magnus Bruaset, PhD      Email : amb at nobjects.com
   VP Development               Phone : +47 22 06 76 10
   Numerical Objects AS         Mobile: +47 92 61 00 63
   P.O. Box 124 Blindern        Fax   : +47 22 06 73 50
   N-0314 Oslo, Norway          WWW   : http://www.nobjects.com
**********************************************************************
 Diffpack 3.0 Released: The Next Step in Object-Oriented PDE Software
                    See http://www.nobjects.com
**********************************************************************


From fredrik at effbot.org  Tue Dec 19 11:30:53 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 19 Dec 2000 16:30:53 GMT
Subject: Python in "math through programming" curriculum
References: <2g2t3t06r2agrrjmfnr41eu4lc2u39a8qi@4ax.com> <3A3EFF47.67DE69CF@gte.net> <513v3toslkjkggh9oumlf20qc9bi1gdtif@4ax.com>
Message-ID: <13M%5.465$AH6.37131@newsc.telia.net>

Kirby Urner wrote:
> Maybe the Mac people are working on alternatives/enchancements
> to IDLE?  Does anyone have some info on this?

    http://starship.python.net/crew/just/

    "Yes, Guido is my brother"

    "I am the author of the IDE for MacPython. This place
    used to be the download spot for it, but since Python
    1.5.1 it is included in the standard MacPython distribution.
    If you can't find it: it lives in :Mac:Contrib:IDE. Starting
    with MacPython 1.5.2b1 /.../ the IDE is included in the
    distribution with full source code and a Python Style
    License (PSL)."

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From erno-news at erno.iki.fi  Tue Dec 12 18:11:34 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 13 Dec 2000 01:11:34 +0200
Subject: Python-like compiled language
References: <913jr3$dmv$1@sylvester.vcn.bc.ca> <3A35835B.AAA92005@bioeng.ucsd.edu> <kuzoi178kg.fsf@lasipalatsi.fi> <9163mr02v2t@news1.newsguy.com>
Message-ID: <kulmtl6xqx.fsf@lasipalatsi.fi>

In article <9163mr02v2t at news1.newsguy.com>, "Alex Martelli"
<aleaxit at yahoo.com> writes:

| "Erno Kuusela" <erno-news at erno.iki.fi> wrote in message
| news:kuzoi178kg.fsf at lasipalatsi.fi...
|| In article <3A35835B.AAA92005 at bioeng.ucsd.edu>, Curtis Jensen
|| <cjensen at bioeng.ucsd.edu> writes:
|| 
|| | The closest compiled languege to Python is probaby C/C++.
|| 
|| huh? [...]

in retrospect, i can see how C/C++ could be considered close
to python in the sense that they're easyish to interface
to python, and they're (compatible with, in the case of c++)
python's implementation language.

the above quoted sentence just triggered a reflex :)

[java points snipped]

| But in the context in which the question was asked, Java would
| not be much use, as the context was one of speed.  Python plus
| Numeric was judged too slow by the original poster; he wanted
| a faster solution.  I'd bet big bucks that Python+Numerics is
| faster than current Java implementations on all typical numeric
| issues, so Java wouldn't help.

some say that the latest virtual machines should be quite fast, but
then again with java you never know how much truth there is to the
hype...

[good scheme suggestion snipped]

it is also rumored that ocaml has a good optimizing compiler.

| Similarity is in the eye of the beholder... as is "compiledness"!-)

very true!

  -- erno


From kirschh at lionbioscience.com  Fri Dec  8 10:09:33 2000
From: kirschh at lionbioscience.com (Harald Kirsch)
Date: 08 Dec 2000 16:09:33 +0100
Subject: Q for regexp experts
Message-ID: <yv2aea77xw2.fsf@lionsp093.lion-ag.de>

Even with perl-compatible regular expressions it seems to be
impossible to specify a pattern which matches 

   `the longest substring not containing separator' SEP

Put another way, from a given starting point, the pattern shall match
everything until just before the next occurence of SEP. A first
attempt is something like

   [^SEP]*

but this disallows S', 'E' and 'P' in between the separators.

I tried something like (.(?!SEP))+ but it somehow does not work. Is it
at all possible to get this to work with regular expressions?

  Harald Kirsch

-- 
----------------+------------------------------------------------------
Harald Kirsch   | kirschh at lionbioscience.com | "How old is the epsilon?"
LION Bioscience | +49 6221 4038 172          |        -- Paul Erd?s



From aleaxit at yahoo.com  Fri Dec  8 12:37:03 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 8 Dec 2000 18:37:03 +0100
Subject: The secrets of operator.add
References: <g08zpq51vy.fsf@scumbag.ecs.soton.ac.uk>
Message-ID: <90r69r02hmn@news2.newsguy.com>

"Jacek Generowicz" <jmg at ecs.soton.ac.uk> wrote in message
news:g08zpq51vy.fsf at scumbag.ecs.soton.ac.uk...
    [snip]
> It looks as if multiple calls to operator.add are avoided somehow
> . . . or am I failing to understand the profiler's output ?

I think the profiler may be inserting its tracefunction in some
places and not others, thus perhaps distorting measurement.


> Is there a way of getting similar advantages in
>
> run ('map(lambda x:x+1, range(N))')  ?
>
> Is there a Python way of currying add?

There are several, e.g.:

def curry_add(n):
    def curried_add(x,n=n):
        return x+n
    return curried_add

but I doubt the actual performance will be noticeably different.


Alex





From deragon at 8d.com  Mon Dec 18 14:00:54 2000
From: deragon at 8d.com (deragon at 8d.com)
Date: Mon, 18 Dec 2000 19:00:54 GMT
Subject: float to string conversion
Message-ID: <91lmt3$4b8$1@nnrp1.deja.com>

Greetings.


  I am embarrassed to ask this since it is trivial, but
  I cannot figure out how to convert a float to a string (obviously,
  I am new to python).  Somehow, I could not find in the 2.0
  documentation the syntax to perform the operation.

  The following in my function:

    return MSISDN + " check FAILED:  $" + balance + errormsg

  where balance is a floating number and errormsg is a string
  generates the following error:

    TypeError: cannot add type "float" to string


Sincerely,
Hans Deragon


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


From mikael at isy.liu.se  Wed Dec  6 06:56:45 2000
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Wed, 06 Dec 2000 12:56:45 +0100 (MET)
Subject: Python Origins
In-Reply-To: <90l3j9$4bn$1@kapsel.intranet>
Message-ID: <XFMail.001206125645.mikael@isy.liu.se>

On 06-Dec-00 Piotr Kuchta wrote:
 >  I didn't noticed people speaking Esperanto in Amsterdam ;), but that's true
 >  that Dutchmen are very talented: common Dutchman speaks well two foreign
 >  languages, English and German.

That isn't too strange, since Dutch is the union of English, German, 
and to some extent Swedish... At least that's how it sounds.

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    06-Dec-00
Time:    12:55:26

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

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


From aleaxit at yahoo.com  Fri Dec 15 08:30:42 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 15 Dec 2000 14:30:42 +0100
Subject: compiled or interpreted?
References: <w8jZ5.74702$IP1.2487256@news1.giganews.com> <usnnt3oj4.fsf@cs.uu.nl> <G5Lz4o.EyA@leeds.ac.uk>
Message-ID: <91d6f50n5h@news1.newsguy.com>

"Nick Efford" <nde at comp.leeds.ac.uk> wrote in message
news:G5Lz4o.EyA at leeds.ac.uk...
> On 12 Dec 2000 11:46:39 +0100, piet at cs.uu.nl <piet at cs.uu.nl> wrote:
>
> > And Python doesn't have a JIT, like Java.
>
> Has the idea of developing a JIT for Python ever been considered?

Dunno, but developing Python for a JIT surely has -- witness
Jython and the Python .NET beta...!-)

Not that any performance boost is very observable yet, but...:-)


Alex





From bvdpoel at uniserve.com  Wed Dec 20 11:51:47 2000
From: bvdpoel at uniserve.com (Bob van der Poel)
Date: Wed, 20 Dec 2000 09:51:47 -0700
Subject: length of wav files
References: <3A402D8E.81978F3B@uniserve.com> <3A4057E4.9FAB4D1@zip.com.au>
Message-ID: <3A40E3A3.28BD2D9@uniserve.com>

Erik de Castro Lopo wrote:
> 
> Bob van der Poel wrote:
> >
> >
> > I am looking at the docs for the wave library and can't seem to find the
> > one bit of information I want to extract: the running time of a wav
> > file. So, before I have to relearn my math, has someone some code which
> > will deterime the time in seconds of a wav sound file?
> 
> import wave
> 
> wfile = wave.open ("somefile.wav", "r")
> time = (1.0 * wfile.getnframes ()) / wfile.getframerate ()
> 
> The `1.0' above to to force the calculation to use floating point
> arithmetic.

Many thanks. Missed the framerate() stuff.

-- 
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bvdpoel at uniserve.com
WWW:   http://users.uniserve.com/~bvdpoel



From see at below  Tue Dec  5 08:28:17 2000
From: see at below (Paul Foley)
Date: 06 Dec 2000 02:28:17 +1300
Subject: "inherited" keyword in Python?
References: <3A2B9EFA.3A64C13B@schlund.de> <90gevs02uu1@news2.newsguy.com> <m2elznipbl.fsf@mycroft.actrix.gen.nz> <90ievt025bl@news2.newsguy.com>
Message-ID: <m266kzhua6.fsf@mycroft.actrix.gen.nz>

On Tue, 5 Dec 2000 11:11:13 +0100, Alex Martelli wrote:

> "Paul Foley" <see at below> wrote in message
> news:m2elznipbl.fsf at mycroft.actrix.gen.nz...
>     [snip]
>> > Metaprogramming magick was deemed inappropriate here (as I
>> > read the group consensus).  One *might* have a global function
>> > called super(), which tried to gauge the appropriate class
>> > object to use (in single-inheritance cases, only, of course),
>> 
>> Why "of course"?  If you can find out the name of the method which is
>> calling super() (which can be done, but I don't remember how), and the

> Name of calling-function is easy (raise & catch an exception,
> it's in the stackframe you get this way).

That much I knew.  How to get the stack frame was what I didn't know.

> I think your cpl function is much more complicated than what's needed
> to reproduce Python's very simple rule for walking the DAG of base
> classes (it's just depth-first, left-first), even assuming it does
> end up reproducing it perfectly.  But that's not the main point.

It's probably much more complicated than it needs to be to do what it
does, but it's not intended to reproduce Python's rule exactly -- that
rule doesn't allow for a correctly-functioning "super" function in the
presence of multiple inheritance.

My cpl() uses the CLOS rule, which is similar to Python but not
broken.  The entire class hierarchy is sorted in such a way that
classes appearing to the left of other classes in __bases__ appear
before them in the sorted list, every class appears before any of its
bases, and each appears exactly once.  Every method is reachable, not
just the ones up the left branch, as in Python.

> A more interesting theoretical question is how you propose to solve
> the case in which you can't tell which class the current method
> 'belongs' to (also remember a given function-object may be set as
> a method of N classes, for any N, with any set of names -0- but,
> even forgetting that for the moment).  E.g., a very simple case:

If all else fails, take the easy way out and specify it in the call to
super()

> class A:
>     def foo(self):
>         print super()

At first reading, I didn't understand this: I thought it should be an
error, since there are no superclasses.  Reading further, I understand
that you expect super() to return one of the base classes.  Of course,
that won't work.  I thought super() was just supposed to call the
appropriate same-named method of a parent class.  Anything else can be 
done through the usual self.whatever, surely?

> class B(A):
>     def foo(self):
>         print super()
>         A.foo(self)

> class C(A):
>     def foo(self):
>         print super()
>         A.foo(self)

> class D(B,C):
>     def foo(self):
>         print super()
>         B.foo(self)
>         C.foo(self)

In this case, the CPL would be [D B C A], super() in D.foo() would
call B.foo(), whose call to super() would call C.foo(), whose call to
super() would call A.foo(), whose call to super() would signal an error.

> 'super()' in a multiple-inheritance context is a good example.
> There is no single superclass that will do for all attributes
> one might wish to access, so
>     super().foo()

[It was at this point that I realized what your super() was supposed
to be doing...]

> and 'supercall' is always (but always!-) called in the form

>     supercall(self, __thisclass, methodname, *args, **kwds)

> which reduces it to something decently-simple, like:

> def supercall(_obj, _cls, _metnam, *args, **kwds):
>     for _bas in _cls.__bases__:
>         _metobj = getattr(_bas, _metnam, None)
>         if not _metobj is None:
>             return _metobj(_obj, *args, **kwds)
>     else:
>         raise NameError, "No superclass of %s has method %s" % (
>             _cls, _metnam)

> But, how often is it going to be satisfactory to call 'just
> the one' (of several) inherited-method of a certain name, in
> a multiple-inheritance context?  Just as often we may want

It isn't (at least, not to me), which is why I avoided doing it that
way.  Having figured out how to get the frame info, here's an
implementation.  No need to pass in the object and method name:

  import sys

  def super(klass, *args, **kwargs):
      try:
	  1/0
      except:
	  frame = sys.exc_info()[2].tb_frame.f_back
      code = frame.f_code
      name = code.co_name
      self = frame.f_locals[code.co_varnames[0]]
      if args is ():
	  args = map(lambda x, f=frame: f.f_locals[x],
		     code.co_varnames[0:code.co_argcount])
      else:
	  args = [self] + list(args)
      # should do something similar with kwargs here
      classes = cpl(self.__class__)
      for klass in classes[classes.index(klass)+1:]:
	  if klass.__dict__.has_key(name):
	      return apply(klass.__dict__[name], args, kwargs)
      raise AttributeError, "No next method `%s'" % name

If you just call super(__thisclass) with no args other than the class,
it passes on the args the calling method got; or you can specify other
args explicitly [yes, I know this prevents you calling with no args if
super's caller has args]

-- 
The power of accurate observation is commonly called cynicism by those
who have not got it.
                                                    -- George Bernard Shaw
(setq reply-to
  (concatenate 'string "Paul Foley " "<mycroft" '(#\@) "actrix.gen.nz>"))


From news at dorb.com  Sun Dec 10 13:42:26 2000
From: news at dorb.com (Darrell)
Date: Sun, 10 Dec 2000 18:42:26 GMT
Subject: reloading modules question
References: <90scdf$hpe$1@troll.powertech.no> <2APY5.12112$hp.1409369@typhoon.nyroc.rr.com>
Message-ID: <m8QY5.12443$hp.1431105@typhoon.nyroc.rr.com>

Duh!

Could have used:
    sys.exit(1)
or
    raise SystemExit()

Not sure how these will work in this case though.

--Darrell



> I haven't made extensive use of reload, and want to hear how it goes for
> you.
> Although killing Python would clean up memory leaks and the like.
> I assume IIS would restart it, maybe??
>
> # Hacked from the win32 demos
> import win32api, win32pdhutil, win32con#, sys
>
> def killProcSelf():
>     myPid= win32api.GetCurrentProcessId()
> .....




From dsh8290 at rit.edu  Thu Dec 21 23:47:32 2000
From: dsh8290 at rit.edu (D-Man)
Date: Thu, 21 Dec 2000 23:47:32 -0500
Subject: ?Why is 0e-6 invalid?
In-Reply-To: <3A428C7C.5CFBB181@email.sps.mot.com>; from Norman_Shelley-RRDN60@email.sps.mot.com on Thu, Dec 21, 2000 at 04:04:28PM -0700
References: <3A428C7C.5CFBB181@email.sps.mot.com>
Message-ID: <20001221234732.A20586@westsidecnc.rh.rit.edu>

I think that it is illegal in Scientific notation (I presume that is
what you want).

AFIAK scientific notation is [1-9]e-?[0-9]+
(a 1 thru 9 followed by e followed by optional negative sign followed
by any sequence of 0 thru 9)

-D

On Thu, Dec 21, 2000 at 04:04:28PM -0700, Norman Shelley wrote:
> Why is 0e-6 invalid and 0*.000001 is valid?
> 
> They are equivalent and a "normal" mathematically trained individual
> sees them for what they are immediately.
> 
> (I have a trivial engineering notation to number convertor that takes
> "0u" or "5u" and converts them to "0e-6" and "5e-6" and then evaluates
> them. The later works fine, the former fails. Granted I could check for
> this condition, but it seems that the Python evaluator should be able to
> work on this boundary case as it still consistent.)
> 
> 
> >>> 0*.000001
> 0.0
> >>> 1e-6
> 1e-06
> >>> 0e-6
>   File "<stdin>", line 1
>     0e-6
>      ^
> SyntaxError: invalid syntax
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list



From daniel.dittmar at sap.com  Fri Dec  8 07:40:48 2000
From: daniel.dittmar at sap.com (Daniel Dittmar)
Date: Fri, 8 Dec 2000 13:40:48 +0100
Subject: FYI: Python support in SlickEdit
Message-ID: <90qksh$8ld$1@news1.wdf.sap-ag.de>

For those few who occasional ask about the Python support in SlickEdit:

In the new version 6.0, tagging is now properly implemented for Python.
There is no autocompletition for Python (at least none that I found). Apart
from that, most changes seem to be more cosmetic than revolutionary.

Daniel

--
Daniel Dittmar
daniel.dittmar at sap.com
SAP DB, SAP Labs Berlin




From dalke at acm.org  Fri Dec  8 22:47:34 2000
From: dalke at acm.org (Andrew Dalke)
Date: Fri, 8 Dec 2000 20:47:34 -0700
Subject: Python wrapper for C++ STL?
References: <3A31760E.D157EB54@erols.com> <npfY5.1850$x6.2577114@news2.rdc2.tx.home.com> <90s2lm$ig1$1@nnrp1.deja.com> <bDhY5.1999$x6.2840573@news2.rdc2.tx.home.com>
Message-ID: <90s9s1$5oj$1@slb6.atl.mindspring.net>

Rainer Deyke wrote
>I tend to use a dictionary where all values are None for sets.

For a bit more performance, use a number, like 0 or 1.
Using None requires a name lookup.

>>> def check_none(r):
...     for c in r:
...             None == None
...
>>> def check_one(r):
...     for c in r:
...             1 == 1
...
>>> import time
>>> r = range(1000000)
>>> t1=time.time();check_none(r);t2=time.time()
>>> t2-t1
6.01800501347
>>> t1=time.time();check_one(r);t2=time.time()
>>> t2-t1
2.72878599167
>>>

                    Andrew





From billtj at y.glue.umd.edu  Fri Dec 22 13:42:11 2000
From: billtj at y.glue.umd.edu (William Djaja Tjokroaminata)
Date: 22 Dec 2000 18:42:11 GMT
Subject: A Suggestion for Python Colon Syntax
References: <mailman.977446446.16022.python-list@python.org>
Message-ID: <9207a3$av6$1@hecate.umd.edu>

Tim Peters <tim.one at home.com> wrote:
: [William Djaja Tjokroaminata]
:> I saw many replies that give the reasons why the colons are
:> *needed*.

: Not mine:  it told you why the colons are *there*.  It's a human-factors
: thing, not a tech thing.  Desirable, not necessary.

Exactly.  I just regretted the initial study that used newbies and they
preferred colons.  At the present time, correct me if I am wrong, I think
it is still rare that Python is one's first programming language.  People
who start to use Python may already use some other popular programming
languages.  Probably if we do another survey now, the result may be
different.  I don't think survey is a bad thing.  Richard Stallman
sometimes asked opinions on the emacs newsgroup before he decided whether
to go one way or the other.

:> My suggestion is to make it *optional*, like the semicolon at
:> the end of statements.

: Understood.  Won't happen, though:  Python has been around for a decade.  Do
: you imagine that Guido hasn't heard this suggestion before?  In fact there
: are very few "new syntax" ideas that weren't debated and rejected before
: 1991 ended.  And, yes, this is one of them (ditto the whole "while 1" thread
: that's been going on virtually non-stop for a decade).

Probably it was true a long time ago, a decade perhaps.  Nowadays, there
are more programming languages around.  In particular, I like the idea of
the basic syntax of Tcl.  Yes, it is very simple, and too simple in a lot
of cases (thank goodness that Python syntax is not as simple as Tcl), but
it makes it just requires the bare minimum for syntax.  It seems to me
that the colon is redundant, because the next-level indentation is
required afterwards.  Isn't it the basic premise of Python to take the
best features of other programming languages?  Probably a decade ago ABC
was around, but now I think many more people have encountered Tcl rather
than ABC.

:> Yes, for people who cares about making it easier to read for newbies,
:> they can always put colons.  Yes, for people who use emacs to
:> automatically indent the next line, they can put colons too.

: Guido abhors gratuitous variations in coding sytle; that's largely why
: indentation is used instead of curly braces or begin/end to begin with.  A
: choice about whether or not to use a colon is simply not something people
: *need*.  The few places Python allows gratuitous choices today are places
: Guido regrets (although not enough to lose sleep over <wink>).  For example,
: Guido is unhappy that both "<>" and "!=" are accepted to mean "not equal";
: the older "<>" survives only for compatibility with a very early release of
: the language.

I can see clearly the big difference (as I have experienced) between using
indentation and curly braces; it makes really consistent coding
style.  However, I don't see the colons as falling into the same
category.  Yes, Guido wants everybody to use colons at the end of the
clauses, but right now what prevents someone to put semicolon at the end
of every statement?  Just compare the two or three formats:

    xxxx xxxx xxxx:
        xxxx xxxx xxxx
        xxxx xxxx xxxx

    xxxx xxxx xxxx:
        xxxx xxxx xxxx;
        xxxx xxxx xxxx;

    xxxx xxxx xxxx
        xxxx xxxx xxxx
        xxxx xxxx xxxx

Which one do you think is the most consistent in layout?  To me, it is not
the first one.

:> ...
:> I want opinion of you who have programmed for a long time whether
:> there will be some harms if the colons are made optional.

: It's Pythonic to either always require them or never allow them; the former
: won 10 years ago.

But the semicolon at the end of the statement break the Pythonic rule.

:> If not, probably someone can tell me how I can forward this
:> suggestion to the core Python development team?

: You already did <wink> -- Guido's on vacation this week, which makes me as
: close as you'll get until after Christmas.  You can submit a formal feature
: request via the SourceForge bug manager at:

:     http://sourceforge.net/bugs/?group_id=5470

: but you should expect it to be rejected.  Besides that Guido will object on
: principle, the practical support costs of adding such a thing are enormous:
: editing environments from the Emacs Python mode to IDLE would be broken, and
: would be difficult to fix.  For a hint as to why it's difficult, study
: tokenize.py in the std library.  That trailing colon makes partial parsing a
: whole lot easier, and especially when parsing backwards (as e.g. the Emacs
: python-mode and IDLE and PythonWin need to do routinely).

I am not too familiar with the parsing stuff.  However, in my simplistic
opinion, Python is not a free-form language like C or Perl.  Therefore,
probably it is reasonable for any parser to breaks a Python code first
into lines, even in backward parsing, instead of parsing it token by
token first.  In parsing backwards, can then it just detect first that the
line is at different indentation level rather than try to detect the
colon?

Regards,

Bill


From jhauser at ifm.uni-kiel.de  Wed Dec 13 07:27:38 2000
From: jhauser at ifm.uni-kiel.de (Janko Hauser)
Date: 13 Dec 2000 13:27:38 +0100
Subject: Tutorial in Numerical Python?
References: <87k8947e9f.fsf@sheffield.ac.uk>
Message-ID: <87hf48ld51.fsf@lisboa.ifm.uni-kiel.de>

There is a tutorial at OReilly which tries to do exactly this.

http://www.oreillynet.com/topics/python/scientific

-- 
  Institut fuer Meereskunde             phone: 49-431-597 3989
  Dept. Theoretical Oceanography        fax  : 49-431-565876
  Duesternbrooker Weg 20                email: jhauser at ifm.uni-kiel.de
  24105 Kiel, Germany


From aleaxit at yahoo.com  Tue Dec 19 12:24:44 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 19 Dec 2000 18:24:44 +0100
Subject: while true: !!!
References: <mailman.977243949.24318.python-list@python.org>
Message-ID: <91o5lo01k93@news2.newsguy.com>

"Carsten Geckeler" <nospam at no.spam> wrote in message
news:mailman.977243949.24318.python-list at python.org...
    [snip]
> > My claim is that Python is simpler, clearer, and MORE effective,
> > for NOT trying to draw this level of distinction.
>
> Right, it's more effective with regard to programmers' work, but
> definitely less effective according to speed.

Not because of the specific lack of distinction between
character and string-of-length-1, no.  Remember Java DOES
draw the distinction but speed of typical Java and Python
programs are comparable (and have been for years, though
Java keeps *promising* future speedups).


> > > Do you know anything faster than this?  It just writes the byte char
into
> > > the memory space pointer str + 3.  In most cases strings do consist of
> > > chains of bytes in memory.
> >
> > Not sure what you mean by 'chain' here -- Haskell is the only one
>
> Is array better?

Definitely better -- it IS a frequent computer-science term.


> > I program in C when I must (though I'd much rather take a
> > step up to C++, which I can in most situations but *NOT* when
> > I want my Python extensions to be taken seriously by the
> > Python community, alas...).  That doesn't mean I have to
> > _like_ it:-).
>
> Right.  Take what programming language you like.  But the fact that Python
> and kernels are programmed in C shows, that you always use the language
> that fits your problem.  I love programming in Python, I also like Perl,
> and I also program in C.  And probably I will also program in Fortran
> during my thesis (although I HATE Fortran).  But that's it: If you need
> pure speed, Fortran is just a better choice than rare Python.  If you need
> high-level construct whithout -- almost -- the work of type declaration,
> Python is the better choice.

The best choice, for speed and for anything else, is to use "N"
languages for some N >= 0.  N==2 is a frequent case -- some low
level language for the speed-crucial bits, some high-level one
for the rest (often over 80% of the overall functionality).

N==1 is sometimes best, but hardly ever [a] for a big component
and [b] for a language of low level (kernels are not programmed
in ONE language -- they have some machine language as low-level
and C/C++/etc as the higher-level for most stuff).


> C++ with the STL is very nice. ;)

Yes, and with Boost on top (www.boost.org) even better -- Boost
builds on Standard C++ Library ideas (which you also find in STL)
to add many useful functions, such as Python extending in C++.


Alex





From jh at web.de  Mon Dec  4 17:57:51 2000
From: jh at web.de (Jürgen Hermann)
Date: Mon, 4 Dec 2000 23:57:51 +0100
Subject: "inherited" keyword in Python?
References: <3A2B9EFA.3A64C13B@schlund.de>
Message-ID: <90h7hl$hfj$02$1@news.t-online.com>

"Carsten Gaebler" <cg at schlund.de> schrieb im Newsbeitrag
news:3A2B9EFA.3A64C13B at schlund.de...
> Does Python know something like the "inherited" keyword from Borland
> Pascal?

>>> class a:
...  pass
...
>>> class b(a):
...  pass
...
>>> x=b()
>>> print x.__class__.__bases__
(<class __main__.a at 007F15CC>,)




From ben.hutchings at roundpoint.com  Fri Dec 22 07:41:59 2000
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 22 Dec 2000 12:41:59 +0000
Subject: IP adresses are not unique enough.
References: <3a413a17.3416402@192.168.0.1> <3A41475A.5DC4ABD8@san.rr.com> <3a4148ad.7150467@192.168.0.1>
Message-ID: <uzohoy6eg.fsf@roundpoint.com>

gogo at bluedynamics.com (Georg Gogo. BERNHARD) writes:
<snip>
> The MAC adresses are unique, as I understood.

As long as users don't go changing them.  They are not generally
hard-wired into Ethernet interfaces.

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From skip at mojam.com  Wed Dec 20 13:44:58 2000
From: skip at mojam.com (Skip Montanaro)
Date: Wed, 20 Dec 2000 12:44:58 -0600 (CST)
Subject: ANNOUNCE: soundex.py
Message-ID: <14912.65066.958523.18862@beluga.mojam.com>

Python 2.0 no longer ships with a soundex module.  Sometime ago, Tim Peters
and Fred Drake each cooked up replacements written in Python.  I merged them
together into a single module which is available from 

    http://musi-cal.mojam.com/~skip/python/

If you have any questions or comments on the module, please send them my
way.

-- 
Skip Montanaro (skip at mojam.com)
Support the Mojam.com Affiliates Program: http://www.mojam.com/affl/
(847)971-7098



From jflores at codeit.com  Mon Dec 18 14:18:43 2000
From: jflores at codeit.com (Julio Flores Schwarzbeck)
Date: Mon, 18 Dec 2000 11:18:43 -0800
Subject: =?iso-8859-1?Q?Re:_More_docs_in_Spanish/M=E1s_madera?=
In-Reply-To: <3A3E6153.CA888DF5@arrakis.es>
References: <5.0.2.1.0.20001218105723.00a232e0@mail.codeit.com>
Message-ID: <5.0.2.1.0.20001218111745.00a2a0d0@mail.codeit.com>

No te preocupes, hermano, para eso andamos.. ya me agarrar?s en una t?.. je.

Un saludo.

Julio

At 08:11 PM 12/18/00 +0100, Marcos S?nchez Provencio wrote:
>Julio Flores Schwarzbeck wrote:
> >
> > I believe you meant http://pySpanishDoc.sourceforge.net/
> >
> > Probablemente te refieres a: http://pySpanishDoc.sourceforge.net/
> >
> > Saludos,
> >
> > At 06:01 PM 12/18/00 +0000, rapto at arrakis.es wrote:
> > >New files in
> > >Ficheros nuevos en
> > >http://pySpanishDoc/sourceforge.net/
> > >
> > >
> > >Sent via Deja.com
> > >http://www.deja.com/
> > >--
> > >http://www.python.org/mailman/listinfo/python-list
>
>Indeed. Thank you very much.
>Corcho, vaya gamba. M?s vale vigilarme.




From gogo at bluedynamics.com  Wed Dec 20 18:07:08 2000
From: gogo at bluedynamics.com (Georg Gogo. BERNHARD)
Date: Wed, 20 Dec 2000 23:07:08 GMT
Subject: How to make a GUID with plain python
Message-ID: <3a413a17.3416402@192.168.0.1>

Hello!

I am looking for a way to create a global unique identifier / UUID
with python. I found a way to do this with ActivePython in Windows,
but I need a solution for other platforms, too. (In ActivePython one
can use >>> pythoncom.CreateGuid <<< to get a global unique
identifier.)

The GUIDs use the MAC adress of the machine they are created at in
their algorithm. I start to think that this can not be done in python,
but I still hope that I am wrong in that point.

A plain python algorithm itself would be very useful, since one could
use it for RPC / COM as well as for creating indexes for distributed
databases.

Maybe somebody knows how to create a GUID with plain python, any help
is welcome.

Regards, 
gogo at bluedynamics.com






From michael at stroeder.com  Thu Dec 28 06:06:10 2000
From: michael at stroeder.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Thu, 28 Dec 2000 12:06:10 +0100
Subject: PEP 222 draft
References: <slrn947bj1.1rs.amk@207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com> <925b3g$5pl$1@nnrp1.deja.com>
Message-ID: <3A4B1EA2.E15B4FB7@stroeder.com>

tlownds at my-deja.com wrote:
> 
> In article <slrn947bj1.1rs.amk at 207-172-57-45.s45.tnt2.ann.va.dialup.rcn.com>,
>   akuchlin at mems-exchange.org wrote:
> >     Utility function: build a query string from a list of 2-tuples
> >
> 
> Perhaps it would be better to extend urllib.urlencode to handle this 
> case. I think it should be extended to handle lists as values as well:
> 
> >>> urlencode({'x': ['a', 'b']})
> 'x=%5b%27a%27,+%27b%27%5d'
> 
> would be:
> 
> 'x=a&x=b'

But you can do that with a list of 2-tuples either. Note: The list
order of dictionary keys is normally not preserved.

Ciao, Michael.


From scarblac at pino.selwerd.nl  Wed Dec  6 08:55:14 2000
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 6 Dec 2000 13:55:14 GMT
Subject: Sort
References: <NlnX5.1444$yv1.2700363@nnrp5.proxad.net>
Message-ID: <slrn92sh4o.g70.scarblac@pino.selwerd.nl>

Alfred <scj at mcom.mcom.fr> wrote in comp.lang.python:
> Hello world,
> 
> I'm trying to sort a List of List, on the third col, but I don't find how to
> do this.

# 'L' is a list of lists with at least three elements.
# No error checking is done.

def cmp_lists(A, B):
   # Function that compares two lists by their third element
   return cmp(A[2], B[2])
   
sortedlist = L[:]          # Make a copy, may be unnecessary
sortedlist.sort(cmp_lists) # Sort it using the function cmp_lists


From erno-news at erno.iki.fi  Tue Dec 12 14:17:51 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 12 Dec 2000 21:17:51 +0200
Subject: Python-like compiled language
References: <913jr3$dmv$1@sylvester.vcn.bc.ca> <3A35835B.AAA92005@bioeng.ucsd.edu>
Message-ID: <kuzoi178kg.fsf@lasipalatsi.fi>

In article <3A35835B.AAA92005 at bioeng.ucsd.edu>, Curtis Jensen
<cjensen at bioeng.ucsd.edu> writes:

| The closest compiled languege to Python is probaby C/C++. 

huh? those are about a million times lower level than python,
with manual memory management and pointers and all...
c++ is also about a million times more complex and pretty ugly
with all the backwards-compatibility baggage.

  -- erno


From fredrik at effbot.org  Tue Dec 19 11:28:08 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 19 Dec 2000 16:28:08 GMT
Subject: Walk through dictionary keys?
References: <3A3F8B89.1DD9E01F@schlund.de>
Message-ID: <s0M%5.463$AH6.37024@newsc.telia.net>

Carsten Gaebler wrote:
> Is there a way of 'walking' through the keys of a dictionary?

keys()

> If I have a large dictionary I'd like to prevent
>
> for key in d.keys():
>     do_something
>
> because this makes a copy of all the keys.

wrong.  it creates references to all the keys, not
copies of them.

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From sasoft at gmx.de  Sat Dec 16 06:05:20 2000
From: sasoft at gmx.de (cruciatuz)
Date: Sat, 16 Dec 2000 12:05:20 +0100
Subject: new with python..
In-Reply-To: <91dt4e$d3u$1@barni-1.tehnicom.net>
References: <91dt4e$d3u$1@barni-1.tehnicom.net>
Message-ID: <1514064385.20001216120520@gmx.de>

Hello Milos,


MH> i am new with this proggraming tool and i would appriciate if you would tell
MH> me where can i get a good tutorial.
    Is the python tutorial bad?
    Plz be more explicit.
    What are you searching for?




+ ---------------------------------- +
| Best regards:   Cruciatuz_AH       |
+ ---------------------------------- +
| AKA:            Stefan Antoni      |
+ ---------------------------------- +
| ICQ:            72292815           |
| web (d2-clan):  www.cruciatuz.de   |
| PGP-Key:        AVAILABLE          |
+ ---------------------------------- +                   
| Nur der Kleingeist h?lt Ordnung,   |
| das Genie beherrscht das Chaos.    | 
+ ---------------------------------- +

12:03 / Samstag, 16. Dezember 2000





From obou at my-deja.com  Tue Dec  5 12:18:50 2000
From: obou at my-deja.com (obou at my-deja.com)
Date: Tue, 05 Dec 2000 17:18:50 GMT
Subject: SAP R/3 automation with win32com
References: <90flpu$gpu$1@nnrp1.deja.com> <kl5n2tsfk3hud1c37gjo56290sueccuegk@4ax.com> <90i6n6$hg0$1@nnrp1.deja.com> <90icc6022q9@news2.newsguy.com>
Message-ID: <90j81i$c4a$1@nnrp1.deja.com>

In article <90icc6022q9 at news2.newsguy.com>,
  "Alex Martelli" <aleaxit at yahoo.com> wrote:
> <obou at my-deja.com> wrote in message news:90i6n6$hg0
$1 at nnrp1.deja.com...
>     [snip]
> > I don't know VB at all and don't really know how to translate VB
> > functions calls to Python function calls. Especially when using COM.
> > There are some strange differences. In my example, there are in VB
> > function calls that takes 3 parameters where in Python I have to put
> > only 2 parameters in the brackets and get the third as a return
value.
> > The salesOrd.DimAs function calls are such cases. In VB the return
> > value is written as a third parameter and the function has no return
> > value.
> > Is there a way to identify which parameter have to stay in the
brackets
> > and which ones will be sent as return value?
>
> Yes.
>
> COM parameters are defined as [in], [out], [out,retval], or [in,out]
> (there are also other attributes, but these are the ones that matter
> here).  The [out,retval], if present at all, is only the last one.
>
> In VB, a parameter [out,retval] appears as a return value, all others
> appear as actual parameters to the function call (and not as returns).
>
> In Python, parameters [in] and [in,out] appear as actual parameters
> to the function call; parameters [out], [out,retval], and [in,out]
> appear as return values (once you've run makepy on the type library
> to let Python know what is what -- otherwise, [out] and [in,out]
> can be troublesome).  Note that [in,out] is BOTH an actual parameter
> (for the 'in' part) AND a return value (for the 'out' part).
>
> [out] and [in,out] are not all that common in well-designed COM
> (Automation) type-libraries, but, they do happen at times.
>
> The VB Object Browser does not clearly display the attributes of
> the parameters.  *However*, another precious tool is called "OLE
> View" and should be also part of VB's higher editions (and I
> think oleview.exe is also one of the tools you can download for
> free as part of Microsoft's "Platform SDK").  THAT one, OleView,
> lets you display any type-library in terms of IDL, the Interface
> Definition Language, which includes the [in], [out], etc,
> notations.  It's a very good tool for exploring COM stuff.  Of
> course, a component's docs SHOULD include the non-trifling
> details of what parameters are input ones, which are output
> ones, etc, etc, but, sometimes, such little crucial things
> drop by the wayside... and OleView can save your life then!-)
>
> Alternatively, if you'd rather read Python than IDL, look at
> the source-file that makepy.py generates for you when you 'hook
> up' to a type-library.  It's placed in win32com\genpy, and its
> filename is the GUID of the library; in the Python classes it
> builds to correspond to each COM interface, it also clearly
> separates and distinguishes between input and output parameters
> of each method.
>
> Alex
>
>

Thanks for these very helpful and precise informations!

As I have VB installed, I'll have a look at the Ole View tool.

I abuse, but is there a way to find the "Makepy" name of a particular
COM object? I have tons of SAP objects in Makepy and cannot find one
that correspond to the objects I'm using...
Do somebody know where Makepy looks for the titles it shows in it's
listbox?

Olivier Boudry


Sent via Deja.com http://www.deja.com/
Before you buy.


From ben.hutchings at roundpoint.com  Fri Dec 22 08:02:48 2000
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 22 Dec 2000 13:02:48 +0000
Subject: IP adresses are not unique enough.
References: <3a413a17.3416402@192.168.0.1> <3A41475A.5DC4ABD8@san.rr.com> <3a4148ad.7150467@192.168.0.1>
Message-ID: <uvgscy5fr.fsf@roundpoint.com>

gogo at bluedynamics.com (Georg Gogo. BERNHARD) writes:
<snip>
> The MAC adresses are unique, as I understood.
<snip>

I have some anecdotal evidence of cases where a manufacturer assigned
duplicate MAC addresses.  Someone found me a documented case, too:
<http://puck.nether.net/lists/cisco-nsp/2731.html>.

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From dsh8290 at rit.edu  Tue Dec 26 18:03:40 2000
From: dsh8290 at rit.edu (D-Man)
Date: Tue, 26 Dec 2000 18:03:40 -0500
Subject: Setting Environment Variables
In-Reply-To: <20001223051758.A25990@bigskytel.com>; from jcm@bigskytel.com on Sat, Dec 23, 2000 at 05:17:59AM -0700
References: <LOBBJMHOIFOLFIPFANHJMEFCCFAA.beno@TheWebsons.com> <"from beno"@TheWebsons.com> <20001223051758.A25990@bigskytel.com>
Message-ID: <20001226180340.B8309@westsidecnc.rh.rit.edu>

More accurately the file is 

$HOME/.bashrc
or
~/.bashrc

since the location of the home directory is up to the administrator.
But David is right since you have a standard installation.

:-)
-D

On Sat, Dec 23, 2000 at 05:17:59AM -0700, David Porter wrote:
> * Ben Ocean <beno at TheWebsons.com>:
> 
> > I'm trying to set environment variables so I can define my own path for my
> > modules. My problem is I don't know where to set the PYTHONPATH. My book
> > says do it in the .cshrc file: where the heck's that? My Linux book says it
> > should be in the /home directory, but it isn't. I own this server and it's a
> > standard RedHat 6.2 install. Please advise.
> 
> What is your shell? It is probably bash. if so, add something like this to
> /home/username/.bashrc:
> 
> export PYTHONPATH=/path/to/modules
> 
> If adding more than one path put a colon between them.
> 
> 
> David
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list



From oivvio at polite.se  Thu Dec 14 14:34:59 2000
From: oivvio at polite.se (oivvio at polite.se)
Date: Thu, 14 Dec 2000 19:34:59 GMT
Subject: Right tools for python info space mapping project?
Message-ID: <91b7d4$9ml$1@nnrp1.deja.com>

Hi all.

I've been thinking of writing an info space mapping program for some
time. (Check out http://www.geog.ucl.ac.uk/casa/martin/atlas/atlas.html
for some
examples of info space mapping.)

I want to retrieve material from the net (my first priority is newspaper
sites), store it in a database, calculate relationships between texts
and make  graphical representations of theese relationships.

I hope that this will mostly consist of gluing together existing
open sourced software. I'll be using python for this gluing work.

This is what I need:

1. RDBMS
MySQL, or PostgreSQL. Which one has better python bindings?

2. Full text indexing.
As neither of the free RDBMS have full text indexing this will have to
been done separately. Sourceforge hosts Wordindexer (a perl thingy that
works with MySQL), mifluz (a c++ word indexer), and pyMiflus (python
bindings for mifluz). What are you're experiences with these and
similar tools.

3. Parsing websites.
I need a way to describe the structure of a website (what's to be
downloaded and what's not.) I haven't found anything in this
category. Have you?

4. Calculating relationships between texts based and word frequencies.
I guess this will be a work for NumPy

5. Making the pictures.
There seems to be about a million graphics/plotting packages for
python. I'm looking for something that has a lot of capabilities for
making the pictures very pretty. ImageMagick, vtk, PIL or what?

All ideas and  pointers are welcome.

regards oivvio


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


From delza at antarcti.ca  Thu Dec  7 15:32:02 2000
From: delza at antarcti.ca (Dethe Elza)
Date: Thu, 07 Dec 2000 20:32:02 GMT
Subject: MVC module anyone?
References: <90nnq6$8n0$1@news.rz.uni-karlsruhe.de>
Message-ID: <6tSX5.767919$8u4.13380209@news1.rdc1.bc.home.com>

Bjoern,

I've toyed around with something like this.  Basically, putting the view in
external resources.  I would use XML for this.  Since most GUI toolkits are
now using XML for their serialized forms, at least optionally, you could
then write a XSLT tranform from your resource to the target GUI's specific
resource format.

I hope the above makes sense, I've been thinking about this a lot and
sometimes jump over several steps when I
try to describe it.

Wish I had more time to work it out.  For a base XML language to jump into
this from there are several choices:

XUL (http://www.mozilla.org/xpfe/xulref/XUL_Reference.html), used in Mozilla
UIML (http://www.uiml.org), proprietary language for doing just what I
describe above
XForms (http://www.w3.org/MarkUp/Forms/), upcoming standard to add real GUI
elements to the web

Some of the xml resource formats to consider transforming to:

libglade xml for GTK+ and GNOME (http://www.daa.com.au/~james/gnome/), hard
to find documentation on the XML format itself
BML (http://www.alphaworks.ibm.com/tech/bml), for Java/Swing
I don't know of an xml dialect for describing/creating Tkinter interfaces,
but there is one in the works for wxWindows.  No URL last time I checked.

HTH

--Dethe


"Bjoern Giesler" <Bjoern.Giesler at gmx.de> wrote in message
news:90nnq6$8n0$1 at news.rz.uni-karlsruhe.de...
> Hi,
>
> I'm thinking of writing a Model-View-Controller based GUI wrapper around a
> couple of the more popular Python GUI toolkits because
> a) I'm tired of apps that work with only a certain GUI, so I want
>    an abstraction layer
> b) I like Model-View-Controller.
>
> Is there already such a project? I don't want to duplicate efforts.
>
> If not, can anyone point me to the Smalltalk MVC specification? I've
worked
> with Smalltalk a while ago, but don't remember exactly how the classes
> looked like. I'd really like to do this the way Smalltalk does it. (If I
> can't, I'll use the OPENSTEP AppKit as a model.)
>
> Regards,
> --Bj?rn




From andreas at andreas-jung.com  Fri Dec 22 03:59:58 2000
From: andreas at andreas-jung.com (Andreas Jung)
Date: Fri, 22 Dec 2000 09:59:58 +0100
Subject: Speed of Python 1.5.x vs 2.0
In-Reply-To: <4vj54t4v0hvjsnstfmtmf36q9irotp59sj@4ax.com>; from tyler@tylereaves.com on Thu, Dec 21, 2000 at 10:53:32PM -0500
References: <4vj54t4v0hvjsnstfmtmf36q9irotp59sj@4ax.com>
Message-ID: <20001222095958.A22123@yetix.sz-sb.de>

On Thu, Dec 21, 2000 at 10:53:32PM -0500, Tyler Eaves wrote:
> 	I'm soon going to attemp to talk my webhost into upgrading to
> 2.0, but I figure the only way I'm going to be able to make that
> happen is if I can show results that show 2.0 being faster / more
> effcient than 1.5.x. Their servers run on Red Hat 7.

2.0 runs a bit faster than 1.5.X but don't expect a performance
boost. You should upgrade when you need some of the new features
of Python 2.0 like unicode support... 

Andreas 



From trentm at ActiveState.com  Wed Dec 20 15:44:41 2000
From: trentm at ActiveState.com (Trent Mick)
Date: Wed, 20 Dec 2000 12:44:41 -0800
Subject: Python in embedded systems?
In-Reply-To: <_w706.11237$TU6.1069571@ptah.visi.com>; from grante@visi.com on Wed, Dec 20, 2000 at 07:12:58PM +0000
References: <_w706.11237$TU6.1069571@ptah.visi.com>
Message-ID: <20001220124441.A21354@ActiveState.com>

On Wed, Dec 20, 2000 at 07:12:58PM +0000, Grant Edwards wrote:
> Has anybody heard of a Python interpreter being used in an
> embedded system to provide user scripting capabilities?

Don't know if it helps for your situation but I believe that the empeg car
MP3 player embeds Python for driving its UI:

http://www.empeg.com/


Trent

-- 
Trent Mick
TrentM at ActiveState.com



From donn at oz.net  Sat Dec 30 02:42:05 2000
From: donn at oz.net (Donn Cave)
Date: 30 Dec 2000 07:42:05 GMT
Subject: Queuing mechanism that is poll/select'able ?
References: <txk4rzo9o4j.fsf@vortex.ukshells.co.uk>
Message-ID: <92k3kd$2dt$0@216.39.151.169>

Quoth Chris Ebenezer <chriseb at pobox.com>:
| I'm re-writing some python code to use active objects (i.e multiple
| threads).  I have been looking at using the Queue class as a means to
| send messages between different objects.  
|
| However, the one problem is that since the Queue class isn't based
| around a unix abstraction that is descriptor based, it doesn't support
| select().  Since this is the only mechanism that is relatively
| ubiquitous in unix for the use of polling, this causes problems when
| you try to use Queue's in-conjunction with other 'eventing objects'
| like sockets etc.
|
| I'm pondering writing a Queue-like class based around UNIX domain
| sockets (pipes would be even better but would have built in
| limitations) and cPickle, but before i do this i thought i would ask
| around to see if such a beast already exists.
|
| So, does such a mechanism exist ?

I have had to do a similar kind of thing, though in the context
of BeOS native thread & message handling vs. sockets.  I don't
have an existing implementation that works with Queue, but I hope
you will find that existence of such parallel implementations
supports the concept.  In other words, good idea, go for it!

	Donn Cave, donn at oz.net


From news at dorb.com  Sun Dec 10 14:24:59 2000
From: news at dorb.com (Darrell)
Date: Sun, 10 Dec 2000 19:24:59 GMT
Subject: Problems With socket Under NT & PythonWin 2.0
References: <910jtt$nnh$1@nnrp1.deja.com>
Message-ID: <fMQY5.12943$hp.1457332@typhoon.nyroc.rr.com>

Copy the example to a .py file and run it from the command line or just
double click it.
The example blocks on conn.recv(1024), meaning that you current thread will
stop responding.

Of interest might be select or timeout socket.
Timeout socket is something you would need to download.


--Darrell

<swanky_mode at my-deja.com> wrote:
> Greetings;
>
> I'm trying to learn a bit more about Python programming under Windows
> NT, and came across a problem with the Socket EchoServer.py example in
> section 7.2.2 inside the Python reference pages.
>
> As soon as I run EchoServer inside the PythonWin IDE with Python 2.0,
> the entire IDE "hangs"; e.g., the mouse pointer icon is set to an hour
> glass, and there is no change in its settings, as well as not being
> able to click on any of the toolbar buttons.  The only way I can stop
> the application is by closing PythonWin from the NT Task Manager.
>





From aleaxit at yahoo.com  Sat Dec 16 03:35:59 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 16 Dec 2000 09:35:59 +0100
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <tcmay-FCA22F.14073713122000@news.got.net> <918v7a02scn@news1.newsguy.com> <tcmay-04E80C.16181313122000@news.got.net> <91a1ed02a4j@news2.newsguy.com> <slrn93l1tf.5l.cerutti@fiad06.norwich.edu>
Message-ID: <91f9jr0ado@news2.newsguy.com>

"Neil Cerutti" <cerutti at together.net> wrote in message
news:slrn93l1tf.5l.cerutti at fiad06.norwich.edu...
    [snip]
> Luckily, the objects we model in software are generally simpler
> and easier to define than a Vehicle class. In the real world,

I'd put it another way: the _models_ we build in our minds
(particularly for software-design use, but not only!) are
simpler than the 'real world' objects.  That's why they're
models: the map is not the territory.  It's the source of both
their usefulness and their limitation.

> things have "family relationships", (Wittgenstein?)--you can't

Wittgenstein is talking about _human language_ (which in his
terms covers human thoughts, and actions in as much as they
communicate, to another or to oneself): a word (a concept)
denotes a 'cloud' of entities (he might have called it a fuzzy set,
had the term existed) which are joined by a vague 'family
resemblance'.

There is no implied assertion on any 'deep nature of reality'
(no needed noumena under those phenomena, if you will),
because "whatever we're talking about, we're only talking
about natural history of human beings" -- i.e., all of our
words/thoughts/concepts/models/actions/abstractions are
ALWAYS in the context of our biology/needs/desires/
abilities (that's in the _mature_ Wittgenstein; the Tractatus
of the 20-something Wittgenstein foreshadowed that key
realization but didn't truly and fully grasp it).

> always make an easy determination about what class a thing
> belongs to. In an ocean simulation, you might have a Fish class
> and it would be useful for the Dolphin class to inherit, but it
> IS-NOT a fish.

At roughly the same time as Wittgestein's, Alfred Korzybski was
busy building general semantics, and the law of non-identity.

The map-territory relationship was his conception, and more
fully stated: "The map is not the territory it represents, but, if
correct, it has a similar structure to the territory, which accounts
for its usefulness...." and a few pages later "it is enough to train
children to abandon the ?is? of identity", to build "consciousness
of abstracting".

The focus on _usefulness for our purposes_ of our models,
which need to be consciously built as such, as abstractions,
parallels the 'natural history of human beings' idea.

If this be an ontology/epistemology split, then, make the most
of it... it seems to me that most thinkers after W, K and other
contemporary analyzers of language (Sapir and Whorf also
come to mind:-) have more or less accepted this split, i.e,
focused on 'how we know what we know and what do we do
with it' more than on 'deep nature of reality' issues.


> Perhaps it makes sense to instantiate a Vehicle if you believe in
> Plato's idea of an ideal vehicle. All other vehicles will be
> different or imperfect in some way, but the ideal could exist and
> be instatiated--by God, er... by the programmer. ;-)

If you accept Plato's conceptions, the Ideal cannot be instantiated
in this world; there is a different, 'true' world, where _only_
Ideals lie.  Plotinus (and other Platonists before the Renaissance)
may be closer to this 'Ideal Instantiated' possibility.  But _this_
is indeed ontological more than epistemological, and lacks the
key distinction of maps/models/concepts (in our minds, for our
purposes) vs 'noumena'.


Alex





From nospam at no.spam  Mon Dec 18 17:53:55 2000
From: nospam at no.spam (Carsten Geckeler)
Date: Mon, 18 Dec 2000 23:53:55 +0100 (MET)
Subject: while true: !!!
In-Reply-To: <slrn93snve.tvg.grey@teleute.rpglink.com>
Message-ID: <Pine.LNX.4.10.10012182334180.1345-100000@erde.kosmos.all>

On Mon, 18 Dec 2000, Steve Lamb wrote:

> On Mon, 18 Dec 2000 17:05:51 +0100, Alex Martelli <aleaxit at yahoo.com> wrote:
> >I think it would be inconsistent, un-Pythonic, to add a fundamental type for
> >'conceptual' reasons, without any practical, measurable benefit.  For
> >example, Python does away with the distinction between 'a character' and 'a
> >string of length 1' -- a distinction that is religiously maintained in C,

Not fair!  C is a lowlevel language (relatively to interpreter languages
as Python etc.).  In C you can say
	s[3] = char
Do you know anything faster than this?  It just writes the byte char into
the memory space pointer str + 3.  In most cases strings do consist of
chains of bytes in memory.

> >Pascal, Java, etc, and does have a basis in both 'conceptual' and
> >machine-level considerations -- yet the strong conceptual *simplification*
> >arising from just doing away with the distinction appears to be quite
> >positive.
> 
>     I can attest to that.  I have tough times in C, Pascal and other such
> languages because of one consideration they make that Perl and Python do not.
> Data is data.  Forgiving me my discussion earlier where I said Python doesn't
> do that (when compared to Perl) it still does it in all the ways that count.
> I don't have to worry about how large or small of a block of space to allocate
> to my strings or arrays.  It's a string, not a string[80] that I might one day
> want to stuff 81 characters into.  :/

Please, don't be so mean to C.  It's such a nice programming language!
Remember, Python itself is programmed in C.  ;)

But to your point:  If you are sure that your string is always shorter
than 79 characters, it'n no faster way than doing it as C.  No overhead
due to range checking and reallocating.  But of course, from the everyday
user's point, interpreter languages as Python or Perl do a nicer job.  But
every language has it's field of appliance.  Just imagine programming a OS
kernel in Python... ;)

Cheers, Carsten
-- 
Carsten Geckeler:  carsten dot geckeler at gmx dot de
To get proper email-address replace `dot' and `at' by the corresponding symbols.






From grey at despair.rpglink.com  Mon Dec 18 06:18:59 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Mon, 18 Dec 2000 11:18:59 -0000
Subject: while true: !!!
References: <mailman.977127721.4398.python-list@python.org>
Message-ID: <slrn93rsl2.qb2.grey@teleute.rpglink.com>

On Mon, 18 Dec 2000 01:19:51 -0700, Blair Lowe <Blair.Lowe at compeng.net> wrote:
>This is exactly what a python program should be NOT doing. "while 1" 
>is one of the most cryptic statements in Computer Science. 

    Really?  I think it is actually one of the most obvious.  Esp. when one
takes into consideration that you generally learn it once while learning how
different people define their own way of doing an endless loop is a once per
program experience.

>other languages are known for this stuff, but I do not know why python 
>doesn't auto define true and false, so that we don't have to write unreadable
>stuff.

    Maybe because defining a true and then making it false is more cryptic
than an endless while with clear escapes from the loop?  

>While "while 1" (pardon the pun) is a standard fixture that anyone 
>with any experience knows what is going on, the whole point is to 
>make all our code readable so that we, and others can maintain it and 
>reuse it easily.

    I sure hope that isn't the case.  As many people have rightfully pointed
out on a great many things regarding computers, when you design something for
idiots only idiots will use it.  "while 1:" in Python is a very easy concept
to graps compared to, say, basic file IO or, heaven forbid, regular
expressions!  If someone can decode (much less write) the following line...

    self.re = re.compile(r'^(\d{1,2}|)[Dd](\d{1,3}|)([+-]\d{1,2}|)$')

     ... then even if they have /never/ seen "while 1:" before they will know
wiat it is.  And guess what, people are going to have to know regex if they
are going to maintain other people's code.  

>Imagine the newbie looking at this statement: while 1 ... hmmm must 
>be the first condition of a while loop ????

    Imagine a newbie maintaining someone else's code.  Wait, a newbie wouldn't
be maintaining someone else's code.  A newbie would be reading through
Python's web page and hit the documentation for while at
<http://www.python.org/doc/2.0/ref/while.html> or maybe even reading the
Beazley book and hit Chapter 5, Control Flow, read up on while there and even
see an example of "while 1:".  In fact, the example given is reading lines
from STDIN until a blank line is entered.  

>It is cheap to define true and false once, or (my favourite) define 
>ever (;;). Although the definition is cryptic, the resulting code is 
>very readable. That is what we want.

    Is it?  Readable for whom?  Personally, dealing with Perl code day in and
day out, I find the "line noise" quite readable.  Others dislike it but there
are things about it I do like.  I also find regex quite readable.  Let's take
the previous example.

    ^(\d{1,2}|)[Dd](\d{1,3}|)([+-]\d{1,2}|)$

    Now, how would you rewrite that to be readable?  Well, we could expand all
of the \d's to ::DIGIT:: for example, but does that make it any more readable?

    ^(::DIGIT::{1,2}|)[Dd](::DIGIT::{1,3}|)([+-]::DIGIT::{1,2}|)$

    Does that help?  Well, lets expand a few more.  ^ and $ aren't obvious, so
let's make that ::BOL:: and ::EOL::.  Not that those are obvious either, but
at least one might be able to puzzle out what they mean.  Oh, and | for "or"?
Tsk, tsk.  Not to mention [Dd] meaning D or d?  That won't do at all, either!

    ::BOL::(::DIGIT::{1,2}::OR::)D::OR::d(::DIGIT::{1,3}::OR::)(+::OR::-::DIGIT::{1,2}::OR::)::EOL::

    Yeah, that doesn't exist, but it is a play on the POSIX regex of having
character classes defined as words to be "readable" instead of, well, being
readable.  "for ever" and "while true" are cute, but aren't really needed
except for those who are more worried about the look of the language over the
function.  "while true" dictates that true, at some point, equals false.  All
to avoid a break or next (depending on the language)?  No thanks.  "for ever"
means having multiple forms after for.  That breaks consistancy.  All for
"readabilty".  Certainly not for maintainability since consistancy, to me,
means better maintainability, not less.  If you disagree, maybe you'll tell me
why the following is so nice in Perl.

if (!$foo){
  next;
}

next if (!$foo);

unless {$foo){
  next;
}

next unless ($foo);

    I can tell you why three of the four forms are bad and, although I like
two of them, only the remaining one should be used.  Maintainability.

if (!$foo){
  next;
}

    That is the best one to use because if the statement needs to be change to
true you're not changing unless to if.  It is better than placing the
operation before the test because in that form you cannot have multiple
operations.  What if you need to add a number and then do a next?  Well, now
you have to rewrite the code.  And finally unless does not allow an "else
unless".  I don't recall if else is allowed, either.  

    So, getting back to your "for ever" statement.  Yeah, it looks pretty, but
it costs more work in the end and causes more thinking for people beyond the
novice stage.  Sure, everyone is a novice once, but hardly anyone who uses any
language for any length of time is a novice forever.

    IMHO trying to cater to the novice at the expense of maintainability for
the experts is not something to strive for in a language.  Introducing a
design where a statement is contradictory is also not something to strive for.
That is why we have "while 1:" because it is obvious to all who see it what it
is beyond, say, the first day or so with the language and it doesn't have the
problems "for ever" and "while true" impose.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From sill at localhost.kitenet.net  Wed Dec  6 18:35:29 2000
From: sill at localhost.kitenet.net (Oldayz)
Date: Wed, 06 Dec 2000 23:35:29 GMT
Subject: Newbie question on 'cmpfiles':
References: <90i074$d73$1@nnrp1.deja.com> <JmeX5.39553$II2.3513820@newsread2.prod.itd.earthlink.net> <sZnX5.153284$e5.99945@newsfeeds.bigpond.com>
Message-ID: <53AX5.51696$nh5.3969094@newsread1.prod.itd.earthlink.net>

On Wed, 6 Dec 2000 21:05:31 +1100, Srihari Vijayaraghavan <harisri at bigpond.com> wrote:
>Oldayz wrote:
>
>> On Tue, 05 Dec 2000 05:59:02 GMT, harisri at my-deja.com
>> <harisri at my-deja.com> wrote:

>Hello Andrei,
>
>Thanks. However I am having some problem while using "dircmp" class.
>
>I have the following 2 (identical) directories under my home directory: 
>/home/hari/python1 and /home/hari/python2. 
>
>I am getting few error messages, only while using the following attributes: 
>diff_files, report(), report_partial_closure(), and report_full_closure().
>
>I am enclosing the error messages for your reference. Am I doing something 
>wrong?
>
>Python 2.0 (#1, Oct 16 2000, 18:10:03)
>[GCC 2.95.2 19991024 (release)] on linux2
>Type "copyright", "credits" or "license" for more information.
>>>> from filecmp import *
>>>> x=dircmp("/home/hari/python1", "/home/hari/python2")
>>>> x.left_list
>['about.html', 'acks.html', 'api', 'dist', 'doc', 'ext', 'icons', 
>'index.html',
>'inst', 'lib', 'mac', 'modindex.html', 'ref', 'tut']
>>>> x.right_list
>['about.html', 'acks.html', 'api', 'dist', 'doc', 'ext', 'icons', 
>'index.html',
>'inst', 'lib', 'mac', 'modindex.html', 'ref', 'tut']
>>>> x.common
>['ref', 'ext', 'acks.html', 'inst', 'api', 'mac', 'about.html', 
>'modindex.html', 'tut', 'doc', 'icons', 'lib', 'dist', 'index.html']
>>>> x.left_only
>[]
>>>> x.right_only
>[]
>>>> x.common_dirs
>['ref', 'ext', 'inst', 'api', 'mac', 'tut', 'doc', 'icons', 'lib', 'dist']
>>>> x.common_files
>['acks.html', 'about.html', 'modindex.html', 'index.html']
>>>> x.same_files
>Traceback (most recent call last):
>  File "<stdin>", line 1, in ?
>  File "/usr/local/lib/python2.0/filecmp.py", line 147, in __getattr__
>    self.phase3()
>  File "/usr/local/lib/python2.0/filecmp.py", line 214, in phase3
>    xx = cmpfiles(self.left, self.right, self.common_files)
>  File "/usr/local/lib/python2.0/filecmp.py", line 288, in cmpfiles
>    res[_cmp(ax, bx, shallow, use_statcache)].append(x)
>TypeError: too many arguments; expected 2, got 4                            
>>>> x.report()
>diff /home/hari/python1 /home/hari/python2
>Traceback (most recent call last):
>  File "<stdin>", line 1, in ?
>  File "/usr/local/lib/python2.0/filecmp.py", line 241, in report
>    if self.same_files:
>  File "/usr/local/lib/python2.0/filecmp.py", line 147, in __getattr__
>    self.phase3()
>  File "/usr/local/lib/python2.0/filecmp.py", line 214, in phase3
>    xx = cmpfiles(self.left, self.right, self.common_files)
>  File "/usr/local/lib/python2.0/filecmp.py", line 288, in cmpfiles
>    res[_cmp(ax, bx, shallow, use_statcache)].append(x)
>TypeError: too many arguments; expected 2, got 4                            
> >>> x.report_partial_closure()
>diff /home/hari/python1 /home/hari/python2
>Traceback (most recent call last):
>  File "<stdin>", line 1, in ?
>  File "/usr/local/lib/python2.0/filecmp.py", line 258, in 
>report_partial_closure
>    self.report()
>  File "/usr/local/lib/python2.0/filecmp.py", line 241, in report
>    if self.same_files:
>  File "/usr/local/lib/python2.0/filecmp.py", line 147, in __getattr__
>    self.phase3()
>  File "/usr/local/lib/python2.0/filecmp.py", line 214, in phase3
>    xx = cmpfiles(self.left, self.right, self.common_files)
>  File "/usr/local/lib/python2.0/filecmp.py", line 288, in cmpfiles
>    res[_cmp(ax, bx, shallow, use_statcache)].append(x)
>TypeError: too many arguments; expected 2, got 4
>>>> x.report_full_closure()
>diff /home/hari/python1 /home/hari/python2
>Traceback (most recent call last):
>  File "<stdin>", line 1, in ?
>  File "/usr/local/lib/python2.0/filecmp.py", line 264, in 
>report_full_closure
>    self.report()
>  File "/usr/local/lib/python2.0/filecmp.py", line 241, in report
>    if self.same_files:
>  File "/usr/local/lib/python2.0/filecmp.py", line 147, in __getattr__
>    self.phase3()
>  File "/usr/local/lib/python2.0/filecmp.py", line 214, in phase3
>    xx = cmpfiles(self.left, self.right, self.common_files)
>  File "/usr/local/lib/python2.0/filecmp.py", line 288, in cmpfiles
>    res[_cmp(ax, bx, shallow, use_statcache)].append(x)
>TypeError: too many arguments; expected 2, got 4                            
>    
>Thanks in advance,
>
>Hari.      

According to 
http://sourceforge.net/bugs/?group_id=5470&func=detailbug&bug_id=124120 
this is a known bug and it can be fixed by applying the latest patch, which 
you can get at 
http://sourceforge.net/patch/?func=detailpatch&patch_id=102628&group_id=5470

-- 
Hope this helps,

	Andrei


From frank.neuber at gmx.de  Mon Dec  4 04:28:46 2000
From: frank.neuber at gmx.de (Frank Neuber)
Date: Mon, 4 Dec 2000 10:28:46 +0100
Subject: setsockopt and urllib
Message-ID: <slrn92moue.hc.frank.neuber@mars.private.de>

Hi group,
I have to set special timeouts on receive and send calls in conjunction with 
urllib-calls.
Is there a simple solution to do this without modifying the urllib itself.
For Example:
I wish to set the timeout to 60 s for the following command:
(tmpfile, hdr) = urllib.urlretrive(urlname)
It is possible to to this with the socks module call 
setsockopt(SOL_SOCKET,SO_RCVTIMEO, ttime) over the urllib module.

thanks in advance
Frank


-- 
     _/_/_/_/ _//   _/ Frank Neuber
    _/       _/_/  _/  frank.neuber at gmx.de (private)
   _/_/_/   _/ _/ _/
  _/       _/  _/_/    neuber at opensource-systemberatung.de
 _/       _/    // http://www.opensource-systemberatung.de


From ssartor at bellatlantic.net  Sat Dec 16 11:22:04 2000
From: ssartor at bellatlantic.net (Steven Sartorius)
Date: Sat, 16 Dec 2000 16:22:04 GMT
Subject: Can't get Tkinter to run from command line
Message-ID: <MEM_5.1372$L3.323917@typhoon2.ba-dsg.net>

I'm just starting out with Tkinter and have been fooling around with the
following (stored in a file called application.py)....

from Tkinter import *
from ScrolledText import ScrolledText

class gui(Frame):

	def createWidgets(self):
		self.text = ScrolledText(self)
		self.text.pack()

	def __init__(self,parent=None):
		Frame.__init__(self,parent)
		self.pack()
		self.createWidgets()

if __name__ == 'main': gui().mainloop()

This code runs fine from within python itself but I get no window if I
run it from the command line (i.e. python application.py).
I'm sure I'm doing something obviously wrong but what???

Any help much appreciated...

Thanks,

Steve


From junaftnoon at nospamplzyahoo.com  Tue Dec 19 11:39:14 2000
From: junaftnoon at nospamplzyahoo.com (June Kim)
Date: Wed, 20 Dec 2000 01:39:14 +0900
Subject: tkinter event binding to mouse double click
References: <91n7l6$7ij$1@news.nuri.net> <vpL%5.459$AH6.36619@newsc.telia.net>
Message-ID: <91o31f$vi$1@news.nuri.net>

"Fredrik Lundh" <fredrik at effbot.org> wrote in message
news:vpL%5.459$AH6.36619 at newsc.telia.net...
> "June Kim"
> > In the code following, the binding doesn't seem to work; it can't detect
> > mouse double-clicking at all.
>
> assuming you're really double-clicking in the frame widget,
> there's nothing wrong with your code.
>
> maybe you're not clicking fast enough?
>
> > (Maybe a platform-specific problem, I reckon)
>
> works on win95, win98, win NT 4, and win 2K.  maybe you
> have some other kind of win32?
>
> </F>
>

Both the codes work perfectly with "<Button-1>", single-clicking. Can't
figure out what the problem is. Am I getting too old to double click
properly?
I'm using Win98 with Korean char set. No ergonometric design either in the
os or the gadget?




From DanielK at aracnet.com  Sun Dec 24 01:42:50 2000
From: DanielK at aracnet.com (Daniel Klein)
Date: Sat, 23 Dec 2000 22:42:50 -0800
Subject: How to read keyboard input
References: <3A443B15.C7E926A@stu.ust.hk>
Message-ID: <8Qg16.261$LU6.105459@typhoon.aracnet.com>

If you want to get input from stdin then:

    >>> reply = raw_input("Do you have a dead parrot? ")
    Do you have a dead parrot? Yes
    >>> reply
    'Yes'

HTH,

Dan:

"Calvin Fong" <ee_fhw at stu.ust.hk> wrote in message
news:3A443B15.C7E926A at stu.ust.hk...
> Dear all,
>     I'm finding a fcuntion to read the input of keyboard
>     is it something like getchar, i just get lost.
>     anyone know it?
> Yours
> Calvin
>




From lg at rgz.ru  Mon Dec 25 06:16:55 2000
From: lg at rgz.ru (lg at rgz.ru)
Date: 25 Dec 2000 14:06:55 +0250
Subject: is there any troubles with select.select in python
References: <3a47297e@news.edunet.ru>
Message-ID: <3a472a4f@news.edunet.ru>

lg at rgz.ru wrote:
> subj?
> i've got python script with long time select.select waiting with no timeout
> and in half a hour or may be hour .. python exited on signal 11.
> gdb shows the top of stack - 0x80b1649 in select_select (self=0x0, args=0x829c4cc) at ./selectmodule.c:177
I forgot to mention that i'm using 
Python 2.0 (#35, Oct 29 2000, 15:16:39) 
[GCC 2.95.2 19991024 (release)] on freebsd4.0-release


-- 
zev


From dale at out-think.NOSPAMco.uk  Thu Dec  7 16:19:03 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Thu, 07 Dec 2000 21:19:03 +0000
Subject: Dare I say it...
Message-ID: <0ivv2tk3h07rp31c2mtheqsns6voriqq71@4ax.com>

Unicode again.

Shouldn't there be a struct.pack format chacter to produce a Unicode character?

Or have I missed something here?

Thanks
--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants


From aleaxit at yahoo.com  Tue Dec 19 04:24:56 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 19 Dec 2000 10:24:56 +0100
Subject: Python in "math through programming" curriculum
References: <2g2t3t06r2agrrjmfnr41eu4lc2u39a8qi@4ax.com> <3A3EFF47.67DE69CF@gte.net>
Message-ID: <91n9i30pr7@news2.newsguy.com>

"Clifford J. Nelson" <cnelson9 at gte.net> wrote in message
news:3A3EFF47.67DE69CF at gte.net...
    [snip]
> > Awhile back, Tim Peters recommended 'Concrete Mathematics'
> > (Knuth a co-author) as a good example of a "math through
> > programming" text book.  I subsequently bought it, and
> > had to agree.  So how do we "lower a ladder" to middle
> > and high schoolers starting from there?  I think Python
> > is part of the answer.
>
> I don't think Python should be a standard at all, because of
> my experience with it on the Mac. Keep the options open for
> computer programming languages.

I'm not sure what being 'on the Mac' has to do with it.
Either it's irrelevant, or you're letting platform issues
(e.g., integration with native GUI's) cloud judgment on
_language_ issues.

I'm a "multilanguage GUI" -- I strongly believe that most
software development projects (of any substantial richness
and complexity) are best served by a mixture of different
programming languages, interoperating smoothly in various
components and/or as 'glue' among those components.  Thus,
I'll hardly object to 'keeping the options open' -- indeed,
I believe smooth interoperation between different languages
is a key platform consideration (dunno what Mac has to offer
for it, and I _am_ worried that Linux is weakish in such
multilanguage-oriented infrastructure; MS is leading the
way with their '.NET' initiative, which at low levels is
ALL about multilanguage interop... let's hope it spreads!).

But you won't typically hear arguments against this from ANY
Pythonista -- Python is *ABOUT* "playing well with others"...
*designed* and *implemented* to interoperate, like only a few
other languages; most languages appear to be designed "as if"
they were to be the sole implementation language for a typical
project, with "foreign function interfaces" or the like that
sometimes look as if they were tacked on as an afterthought.
And that may be because many people disagree with my "many
languages are better than one" stance -- they'd much rather
stay monolingual, at least within a given project.


But anyway, it would not make sense to argue _against_
"making a language a standard" in one field, because other
languages may prove better at some different task, or
sub-task.  Standardization offers substantial advantages
('network externalities' in economists' parlance), and
the 'keeping open to other languages' part need not
interfere with such advantages at all.

For the specific task of "teaching some maths to middle
and high schoolers", for example, establishing some sort
of 'standard' language would have enormous benefits; and
the relevant issues are quite different from those that
would affect 'standardization upon one base language' in
other areas.  For example, issues of performance on large
real-world tasks, which are a weak point of Python (and
to a lesser extent, lazy-semantics functional languages
such as Clean and Haskell) in comparison to traditional
lower-lever languages such as C++ (and, to a lesser extent,
non-lazy functional languages, such as ML variants), and
thus provide a good part of the motivation for projects
to use multiple languages, are not all that relevant when
the task at hand is _teaching_.


On the other hand, I think an issue of crucial importance
is 'paradigm choice'.  Some languages espouse a specific
programming paradigm very strongly: it would make little
didactical sense to use C for functional or O-O programming,
for example -- or Haskell for anything BUT (lazy-centered)
functional programming, or Smalltalk (or Java) for anything
BUT O-O programming, &c.

Other languages are multi-paradigm: C++, O'Caml, Python,
just to name three, let you use several paradigms together,
in reasonably smooth ways.  The price of that is typically
_complexity_ -- but Python is an exception... it "reaches
out" to multiple paradigms *through simplicity*.  This,
IMHO, is part of what makes it such an outstanding language
for teaching: you can use it to help students learn to
wrap their minds around programming problems in different
ways, _without_ burdening them with complication, including
"strange" syntax.

People who "know" that the only one "right" way to program
is through a single given paradigm 'X', for any 'X', may
have little interest in Python (or C++ or O'Caml...).  If
X is 'lazy functional programming', then Haskell and Clean
have it covered; if it's 'statically typed object-oriented
programming', Eiffel or Sather; if it's 'dynamical object
oriented programming (single-inheritance)', Smalltalk; and
so on, and so forth (not to be confused with 'so Forth':-).

People who 'know' the very opposite -- that multiple paradigms
widen one's mind, and are typically needed for rich, complex
programming tasks in the 'real world' -- may still be rightly
reluctant to give multi-paradigm exposure to students at the
price of using a rich, composite, complicated language.  But
Python lets you have your paradigms and eat them too...!-)



> Here are two:
>
> Clean at: http://www.cs.kun.nl/~clean/
>
> Caml at:  http://pauillac.inria.fr/caml/
>
> Both of those languages are OOP too, and easier to use than Python.

I dispute both of your claims.  _O'Caml_ (the current dialect)
does have OO (as well as other paradigms: the 'basis' is non-lazy
FP, but OO and laziness can be quite smoothly 'tacked on'), but
'Caml' without the leading 'O' doesn't.  And Clean doesn't even
*claim* any OO-ness -- "The programming language Clean is a general
purpose, state-of-the-art, pure and lazy functional programming
language", states
http://www.cs.kun.nl/~clean/About_Clean/about_clean.html,
and continues "designed for making real-world applications" -- a
key issue: it's NOT designed for ease of learning, most particularly
NOT for ease of learning by somebody not yet steeped in higher
maths!  If a lazy, strongly-typed FP language were to be used as
somebody's first introduction to programming, I think Haskell or
HUGS might prove easier (but I have no real experience in attempting
such teaching tasks); Clean's claimed advantages over Haskell are
centered on performance issues, minor ones in teaching.  OTOH,
practical disadvantages include (last time I looked, at least)
no Linux support; the latest GUI library is currently Windows
only, with a hope to have a Macintosh port -- no Linux plans that
I know of.  This looks like a huge practical handicap to me!

Caml, IMHO, has rather cluttered syntax when compared to Haskell
(or Python!); it only gets worse when the Object-oriented extensions
are also considered.  Admittedly, a minor consideration, but, _for
the specific issue of programming-newbies_, not a negligible one
IMHO.  I consider the full O'Caml language too rich and complex
for an _introduction_ to programming, although, exactly because
it is so big, I guess it could be "subsetted" with some success
(just as those [IMHO] so misguided as to use C++ or Ada for a first
course in programming basically _have_ to subset it heavily to
have any chance -- O'Caml is NOT as huge/complex as C++ or Ada,
so the subsetting would not be as 'heavy' &c, fortunately!).

Again, the advantage of O'Caml vs, say, other more standard ML
variants, is basically one of _performance_ (in fact, Clean might
claim that its superior performance wrt Haskell comes in part
from a 'superior', better-cohesive, underlying conceptual model;
I don't think any O'Caml advocate would claim better conceptual
cohesion that standard ML...), stemming in part from richness.
These are not issues that would recommend the language for the
specific purpose of teaching a programming-introduction course.


In fact, if I _did_ have to base a first course in programming
on a purely-functional language, I think I'd choose Erlang -- not
without some misgivings (Erlang's concurrency model may not be
easy for programming beginners to grasp...!), but I strongly
suspect the dynamic typing model would ease my life as an
instructor, by enough to make the choice worthwhile.  (Scheme,
Lisp, or Dylan would also offer some dynamic typing, of course,
but then, they're not purely functional -- and there are other
issues with their use in didactic settings).

But my Numero Uno choice would still be Python, unless strong
external constraints precluded it.  And I hope I've shown this
is NOT because of a lack of open-mindedness, or interest in
other languages, but rather comes from a wide (if sometimes
not as deep as one would like!) examination of the field --
although I'll freely admit that teaching total beginners,
particularly ones of middle-to-high-school ages, has NOT been
an extensive part of my experience in the programming field,
and so I'll gladly listen to 'domain experts'... at least, if
their experiences extend to two or more languages and paradigms
in ways that make fruitful comparison and contrast feasible!


Alex





From richard at NOstarfighterSPAM.freeuk.com  Sat Dec 16 06:24:18 2000
From: richard at NOstarfighterSPAM.freeuk.com (Richard)
Date: Sat, 16 Dec 2000 11:24:18 -0000
Subject: __del__() in a Tkinter app
Message-ID: <MlI_5.53768$eT4.4433606@nnrp3.clara.net>

I have tried adding a __del__() method to a Tkinter application so it can
remove temporary files when the program exits.

However the __del__() method is never called, as in this simple example:

class GUI:
    def __init__(self):
        self.root = Tk()

        quit_pb = Button(self.root, text="Quit")
        quit_pb.bind("<Button>", self.quit)
        quit_pb.pack(side=LEFT)

        self.root.mainloop()

    def __del__(self):
        print 'GUI destructor called'

    def quit(self, event=None):
        self.root.quit()

class NonGUI:
    def __del__(self):
        print 'NonGUI class destructor called'

if __name__ == '__main__' :

    g = GUI()
    print 'after creating GUI object'

    n = NonGUI()
    print 'after creating NonGUI object'

When the app is run and the Quit button is pressed, this produces:

NonGUI class destructor called

i.e. the destructor in the NonGUI object is called, but the destructor in
the GUI object is not.

Does anybody know why this happens? Is it possible to fix it so the GUI
object destructor is called?

Richard





From guy at obstruction.com  Wed Dec  6 23:22:53 2000
From: guy at obstruction.com (Guy Middleton)
Date: 6 Dec 2000 23:22:53 -0500
Subject: reading X resources from Tkinter
Message-ID: <90n3at$130f$1@chaos.obstruction.com>

I just got my copy of Grayson's "Python and Tkinter Programming" (very
nice book, by the way).

He says that one can set entries in the "Tk option database" with
Tk().option_readfile(), and that the syntax of the file read is the same
as the .Xdefaults or .Xresources file.

It's not clear to me (maybe I am not reading carefully enough) if one can
query the X resources directly, or if you can copy X resources into the
option database.  Is this possible?  If it is, how?


From sill at localhost.kitenet.net  Wed Dec 13 22:39:15 2000
From: sill at localhost.kitenet.net (Oldayz)
Date: Thu, 14 Dec 2000 03:39:15 GMT
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <kuzoi15d60.fsf@lasipalatsi.fi> <Pine.BSF.4.10.10012122310510.48551-100000@tcfreenet.org> <u7l54ggyx.fsf@ctwd0143.fitlinxx.com>
Message-ID: <DhXZ5.485$UN1.55303@newsread2.prod.itd.earthlink.net>

On 13 Dec 2000 16:16:06 -0500, David Bolen <db3l at fitlinxx.com> wrote:
>Aaron Jon Reichow <reichowa at tcfreenet.org> writes:
>
>> On 13 Dec 2000, Erno Kuusela wrote:
>> 
>> > the thing that struck me as a little foreign was the apparent
>> > isolation of the smalltalk image from the rest of the operating system
>> > and the filesystem. i want to be able to grep code!
>> 
>> While I'm not a proponent of Smalltalk isolationism, grep is a useless
>> tool (more or less), simply an artifact, ye, a symptom of the
>> sickness known as file-based development.
>
>I'm not sure I'd call it a "sickness" since any development model is
>just that - a model - it may be better or worse than some other model,
>but at least for me, a file-based model works well in many cases.
>
>My biggest problem with the integrated models (with database or other
>repository back-ends) tends to be that often they aren't yet fleshed
>out enough, or in many cases flexible enough, to replicate all the
>functionality I can get out of a file based system with a current
>collection of file based tools that have had lots of time to evolve
>for efficiency and flexibility.
>
>With file-based environments, I generally feel safe that if the
>environment itself doesn't support something, there are enough tools
>that understand the back-end repository (e.g., files) that I can get
>what I want even if I have to go outside the integrated system.  That
>may not be possible with a closed back-end repository.
>
>As was noted elsewhere in this thread, and to tie a little more back
>into Python, this can also be an issue with Zope.  For example, Zope
>may provide grep-like functionality that works through the object
>store, but it may be difficult to replicate what in a filesystem might
>be a grep piped into an awk to produce summary information.  Now maybe
>the right answer is that the same functionality should be done in some
>other manner with Zope, but it may also be a point of needed
>functionality just not yet being supported by the development model.
>(I know in my case there's also a little "queasiness" just because of
>a feeling of lack of control over the store, which I think is just an
>side-effect of my primary file-based focus for so long).
>
>But if the repository model is flexible enough to permit the eventual
>creation of the smaller tools and then their coupling on an as-need
>basis (which Zope may well be for example), then it's more a question
>of a minimal set of functionality to feel comfortable to those of us
>more used to the file system and utilities and tools that we've gained
>proficiency in - that tool proficiency generally leading to
>development efficiency.
>
>--
>-- David

On a related note, I personally dislike IDE/integrated
approach because of exactly one example of missing
functionality you're talking about - I'm addicted
to vi(m) commands, and I can't imagine working with
large bodies of text with regular pico-style commands
(or even emacs-style). Slrn and mutt can both pass on
the task of editing to vim so this approach is really
flexible. Besides, vim supports python as a scripting
language.


-- 

	Andrei


From pbjorn at uswest.net  Wed Dec 27 16:19:36 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Wed, 27 Dec 2000 14:19:36 -0700
Subject: newbie : array initialization, anyone ?
References: <3a4a1352.9818548@news.t-online.de> <wgr26.1691$JG.213419@news.shore.net>
Message-ID: <3A4A5CE8.AC3EBFD9@uswest.net>

arcege at shore.net wrote:

> Gerson Kurz <gerson.kurz at t-online.de> wrote:
> : you can initialize a fixed-size array using something like
>
> :   array = [0] * 10
>
> : now, one would think that you can initialize a multidimensional array
> : like this:
>
> :   multidim = [ [0] * 10 ] * 3
>
> : but now, all 3 elements in multidim are references to the same single
> : array, so that if you type
>
> :   multidim[0][0] = 1
> :   print multidim
>
> : you actually get three 1 in all those 0s. Where did my thinking go
> : wrong along these lines ?
>
> This is in the FAQ (#4.50).  You will want to create each inner dimension
> individually:
>
> >>> multidim = [ None ] * 3
> >>> for i in range(3):
> ...   multidim[i] = [ 0 ] * 10
> ...
> >>>

The FAQ should probably be changed to something like:

  multidim = [ ([0] * 10) for i in range(3) ]

-- bjorn



From kern at caltech.edu  Wed Dec 20 00:36:12 2000
From: kern at caltech.edu (Robert Kern)
Date: Wed, 20 Dec 2000 00:36:12 -0500
Subject: embedding help
References: <3A3F0EFC.D2F2BAE9@ic.sunysb.edu> <91nnek0166d@news2.newsguy.com>
Message-ID: <3A40454C.D8F85313@caltech.edu>

[snip]

> For C-level work, the ideal read would be "How to
> write a Python extension",
> http://starship.python.net:10080/crew/arcege/extwriting/pyext.html
> except that accessing anything on Starship is always
> a bother these days (I wonder if mirrors of this may
> be found anywhere else...?).  

Remove the port specification and then the URL works.

[snip]

Robert Kern


From max at alcyone.com  Tue Dec 12 11:59:22 2000
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 12 Dec 2000 08:59:22 -0800
Subject: [Q] Are Exceptions used that much in practice?
References: <3A35D546.A4731B5B@enteract.com>
Message-ID: <3A36596A.89EEA210@alcyone.com>

Jerome Mrozak wrote:

> I realize that Python allows me to catch exceptions.  But being a
> novice
> I'm not sure if they are used much in the 'real world'.
	...
> Another way of putting it is, if the try-catch-else didn't exist would
> anyone be terribly put out?  (If not then the construct isn't much
> used...)

Yes.  Yes.  Yes.

Anyone who has done substantial programming in Python, Java, or C++ has
used exceptions extensively.  If you haven't found a significant use for
exceptions, then you're either writing extremely trivial programs (ones
so trivial they cannot fail, which has to be _really_ trivial) or you're
not handling error conditions properly.  Truth be told, handling
arbitrary error conditions properly is one of the more difficult things
to do in computer science.  Exceptions, although they require a
significant change in the way you write code, make this a lot easier. 
Draft Standard C++ programming without exceptions was obnoxious.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Love is not love which alters when it alteration finds
\__/ William Shakespeare, _Sonnets_, 116
    Alcyone Systems' CatCam / http://www.catcam.com/
 What do your pets do all day while you're at work?  Find out.


From tim.one at home.com  Thu Dec 28 05:15:29 2000
From: tim.one at home.com (Tim Peters)
Date: Thu, 28 Dec 2000 05:15:29 -0500
Subject: split hangs
In-Reply-To: <slrn94l4d3.550.jp@localhost.localdomain>
Message-ID: <LNBBLJKPBEHFEDALKOLCIELKIFAA.tim.one@home.com>

[Jacek Pop?awski]
> ourregex=re.compile(r'(<(?:[^<>"]+(?:"[^"]*")?)+>)')

Nesting + and * operators is likely to yield a regexp that takes time
exponential in the length of the string, and especially when the overall
regexp doesn't match (the regexp tries all possible nested ways but none
works in the end).  Here's a small variation of Harald Kirsch's program;

import re, sys, time

boing = re.compile(r'(<(?:[^<>"]+(?:"[^"]*")?)+>)')

s = 'p < 0.05, all data in arbitrary'

for i in range(1, len(s)):
    t = s[:i]
    start = time.clock()
    boing.search(t)
    finish = time.clock()
    elapsed = finish - start
    print "%.3f seconds for '%s'" % (elapsed, t)

And here's as much as it printed before I got bored waiting for it:

0.000 seconds for 'p'
0.000 seconds for 'p '
0.000 seconds for 'p <'
0.000 seconds for 'p < '
0.000 seconds for 'p < 0'
0.000 seconds for 'p < 0.'
0.000 seconds for 'p < 0.0'
0.000 seconds for 'p < 0.05'
0.000 seconds for 'p < 0.05,'
0.000 seconds for 'p < 0.05, '
0.000 seconds for 'p < 0.05, a'
0.000 seconds for 'p < 0.05, al'
0.001 seconds for 'p < 0.05, all'
0.003 seconds for 'p < 0.05, all '
0.003 seconds for 'p < 0.05, all d'
0.007 seconds for 'p < 0.05, all da'
0.014 seconds for 'p < 0.05, all dat'
0.028 seconds for 'p < 0.05, all data'
0.086 seconds for 'p < 0.05, all data '
0.172 seconds for 'p < 0.05, all data i'
0.406 seconds for 'p < 0.05, all data in'
0.581 seconds for 'p < 0.05, all data in '
0.854 seconds for 'p < 0.05, all data in a'
1.708 seconds for 'p < 0.05, all data in ar'
3.412 seconds for 'p < 0.05, all data in arb'
6.836 seconds for 'p < 0.05, all data in arbi'

The tail end of the output is clear:  adding another character to the string
doubles the time it takes to figure out that the regexp doesn't match.  This
isn't a bug, it's the way Python (and Perl, and Emacs, and ...) regexps
work.  A full explanation is quite involved; you can find one in Jonathan
Friedl's excellent book "Mastering Regular Expressions" (O'Reilly).

For Harald's particular flavor of non-matching test data, the regexp can be
made very fast by adding a positive lookahead assertion, to make sure
there's *some* hope of eventually matching the ">":

boing = re.compile(r'(<(?=>)(?:[^<>"]+(?:"[^"]*")?)+>)')
                       ^^^^^ new lookahead inserted

A more robust way is to use what Friedl calls "unrolling":

boing = re.compile(r'(<[^<>"]+(?:"[^"]*"[^<>"]*)*>)')

I'm afraid that understanding why that pays off requires a non-trivial grasp
of how the matching engine works; see Friedl for a full explanation.

or-use-a-real-html-parser-ly y'rs  - tim




From othello at javanet.com  Fri Dec 29 14:49:31 2000
From: othello at javanet.com (Raymond Hettinger)
Date: Fri, 29 Dec 2000 14:49:31 -0500
Subject: 2001 Enchancement Wishlist
Message-ID: <3A4CEACB.220998A0@javanet.com>

If these ideas meet with favor, I'll write-up a PEP.

1.  List.append(elem) and List.pop() are great.  Let's add two more
     methods, List.prepend(elem) and List.leftpop().  The goal is to
     make queues,deques, and reverse stacks easier, faster, and clearer
     to implement.

2. There is already a PEP to extend the eval() function which is currently
    limited to dictionaries.  The goal is to accept any object defining
    __getitem__() so that any mapping object can be used to specify the
   globals and locals.   I recommend that we extend the same courtesy to
   the % string formatting operator.  For example:

             print "Spam with %s, eggs, and spam" % aUserDictionary

3.  __init__ is the only special method that doesn't allow a return value.
   I think it should continue to return the new object by default but also
   allow another return item to be substituted.  For example:

            def __init__( self ):
                   if MyClass.aSingleton == None:
                         MyClass.aSingleton = self
                   return MyClass.aSingleton

4.  Allow dictionaries to respond to sequence operations as if .keys() were
     in the list object.  Examples:

    knights = { 'lancalot':'good', 'gallahad':'brave', 'robin':'coward' }
    for name in knights:
           print name, knights[name]
    print knights.len()
    if 'arthur' in knights:
           print 'Not the King'
    print knights[0]              # may be any one depending on the hash table
order
    try:
        print knights[5]
    except IndexError:
        print 'Too few knights for the ladies in the castle'
    del knights['lancelot']      # regular mapping operations left unchanged.

I believe all of the enhancements can be added without impacting existing
programs.
Hope suggestion #4 isn't too radical.

Happy New Year,


Raymond Hettinger
othelloNOSPAM at javanet.com



From billtj at my-deja.com  Wed Dec 27 09:55:44 2000
From: billtj at my-deja.com (billtj at my-deja.com)
Date: Wed, 27 Dec 2000 14:55:44 GMT
Subject: A Suggestion for Python Dictionary/Class
References: <920fe9$c20$1@hecate.umd.edu> <920m9h011uh@news2.newsguy.com>
Message-ID: <92cvte$b0i$1@nnrp1.deja.com>

Thanks for all the suggestions in making the Python class more like
C/C++ struct or class.  However, it looks that the current solution is
based on a "convention" to derive Python classes from a certain class.

I am not too familiar with how the Python interpreter works.  But based
on my current understanding, the current Python class is implemented
using dictionary, unlike in C/C++.  Therefore Python's

    var_1.member_1.member_2.member_3

will invoke three dictionary lookups, while C/C++'s

    var_1.member_1.member_2.member_3

is a (short) constant time operation, after compilation, of course.

My question is whether by defining a special class (call is struct?) in
Python will result in better Python run-time performance.  For example,

    struct MyStruct:
        def __init__ (self):
            self.member_1
            self.member_2
            self.member_3

Because it is a "struct" instead of a "class", we cannot add any new
member.  Can then the Python interpreter be optimized to handle struct's
so that the run-time performance is better than a class?  Coming from
C/C++ background and never saw the Python C code, I imagine that the
interpreter allocates memory space for three labels (which are
pointers/handles to Python objects?) and during
interpretation/compilation, Python replaces the references for member_1,
member_2, and member_3 simply to the pointers to the corresponding
labels.  (I may not be precise in describing this, but hopefully you get
the idea).  So no dictionary lookup is involved.

Any comments?

Bill


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


From loewis at informatik.hu-berlin.de  Wed Dec 27 20:12:11 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 28 Dec 2000 02:12:11 +0100
Subject: adding fields to picked structures
References: <3A478D58.F1B2ABA@uniserve.com>
Message-ID: <p6qr92te4d0.fsf@informatik.hu-berlin.de>

Bob van der Poel <bvdpoel at uniserve.com> writes:

> So, I suppose I could write a conversion utility...but, I'd like to
> avoid that if possible. Am I missing something...or is this just the way
> life should be?

Your new class could implement a __setstate__ method, which could
analyse the structure it receives, and set the missing fields to some
default value.

Regards,
Martin



From gregj at pobox.com  Thu Dec 21 02:18:07 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Thu, 21 Dec 2000 07:18:07 GMT
Subject: Need Help sorting alphabetically.
References: <3A3F1A91.A7ED9182@araidesign.com> <slrn93u8qc.5lh.scarblac@pino.selwerd.nl> <OlY%5.172268$U46.5446765@news1.sttls1.wa.home.com> <91q415$nmm$1@nnrp1.deja.com>
Message-ID: <P8i06.176416$U46.5556469@news1.sttls1.wa.home.com>

<eff at my-deja.com> wrote:

> ...but they don't "throw out all other characters".  the original
> code matches his description, your proposed replacement doesn't.

I thought I was in your kill file already. Is it not working?

My proposed solution addresses the original problem as stated:

> > I would like to sort alphabetically (I'm not sure that is the correct
> > term) so that strings of alpha numerics sort alphabetically instead of
> > by ASCII order.

My code does that. It was someone else who proposed throwing out all the
non-alphabetic characters, which to me seemed wrong because the original
question specifically mentioned "alpha NUMERICs."

In any case I was legitimately trying to help, even if I didn't guess that
the question was really about sorting file names that may or may not be
missing leading zeroes in a numeric portion of the name. I wasn't trolling
the newsgroup looking to pick a fight.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From fit0298 at fitec.co.jp  Wed Dec 13 05:47:43 2000
From: fit0298 at fitec.co.jp (ARIMA Yasuhiro)
Date: Wed, 13 Dec 2000 19:47:43 +0900
Subject: Ruby, Python, Eiffel and Smalltalk
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <mdne3t8q8670mbplnle1jrettaqnf8bssq@4ax.com>
Message-ID: <3A3753CF.2DFA1973@fitec.co.jp>

Welcome to Ruby world.
Append some informations.

israel raj thomas wrote:
> 
> >Do you think that it would be possible to use Python as a Smalltalk
> >replacement if an IDE existed that hid some of Python's limitations?
> 
> Would it perhaps not be better to use Ruby ?
> See below for information and for comparisons of Ruby, Python, Eiffel
> and Smalltalk:

news:comp.lang.ruby

http://www.ruby-lang.org/
http://www.rubycentral.com/


From hamish_lawson at yahoo.co.uk  Wed Dec 27 16:12:37 2000
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: Wed, 27 Dec 2000 21:12:37 GMT
Subject: newbie : array initialization, anyone ?
References: <3a4a1352.9818548@news.t-online.de>
Message-ID: <92dlvu$tm1$1@nnrp1.deja.com>

    array = [0] * 10

    multidim = [ [0] * 10 ] * 3

It should be noted that *both* 'array' and 'multidim' are lists that
consist of references to the same respective single object; in the case
of 'multidim' that shared object is the list object [0, 0, 0, 0, 0, 0,
0, 0, 0, 0], while in the case of 'array' that shared object is the
integer object 0.

A list can be modified and all references to it will see that change.
The statement

    multidim[0][0] = 1

is modifying the list multidim[0]; this list is the same one referred
to by multidim[1] and multidim[2], so they will reflect the change.

Likewise array[0] through array[9] refer to the same object, the
integer 0. *If* there were operations available that allowed you to
modify integers, then it would be possible to modify the object
referred to by array[0] and have that change reflected in array[1]
through array[9]. (But of course Python doesn't provide operations to
modify integers, since it wouldn't make sense to be able to change the
zeroness of 0 - though of course Python does provides operations for
producing integers from other integers, e.g. 2 + 3).

To get independent copies of [0] * 10 you could do this (in Python 2.0):

    multidim = [([0] * 10)[:] for i in range(3)]

or more efficiently:

    zerolist = [0] * 10
    multidim = [zerolist[:] for i in range(3)]

or also more generally:

    def duplist(a, n=1):
        return [a[:] for i in range(n)]

    multidim = duplist([0] * 10, 3)


Hamish Lawson

---------------------------

Gerson Kurz wrote:

> you can initialize a fixed-size array using something like
>
> 	array = [0] * 10
>
> now, one would think that you can initialize a multidimensional array
> like this:
>
> 	multidim = [ [0] * 10 ] * 3
>
> but now, all 3 elements in multidim are references to the same single
> array, so that if you type
>
> 	multidim[0][0] = 1
> 	print multidim
>
> you actually get three 1 in all those 0s. Where did my thinking go
> wrong along these lines ?
>


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


From slhath at home.com  Tue Dec 19 19:29:47 2000
From: slhath at home.com (Scott Hathaway)
Date: Wed, 20 Dec 2000 00:29:47 GMT
Subject: xml-rpc module question
References: <dmC%5.75166$yR4.2151642@news1.rdc1.tx.home.com> <j7L%5.879$VF3.1528@news.more.net>
Message-ID: <%3T%5.76381$yR4.2212181@news1.rdc1.tx.home.com>

It is at a level equal with call.

I made the changes you suggested, but I am still having trouble.

Anyone have any ideas?

Scott

"Jeff Kunce" <kuncej at mail.conservation.state.mo.us> wrote in message
news:j7L%5.879$VF3.1528 at news.more.net...
> > I have taken the xml-rpc module and added the following to the call
area:
> >
> > def CubeIt(x):
> >    return x**3
> >
> > def call(self, method, params):
> >    if method == 'CubeIt':
> >    return CubeIt(params)
> >
>
> This is a wild guess, but did you define CubeIt() at the
> global level of the module, or was it indented the same
> as call() - making it a method of the RequestHandler class?
>
> If the former, I'm not sure what the problem is.
>
> If the latter, try:
>
>     def CubeIt(self, x):
>         return x**3
>
>     def call(self, method, params):
>         if method == 'CubeIt':
>         return self.CubeIt(params)
>
>
>   --Jeff
>
>
>




From jmurray at agyinc.com  Tue Dec 19 14:08:47 2000
From: jmurray at agyinc.com (Joe Murray)
Date: Tue, 19 Dec 2000 11:08:47 -0800
Subject: Tkinter text widget dump method ?
Message-ID: <3A3FB23F.E73D5286@agyinc.com>

Dear All,

In versions Python v 1.5.2 and > and Tkinter v 1.129 and >, there are no
implementations of the Tkinter.Text widget dump methods.  I checked out
the Tkinter docs and it mentions something about the dump method being
implemented.  Is anyone aware of a resolution?  I was confused by the
pythonware documentation.  The dump method would be useful when
outputting formatted text in a 'memorable' way.

Thanks,

joe

-- 
Joseph Murray
Bioinformatics Specialist, AGY Therapeutics
290 Utah Avenue, South San Francisco, CA 94080
(650) 228-1146



From rodsenra at correionet.com.br  Fri Dec  8 11:20:31 2000
From: rodsenra at correionet.com.br (Rodrigo Senra)
Date: Fri, 08 Dec 2000 14:20:31 -0200
Subject: parallel port access?
References: <3A268378.4DA1B831@sympatico.ca> <3A272881.6F444750@engcorp.com> <90lm04$b3r$1@nnrp1.deja.com> <slrn92str2.12g.ssthapa@localhost.localdomain>
Message-ID: <3A310A4F.297D94E8@correionet.com.br>

Suchandra Thapa wrote:
>     You can directly read and write ports on windows 95/98/me(?) but
> you need to create a vxd to access ports on NT and 2000.

AFAIK, there are no vxd under NT neither 2000. VxD's exclusively
dwell under Win9x (since W4Wg when they were called .386). Though,
there is a vxd in Win9x that emulates WinNT miniport-driver layer.

regards
Rod
-- 
Rodrigo Senra         
Computer Engineer   (GPr Sistemas Ltda)  rodsenra at correionet.com.br 
MSc Student of Reflection (IC- UNICAMP) Rodrigo.Senra at ic.unicamp.br
http://www.ic.unicamp.br/~921234 (see also   http://www.gpr.com.br)





From ljohnson at resgen.com  Wed Dec  6 18:11:26 2000
From: ljohnson at resgen.com (Lyle Johnson)
Date: Wed, 6 Dec 2000 17:11:26 -0600
Subject: sendmail: under Windows NT/2000
References: <90mapg$u7m$1@nnrp1.deja.com>
Message-ID: <t2thnjiksfqj7a@corp.supernews.com>

You are correct that invoking "sendmail" pretty much ties you down to Unix
platforms. Take a look at the platform-independent "smptlib" module (part of
the standard Python library) to see if that could be used instead.

"Shicheng" <S.Tian at shu.ac.uk> wrote in message
news:90mapg$u7m$1 at nnrp1.deja.com...
> Hello all,
> I test a python cgi sendmail program; in the code
> there are the following lines:
>
> ...
> sendmail_cmd = "/usr/lib/sendmail -oi "
> ...
> try: mfd = os.popen(sendmail_cmd + real_to, "w")
> ...
>
> I think the string "/usr/lib/sendmail -oi " suggests
> that the code should be used under the UNIX platform.
>
> My query is: under Windows NT/2000, how could I change the
> above line of code so that I could make use of the NT/2000
> mailing facility in the code?
>
> Thanks,
>
> Shicheng
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.




From jflores at codeit.com  Mon Dec 18 13:58:41 2000
From: jflores at codeit.com (Julio Flores Schwarzbeck)
Date: Mon, 18 Dec 2000 10:58:41 -0800
Subject: =?iso-8859-1?Q?Re:_More_docs_in_Spanish/M=E1s_madera?=
In-Reply-To: <91ljd7$17v$1@nnrp1.deja.com>
Message-ID: <5.0.2.1.0.20001218105723.00a232e0@mail.codeit.com>

I believe you meant http://pySpanishDoc.sourceforge.net/

Probablemente te refieres a: http://pySpanishDoc.sourceforge.net/

Saludos,

At 06:01 PM 12/18/00 +0000, rapto at arrakis.es wrote:
>New files in
>Ficheros nuevos en
>http://pySpanishDoc/sourceforge.net/
>
>
>Sent via Deja.com
>http://www.deja.com/
>--
>http://www.python.org/mailman/listinfo/python-list




From nospam at ktgroup.co.uk  Thu Dec  7 04:55:06 2000
From: nospam at ktgroup.co.uk (nospam at ktgroup.co.uk)
Date: 07 Dec 2000 09:55:06 +0000
Subject: Schwartzian Transform (was Re: Sort)
References: <NlnX5.1444$yv1.2700363@nnrp5.proxad.net> <oa8s2tsls5foll62r8fpa6nlcpisb6v1ek@4ax.com> <90l9ir02avk@news2.newsguy.com> <3A2E5F2A.9E650D50@erols.com> <90lrhl06t@news2.newsguy.com> <Pine.NXT.4.21.0012061411450.230-100000@localhost.virginia.edu>
Message-ID: <ud7f4369x.fsf@ktgroup.co.uk>

"Steven D. Majewski" <sdm7g at Virginia.EDU> writes:
> This is however, a pretty ancient Python idiom, so I wonder
> about attributing it to Randall's invention. 

It's an even more ancient sorting idiom. I've just checked a textbook published
in 1975 which describes sorting just the keys instead of an entire record as an
old established technique! It even discusses why you might want to do this when
you're sorting on some function of various fields and need a stable sort.


From Edgar at EdgarSchwarz.de  Sat Dec 23 17:31:40 2000
From: Edgar at EdgarSchwarz.de (Edgar at EdgarSchwarz.de)
Date: 23 Dec 2000 22:31:40 GMT
Subject: mod_python.dll found
References: <921ps0$l2e$1@riker.addcom.de>
Message-ID: <92394c$uri$1@riker.addcom.de>

Hi,
I got an email giving me an URL where to find it (with another name).
So my problem is solved.

Merry Christmas, Edgar

-- 
edgar at edgarschwarz.de                    http://www.edgarschwarz.de
*          DOSenfreie Zone.        Running Native Oberon.         *
Make it as simple as possible, but not simpler.     Albert Einstein


From andreas at andreas-jung.com  Mon Dec 18 03:18:19 2000
From: andreas at andreas-jung.com (Andreas Jung)
Date: Mon, 18 Dec 2000 09:18:19 +0100
Subject: PIL: how to do in-memory transformations ?
Message-ID: <20001218091819.A15144@yetix.sz-sb.de>

I am faced with the following problem using PIL:

- a function gets an image filename and a scale factor
- the function should read the image and determine its type
- the function should scale the image by the given scale factor
- the function should return the binary data of the resized image 
  as a python string (image format must be kept (jpg/gif/...)

I have a solution that works with temporary files on the filesystem
but I need a more elegant and efficient solution that does not
use temporary files. Due to the poor PIL documentation I have
not found a suitable solution yet. Any ideas ?

Tnx, Andreas





From ben at semio.com  Wed Dec  6 20:49:38 2000
From: ben at semio.com (Ben Mitchell)
Date: Wed, 6 Dec 2000 17:49:38 -0800
Subject: Really, Really Confused...
Message-ID: <NCBBLGDECCBLJILGIFMPIECPCJAA.ben@semio.com>

Hi Folks,

I've encountered a very strange problem that'll take a little space to
explain.  I hope some of you will brave your way to the end though because
I'm well and truly stumped by this one.

First some history:  I got started on the road to confusion because I was
trying to turn my Python class into a COM server, though the issue I
ultimately ran into has nothing to do with COM.  It looks like the VB app I
was using to test the COM server i holding a long running reference to the
COM object, and then calling __init__ to initialize it instead of recreating
it from scratch when I  "create" it.  I don't know enough VB/COM to know why
this would be the case, but whatever - the init function was being invoked
repeatedly and it was causing problems.  I decided to remove some complexity
and just reinvoke the initializer a second time from the basic script to
debug.  I have since been able to reduce the issue down to the following
snippet of code:

class cApp:
    object = None
    def __init__(self):
        self.object = None
        self.object = cObject()

class cObject:
    myDict = {}
    def __init__(self):
        if self.myDict.has_key("foo"):
            raise "Already set"
        self.myDict["foo"] = "some value"

testApp = cApp()
testApp.__init__()
print "Finished"


When the init function runs the second time, the "Already set" exception is
raised.  Now I really don't understand why this would be.  When
cApp.__init__() is invoked, it releases its reference to self.object before
creating a new one (self.object = None).  As best I'm aware, it then creates
a completely new cObject with no common references to the old one.  In what
conceivable way could the myDict of the new cObject have a value already set
for "foo"?

And now it gets more interesting.  If I substitute a string in place of a
dict as the value in cObject, everything works fine:

class cApp:
    object = None
    def __init__(self):
        self.object = None
        self.object = cObject()

class cObject:
    myString = None
    def __init__(self):
        if self.myString != None:
            raise "Already set"
        self.myString = "some value"

testApp = cApp()
testApp.__init__()
print "Finished"

This runs to completion without griping.  HOW IS THIS DIFFERENT?

I think the real root of my confusion is that I don't understand what the
constructor is doing here.  Why is it that when I create a seemingly new
cObject, it has any value at all stored within it?


Thanks for reading all this, and thanks in advance for your help.


Best,

-Ben




From jessw at loop.com  Tue Dec 19 17:48:35 2000
From: jessw at loop.com (Jesse W)
Date: Tue, 19 Dec 2000 14:48:35 -0800
Subject: MIDI extensions (WIN32, Mac)?
References: <3A3FACEF.DAF77229@cableinet.co.uk>
Message-ID: <3A3FE5C3.76B4@loop.com>

Dear Richard,
	I am not sure if this is what you are talking about, but I have written
a basic class to print out all the events in a MIDI file; I hope to make
it into a general method to access and modify MIDI files from Python. 
If this might be any help to you please e-mail me.
						Jesse Weinstein

Richard Dobson wrote:
> 
> Hello,
> 
> I have just started looking at Python seriously, with a focus on music
> applications; e.g. algorithmic composition. Python seems to me to be
> idally suited as a first programming language for Music Technology
> students who have never done programming before. So far the net trawling
> I have done has failed to turn up any already-done extensions supporting
> MIDI i/o. I could probably code this myself at least for WIN32 if need
> be, but it is the sort of thing I feel sure somebody must already have
> done. Can anyone suggest anything? Or has it indeed not been done yet?
> 
> Richard Dobson


From fredrik at effbot.org  Tue Dec 12 11:26:28 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 12 Dec 2000 16:26:28 GMT
Subject: Inheriting the @ sign from Ruby
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu>
Message-ID: <UksZ5.3364$O%.525994@newsc.telia.net>

Roy Katz wrote:
> Why do we have to explain to new Python programmers how the
> label-value system Python uses is different than traditional call-by-
> reference and call-by-value?

Do we?  Of course there are people who don't understand that
different languages do things in different ways, but they're not
that common.

The real problem is that many Python tutorials fail to explain
how things work (or hide it somewhere in the exercises), but
that's not really Guido's fault...

> (I've been programming in Python for three years now and I
> understand only that passing an int or float is call-by-value,
> otherwise it is call-by-reference.

Which means that you don't understand how it works...

(Hint: Everything is a reference to an object.  Variables are
named references, not actual objects.  Some objects can be
modified in place.  Some cannot be modified.  That's all)

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From tim.one at home.com  Fri Dec 22 23:15:16 2000
From: tim.one at home.com (Tim Peters)
Date: Fri, 22 Dec 2000 23:15:16 -0500
Subject: A Suggestion for Python Colon Syntax
In-Reply-To: <slrn947q64.jkg.zeitlin@seth.lpthe.jussieu.fr>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEDDIFAA.tim.one@home.com>

[Vadim Zeitlin]
> Isn't it desirable (although not necessary) to attract more
> people to Python?

I think so, yes.

> IMveryHO any simplification of the syntax without any bad
> consequences (are there any in this case?) would be welcome from
> this point of view.

Oh, let's get real here:  the only syntax change that could possibly attract
a significant number of newcomers would be to add curly braces.  That's
because that's the only syntax change left that has a significant
constituency (the "optional trailing colon" thing comes up maybe two times
per year, and Bill is the first person to bother bringing it up twice
<wink>).

> BTW, what was the real rationale for adding composed assignment
> operators (+=) to Python 2.0 (which I'm really extremely glad to
> have now)?

1. According to Guido, it was the single most-requested feature in the
history of the language, and he never disapproved of it.  It took several
years for "the right way" to implement it to become apparent, else it would
have gone in much earlier.

2. The NumPy people had a real problem in that

    x = x + y

could end up generating multimegabyte temp arrays while

    x += y

need not.  That is, for them it wasn't "just another way" to write the
former; it offers crucially different semantics.

> Wasn't it meant to make Python more attractive to the people
> coming from C-lie languages?

And here I thought it was to make Python more attractive to Icon programmers
<wink>.

> And, if by chance it was so, why not continue making
> it even more simple to use to the very same people?

curly-braces-ly y'rs  - tim




From matt at virtualspectator.com  Fri Dec 22 17:59:50 2000
From: matt at virtualspectator.com (matt)
Date: Sat, 23 Dec 2000 11:59:50 +1300
Subject: A Suggestion for Python Dictionary/Class
In-Reply-To: <920fe9$c20$1@hecate.umd.edu>
References: <920fe9$c20$1@hecate.umd.edu>
Message-ID: <0012231202030B.00902@localhost.localdomain>

On Sat, 23 Dec 2000, William Djaja Tjokroaminata wrote:
> Hi all,
> 
> After using Python for a while, a have a wish for another type of
> dictionary.  If you ever used Tcl and Perl, you know that all Python, Tcl,
> and Perl basically provides three different fundamental data types:
> 
>     - scalar (Perl-speak) (which includes numbers and strings)
>     - list (or sequence in Python-speak)
>     - dictionary
> 
> But in Python, we have two different kinds of lists: lists and tuples.  I
> used Tcl and Perl before, so initially I was puzzled by why Python
> provides a 'read-only' list.  But it is fine.
> 
> I also use C and C++ a lot.  Their power and large-scale capabilities come
> from, among other things, the ability to write struct or class.  Among the
> three scripting languages, only Python provides the most natural form of
> struct/class like in C/C++/Java.  That's why I love Python so much; I can
> think in Python much like I think in C/C++/Java.
> 
> However, there is a fundamental difference with Python class.  The members
> are implemented using dictionary.  The danger with dictionary is, even
> though a key does not exist, we still can make an assignment to it.  When
> we make some typo, this can be hard to debug.  For example:
> 
>     class MyClass:
>         def __init__ (self):
>             self.member = 0
> 
>     myClass = MyClass ()
>     myClass.menber = 1    # This is a bug!
> 
>     print myClass.member  # Expected 0, but found 1



you mean Expected 1, but found 0

The only functionality I see you want is a spell checker of sorts?  maybe you
should include in your test units a print out of .__dict__ so you can see
easily that there are mistakes and that perhaps is the reason your test unit
failed.








> 
> In C/C++/Java, if a member does not exist and we try to access or assign
> it (because of some typo, perhaps), it will be caught by the compiler
> simply as syntax error.  Not so with Python class.
> 
> Therefore, is there any way for me to define a class, with some definite
> members just like in C/C++/Java, so that no new members can be defined?  I
> am trying to do it, but using some conventions, not using the Python
> language features.
> 
> Because Python provides a read-only list in terms of tuple, I think it
> will be nice if Python provides another class type with fixed members,
> just like a dictionary with read-only keys.  Any opinion, anyone?
> 
> Regards,
> 
> Bill
> -- 
> http://www.python.org/mailman/listinfo/python-list
-- 
Matt Halstead (PhD)
Research and development
VirtualSpectator
http://www.virtualspectator.com
ph 64-9-9136896



From ldw at us.ibm.com  Mon Dec  4 14:48:18 2000
From: ldw at us.ibm.com (Larry Whitley)
Date: Mon, 4 Dec 2000 13:48:18 -0600
Subject: Graphs from python
Message-ID: <90gse2$14uo$1@news.rchland.ibm.com>

The simulation models I write and run generate a lot of data and to make
sense of it all I put the data into a set of graphs.  The graphs are the
typical spread sheet fare, bar graphs, X/Y graphs, scatter plots, etc.
Today, I read the data into spreadsheets from the spreadsheet and create my
graphs by hand and I'd like to automate the process.

I'm looking at using win32com (by Mark Hammond) to drive excel or
piddle+graphite to generage graphs directly from python.  While I'm sure it
will get the job done, the excel object model is a bit overwhelming.  Piddle
by itself isn't at the right level.  Graphite appears get to the right level
but the pre-alpha rating makes me think that it's a bit early to jump in.

Are there other approachs I should consider?  Whatever I end up with, I'd
like it to work from either Unix or Windows.  If I can't have both, I can
make do with just windows.  Of the two approaches above, is one
significantly better than the other (by whatever measure you want to
choose)?

Thanks,
Larry




From tim.one at home.com  Sat Dec 23 01:45:59 2000
From: tim.one at home.com (Tim Peters)
Date: Sat, 23 Dec 2000 01:45:59 -0500
Subject: Good introduction to functional programming with Python?
In-Reply-To: <91v4m001una@news2.newsguy.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEDIIFAA.tim.one@home.com>

[Alex Maretlli]
> Haskell is my favourite FP language, and it does have some nice
> surface similarities with Python (use of whitespace, the list
> comprehension idea), but it differs deeply from Python in two
> key issues (which are NOT really central to FP...):
>
>     a. lazy evaluation is always the default semantics
>     b. very strict (and elegant) static typing
>
> Python evaluates eagerly by default (like almost all languages!),
> and its typing is always dynamic.
>
> If one wants to broaden one's mind, looking at the world from a
> very different angle than Python's, then Haskell makes a LOT of
> sense.  A 'triangulation' is best done by choosing two 'base
> points' that are not TOO close to each other, and nobody would
> accuse Haskell and Python of being too close, given [a] & [b].

I agree it depends on what the original questioner wants to get out of this.
That I don't know.  Haskell is a pure functional language with a wealth of
supporting material and easy-to-get, easy-to-use free ports, so it's a good
choice if FP is really what he wants.

If it is, then if FP is "about" anything, it's about the art of using
higher-order functions, and I believe mere mortals *need* a strong static
typing system to keep those straight.  Maps returning maps returning maps
... can get hard to follow real fast.

And I think anyone setting out to learn FP without a lazy language is
cheating themself out of a world of beauty:  learning to sling unbounded
streams in a *natural* setting is simply a joy.

If the guy just wants to putz around, though, I'd steer him to Scheme
<wink>.

> Incidentally, while haskell.org is a great site, I would strongly
> suggest that books are by far the best way to learn Haskell.

I imagine that depends too on what he wants out of this.  He should start
with the "Gentle Introduction to Haskell" at

    http://www.haskell.org/tutorial/

and take it from there.

> Hudak's "Haskell School of Expression" (Cambridge University Press)
> may well wow you if you're into multimedia & such; Thompson's more
> traditional "Craft of Functional Programming" 2nd ed (Addison
> Wesley) is workmanlike, thorough, and clear.

The last is what I recommend to most people:  not too abstract, and very
careful to explicitly declare the types of all identifiers <wink>.

> ...
> On the other hand, if one's focus is on understanding how best
> to apply FP *in a Python setting*, his FP curriculum may prove
> too round-about -- ...

FP features in Python were never intended to be more than a minor
convenience, and by all signs will never become more than that.  So I steer
him away from that.  Don't forget that the strongest argument in favor of
listcomps (in Guido's eyes) was that they gave a path to eventually throwing
"map" and "filter" away!

> ...
> A NON-pure language one SHOULD consider for its didactical strength
> is Scheme.

Indeed closer to Python in many ways than is Haskell.

> Abelson's and Sussman's "Structure and Interpretation of Computer
> Programs" will teach one Scheme in a strongly FP-oriented setting,
> and it's a great book on its own -- ...

Yup!  Pne of the few true programming classics.

> http://www.cs.rice.edu/CS/PLT/packages/drscheme/ gives a free
> and very good implementation, with all the trimmings and a lot
> of equally-free didactical support, good for any Win32 box, Mac,
> or Unixoid.  (Not that Haskell's and Erlang's implementations
> are any less free, cross-platform, and good-quality, but, IMHO,
> Rice's Scheme has characteristics that make it even better).

A caution that this depends on how powerful his computer is:  DrScheme is
too much a resource hog for a satisfying experience on older Windows boxes
(voice of experience).  This is a bit ironic, since Scheme is ported at
least as often as Python to tiny platforms.  You pay through the nose for
"all the trimmings" that come with DrScheme.  Worth it if you can afford it!
Else he can pick up implementations than run fine even on palmtops; e.g.,
from the links at (sorry, but my mailer will break this line):

http://www.dmoz.org/Computers/Programming/Languages/Lisp/Scheme/Implementati
ons/

betting-he's-got-more-advice-than-he-can-use-now-ly y'rs  - tim




From max at alcyone.com  Tue Dec 12 11:56:51 2000
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 12 Dec 2000 08:56:51 -0800
Subject: Inheriting the @ sign from Ruby
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu>
Message-ID: <3A3658D3.4AAA9AB0@alcyone.com>

Roy Katz wrote:

> I disagree; Like I mentioned above, this would enforce
> standardization.

It wouldn't enforce anything, since programmers wouldn't be mandated to
use it.  Programmers who didn't want to use the @ shorthand could still
call the `self' argument anything they liked.

To be honest I have seen very little variation in the name of the `self'
argument, and I doubt it would be a huge problem to decipher someone
else's code who used, say, `this' instead.  I don't think it's a serious
problem.  I don't think the @ shorthand is a bad idea, but at the same
time all it would do is make Python code more obfuscated, start
arguments about whether it was a good idea in the first place, and
unless mandated (which would _really_ piss people off) wouldn't
accomplish the goals it was set out to do.

> Python's language is not crystal-clear, imho.

I don't see how making it more obfuscated would help that.

> If we're so tough on
> perl for having a kludgy interface, why do we have {} for
> defining dictionaries instead of [ key:value ]?

Makes perfect sense, actually -- you have tuples, lists, and
dictionaries, and each are set off with different punctuation to make it
clear to everyone (including the interpreter) what type they are.

> (I've been programming in Python for three years now
> and I understand only that passing an int or float is call-by-value,
> otherwise it is call-by-reference.  At the risk of naivete, I would
> like a simple & referencing operator so that I can explicitly,
> finally, have control over this.)

This is actually something I wish had been done explicitly as well; this
is one of the only areas of Python that I found truly irregular and
strange.  It would have been nice to either have pass-by-value be the
default and have an explicit operator for referencing, or vice versa
(doesn't matter to me which one, as long as it's consistent;
pass-by-value as the default makes a little more sense).  Alas, it is a
little too late for that.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ If I had another face, do you think I'd wear this one?
\__/ Abraham Lincoln
    Official Omega page / http://www.alcyone.com/max/projects/omega/
 The official distribution page for the popular Roguelike, Omega.


From serdars at ixir.com  Mon Dec 25 15:44:34 2000
From: serdars at ixir.com (Serdar Soydemir)
Date: Mon, 25 Dec 2000 22:44:34 +0200
Subject: Ynt: html filtering
References: <3A46900D.BD118A0A@wielun.dhs.org>
Message-ID: <005a01c06eb3$8716e5c0$60452bd5@ss>

Hello,

Using translate method and maketrans method in string module may be a good
idea. I didn't try them but they seem more practical than replace..


----- Original Message -----
From: QdlatY <qdlaty at wielun.dhs.org>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Monday, December 25, 2000 2:08 AM
Subject: html filtering


> Hello!
>
> Big problem in writeing message boards is filtering html tags. I use
> very simple method: Each occurence of "<" sign or ">" sign I replace
> with &lt; and &gt;. Is this safe method? Maybe You've got better ideas?
>
> Yours
> QdlatY
> --
> http://www.python.org/mailman/listinfo/python-list




From SBrunning at trisystems.co.uk  Tue Dec 12 11:59:32 2000
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Tue, 12 Dec 2000 16:59:32 -0000
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
Message-ID: <31575A892FF6D1118F5800600846864D5B15D6@intrepid>

> From:	Roy Katz [SMTP:katz at Glue.umd.edu]
> Okay! tell me, which are the objects which cannot be modified in place? I
> assume these are int's, long's, float's, etc.  Strings, lists, objects and
> dictionaries are call-by-reference.  Pardon my choice of diction, but I am
> against inventing new terminiologies (modified-in-place??). 
 
The correct terminology is mutable (for objects which can be modified in
place), and immutable (for objects which cannot). See
<http://www.python.org/doc/current/ref/objects.html>.
 
> I want to know where this is documented.  Specifically, where is the
> dichotomy of ints/longs/floats vs. objects/strings/lists/dictionaries. 
 
The built-in types are documented here -
<http://www.python.org/doc/current/ref/types.html> - including whether or
not they are mutable.
 
> While I'm brainstorming with Python grammar changes, what about a real &
> operator.  No more of this hokey is-it-modifiable-in-place-or-not
> business.  Explicit control is important. 
 
Ugh.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From tbryan at python.net  Tue Dec 26 01:46:40 2000
From: tbryan at python.net (Thomas A. Bryan)
Date: Tue, 26 Dec 2000 01:46:40 -0500
Subject: html filtering
References: <3A46900D.BD118A0A@wielun.dhs.org> <mailman.977727367.19724.python-list@python.org> <3A478E1E.97BD8B67@wielun.dhs.org>
Message-ID: <3A483ED0.1189DDBA@python.net>

QdlatY wrote:
> 
> Hello!
> 
> Moshe Zadka wrote:
> >
> > On Mon, 25 Dec 2000, QdlatY <qdlaty at wielun.dhs.org> wrote:
> >
> > > Big problem in writeing message boards is filtering html tags. I use
> > > very simple method: Each occurence of "<" sign or ">" sign I replace
> > > with &lt; and &gt;. Is this safe method? Maybe You've got better ideas?
> >
> > Well, it is safe as long as you remember to replace every "&" with
> > "&amp;" *before* you replace < and >.
> 
> Yes! Thanks. I forget about ampersands;)).

Then again, you could use cgi.escape and save yourself the trouble. :)

always-check-your-library-before-you-write-a-book-ly yours
---Tom


From ommadawn at club-internet.fr  Wed Dec 13 10:56:13 2000
From: ommadawn at club-internet.fr (Christophe Prevost)
Date: Wed, 13 Dec 2000 15:56:13 -0000
Subject: Probleme with locale
References: <917pjb$ii2$1@reader1.imaginet.fr> <917sre02fh5@news2.newsguy.com>
Message-ID: <9182lh$kt9$1@reader1.imaginet.fr>

"Alex Martelli" <aleaxit at yahoo.com> a ?crit dans le message news:
917sre02fh5 at news2.newsguy.com...

> > this doesn't work ->
> > >>>locale.setlocale(locale.LC_ALL, "fr_FR")
 > Because the string 'fr_FR' is not a valid
> locale descriptor.
> >>> locale.setlocale(locale.LC_ALL, "French_France.1252")

Thanks!




From jj at void.si  Thu Dec 28 04:59:02 2000
From: jj at void.si (jj at void.si)
Date: Thu, 28 Dec 2000 09:59:02 GMT
Subject: [Pythonwin] Pywin encoding (again)
References: <0sNX5.168$M47.3490@m2newsread.uni2.es> <3A301090.3060304@ActiveState.com>
Message-ID: <92f2t6$vji$1@nnrp1.deja.com>

Is this issue fixed, because I cannot edit scripts in cp1250 using
pywin, but it works on PyWin Interactive Prompt and in SciTE 1.34
(code.page=1250; character.set=238)

for example this code entered in pywin iprompt creates a file 5 bytes
long, but if its run as a script saved with pywin it
creates a file 10 bytes long.

f=open('/tmp/a.txt', 'w')
f.write('?????')
f.close()

is there some workaround to write cp1250 scripts with pywin?

JJ



In article <3A301090.3060304 at ActiveState.com>,
  Mark Hammond <MarkH at ActiveState.com> wrote:
> Arturo P?rez Mulas wrote:
>
> >     - Is it possible to write Python scripts under Pythonwin using
> > non-ascii characters (latin-1 or cp1252 in my case) in comments,
> > docstrings, etc?
>
> Not with any of the relased win32all versions :-(
>
> There have been a few patches posted here just after 135 was
released,
> and by all accounts they work fine.
>
> So the next win32all/ActivePython should have it fixed.
>
> Mark.
>
>


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


From davidw at prosa.it  Thu Dec 28 10:45:32 2000
From: davidw at prosa.it (David N. Welton)
Date: 28 Dec 2000 16:45:32 +0100
Subject: Python vs PHP (& Java?)
References: <mailman.977897649.4877.python-list@python.org> <92cdn00hck@news2.newsguy.com> <87k88mxcf4.fsf@eugene.prosa.it> <D722E432FF377E01.657BAF2ECAA117A3.909A66D671189EE3@lp.airnews.net> <vIv26.94619$yR4.2590443@news1.rdc1.tx.home.com>
Message-ID: <878zp0r1lv.fsf@eugene.prosa.it>

"Scott Hathaway" <slhath at home.com> writes:

> Also, I have fallen in love with python, but I would not use it for
> server-side web scripting.  PHP is very easy and extremely powerful
> for this task.  Why not use both?

Because you are using Tcl?;-)

-- 
David N. Welton
     Personal:           http://www.efn.org/~davidw/  
Free Software:           http://people.debian.org/~davidw/
   Apache Tcl:           http://tcl.apache.org


From dnew at san.rr.com  Thu Dec  7 14:56:39 2000
From: dnew at san.rr.com (Darren New)
Date: Thu, 07 Dec 2000 19:56:39 GMT
Subject: Attention NYC Shoppers & Visitors - Beware of Rip-off SY Stores!!! ...... kHP2pr9I
References: <66FF28553F3.hlfeano@wmwnkhkpvfv.gov> <3a2f1968_4@corp.newsfeeds.com> <3a2f30aa$1@news.iprimus.com.au> <3a2fdb88_4@corp.newsfeeds.com>
Message-ID: <3A2FEB78.D263FEF9@san.rr.com>

Joshua Muskovitz wrote:
> Well, duh.  I was sole interested in the cipher at the bottom.  It doesn't
> look like a simple substitution cipher to me, and so I was asking because
> maybe someone else saw an easy way to decipher it.

I think it is added by the spam tool to make all the messages seem unique so
they don't get spam-busted in any automated kind of way.

-- 
Darren New / Senior MTS & Free Radical / Invisible Worlds Inc.
San Diego, CA, USA (PST).  Cryptokeys on demand.
Personal malapropism generator free with purchase!
Steganography: The manual is hidden in the source code.


From tim.one at home.com  Wed Dec 27 03:30:36 2000
From: tim.one at home.com (Tim Peters)
Date: Wed, 27 Dec 2000 03:30:36 -0500
Subject: One Python 2.1 idea
In-Reply-To: <037601c06f89$43d9e510$6401a8c0@home>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEJAIFAA.tim.one@home.com>

[Tim]
> Note:
>
> >>> class Oops:
>         def foo(self):
>             pass
>
> >>> x = Oops()
> >>> x.foo = x.foo
> >>> x.foo.im_self is x
> 1
> >>>
>
> Oops!  That is, x is now involved in a reference cycle (x's
> dict contains a bound method that points back to x).  Before 2.0,
> that's a memory leak.  In 2.0, it's a leak only if the class
> (unlike Oops above) has a __del__ method (or you've disabled
> cyclic gc).

[Darrell Gallion]
> You lost me here Tim.
> ...
> Don't see what's changed after a.f=a.f

The right-hand side a.f found Oops.__dict__['f'], and made a bound method
object out of it and a.  The left-hand side a.f makes that bound method the
value of a.__dict__['f'] (not in Oops's dict again, but in a's dict!).  So
what changed is that a.__dict__.has_key('f') was false before "a.f = a.f",
but is true after.  And a.__dict__['f'].im_self is a.

Why you changed all the names from the orginal example isn't clear or
helpful, so I'll add a few lines to the original instead:

>>> class Oops:
        def foo(self): pass

>>> x = Oops()
>>> x.__dict__ # the instance dict is empty at the start
{}
>>> x.foo = x.foo  # but this puts something into it
>>> x.__dict__
{'foo': <method Oops.foo of Oops instance at 00B24FDC>}
>>> x.__dict__['foo'].im_self is x  # this is the cycle
1
>>> Oops.__dict__['foo']  # in Oops's dict, it's just a function
<function foo at 00B1E6FC>
>>> Oops.foo  # but accessing a class dict via attr notation does magic
<unbound method Oops.foo>
>>>

This is pretty deep magic!  Don't go doing things like "a.f=a.f", and you
can safely ignore it for decades.

and-writing-"x=x"-can-erase-your-hard-drive<wink>-ly y'rs  - tim




From jbalding at hotmail.com  Fri Dec 15 14:57:34 2000
From: jbalding at hotmail.com (jbalding at hotmail.com)
Date: Fri, 15 Dec 2000 19:57:34 GMT
Subject: Is Stackless and uthread.py SMP friendly?
References: <91br9v$qa4$1@nnrp1.deja.com> <ku1yva5ocq.fsf@lasipalatsi.fi> <91dp4j$5fj$1@news.gte.com>
Message-ID: <91dt39$fbk$1@nnrp1.deja.com>

> >| will Stackless and uthread.py make my SMP
> >| experience with Python a happy one?
> >
> >it should work just fine, since microthreads don't use os threads.
>
>    True, but I believe the real question was whether microthreads
> could run on different CPUs, to which I believe the answer is "no".
> It will work fine, but only on one CPU per process.

Okay. My next question: Is there a simple way of taking advantage of
multiple processors using Python, or is this something that is going to
be like pulling teeth? In other words, are there existing packages that
allow relatively high-level, transparent access to multiple processors?

Thanks - John


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


From per at sbc.su.se  Wed Dec 13 04:49:48 2000
From: per at sbc.su.se (Per Kraulis)
Date: Wed, 13 Dec 2000 10:49:48 +0100
Subject: [ANN] crng 1.0: Random-number generators (RNGs) as Python extension
 	types coded in C.
Message-ID: <mailman.976723861.27715.clpa-moderators@python.org>

crng 1.0: Random-number generators (RNGs) implemented as Python
extension types coded in C.

The Python module crng implements random-number generators (RNGs)
based on several different algorithms producing uniform deviates in
the open interval (0,1), i.e. exclusive of the end-point values 0 and
1. A few continuous and integer-valued non-uniform deviates are also
available. Each RNG algorithm is implemented as a separate Python
extension type. The RNG types are independent of each other, but have
very similar interfaces. The entire module is implemented as one
single C source code file.

Official website:  http://www.sbc.su.se/~per/crng/

Documentation:  http://www.sbc.su.se/~per/crng/doc.html

The GNU General Public License (GPL) applies.

[Please note that the crng module has been tested under Python 1.5.2
only. I'd be grateful if someone could check it under Python 2.0.]

-- 
Per J. Kraulis, Ph.D.                        per at sbc.su.se
Stockholm Bioinformatics Center (SBC)        http://www.sbc.su.se/~per
Dept. Biochemistry, Stockholm University     phone +46 (0)8 - 674 78 17
SE-106 91 Stockholm, SWEDEN                  fax   +46 (0)8 -  15 80 57



From aleaxit at yahoo.com  Fri Dec 15 07:32:56 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 15 Dec 2000 13:32:56 +0100
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <tcmay-FCA22F.14073713122000@news.got.net> <918v7a02scn@news1.newsguy.com> <tcmay-04E80C.16181313122000@news.got.net> <91a1ed02a4j@news2.newsguy.com> <3A389B1A.4F33FFD4@cfmu.eurocontrol.be> <91ae8s02llf@news2.newsguy.com> <3A397778.2E6D84D3@cosc.canterbury.ac.nz> <91c7r4$7a0$1@slb1.atl.mindspring.net>
Message-ID: <91d32r0kgt@news1.newsguy.com>

"Andrew Dalke" <dalke at acm.org> wrote in message
news:91c7r4$7a0$1 at slb1.atl.mindspring.net...
    [snip]
> One packet of syntactic sugar, coming up!
>
> def make_a(*classes):
>   class make_a:
>     pass
>   make_a.__bases__ = tuple(classes)
>   return make_a()

or, in a similar vein:

def make_a(*classes, **attribs):
    return new.classobj('blob', classes, attribs)()


Alex





From jepler at inetnebr.com  Wed Dec 13 18:57:45 2000
From: jepler at inetnebr.com (Jeff Epler)
Date: Wed, 13 Dec 2000 23:57:45 GMT
Subject: xreadlines (was Re: while true: !!!)
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com>
Message-ID: <slrn93g37o.9tj.jepler@potty.housenet>

It seems that the most common place that people write the 'while 1:
... if test: berak' idiom is when iterating over lines from a file
where the file is not known to fit in memory.

In Python, there is an existing precedent for creating a special 'iteratable'
object where the entire object would not fit in memory, the 'xrange' function
and its associated object type.

I take this cue to suggest the idea of an 'xreadlines' object.  I have
implemented 'xreadlines' in Python (where it was slower than all other
solutions, probably due to method call overhead) and in C (where it
holds its own with the fastest idiom I know,
        file = open(testfile)
        while 1:
                lines = file.readlines(1024)
                if not lines: break
                for line in lines:
                        pass
)

Attached is a copy of xreadlines.c and test.py (the latter contains the
XReadlines class).. test.py benchmarks 5 methods of reading the file's
contents.  Pick a long text file on your system, and run the test to find the
relative speeds of each method.

On a 433 MHz machine on the 45k-line, 400kbyte file /usr/dict/words:
% ./python test.py 
0.390071988106
1.04547798634
0.354212999344
0.365339994431
1.616106987

Jeff Epler
jepler at inetnebr.com

# ---- cut here for test.py
import xreadlines, time

class XReadlines:
	def __init__(self, file):
		self.lines = []
		self.file = file
		self.index = 0

	def __getitem__(self, i):
		if self.index >= len(self.lines):
			self.lines = self.file.readlines(512 * 1024)
			self.index = 0
		ret = self.lines[self.index]
		self.index = self.index + 1
		return ret

def timer(func, *args, **kw):
	t0 = time.time()
	func(*args, **kw)
	t1 = time.time()
	print t1-t0

testfile = "/usr/dict/words"
def test1():
	for line in open(testfile).readlines():
		pass

def test2():
	file = open(testfile)
	while 1:
		if file.readline() == "": break

def test3():
	file = open(testfile)
	while 1:
		lines = file.readlines(1024)
		if not lines: break
		for line in lines:
			pass

def test4():
	for line in xreadlines.xreadlines(open(testfile)):
		pass

def test5():
	for line in XReadlines(open(testfile)):
		pass

timer(test1)
timer(test2)
timer(test3)
timer(test4)
timer(test5)

# ---- test.py ends here

/* cut here for xreadlinesmodule.c */

#include "Python.h"

typedef struct {
	PyObject_HEAD
	PyObject *file;
	PyObject *lines;
	int lineslen;
	int lineno;
} PyXReadlinesObject;

staticforward PyTypeObject XReadlinesObject_Type;

static void
xreadlines_dealloc(PyXReadlinesObject *op) {
	Py_TRASHCAN_SAFE_BEGIN(op)
	PyObject_GC_Fini(op);
	Py_XDECREF(op->file);
	Py_XDECREF(op->lines);
	PyObject_DEL(op);
	Py_TRASHCAN_SAFE_END(op)
}

#if SIZEOF_INT < 4
#define BIGCHUNK  (512 * 32)
#else
#define BIGCHUNK  (512 * 1024)
#endif

static PyXReadlinesObject *
newreadlinesobject(PyObject *file) {
	PyXReadlinesObject *op;
	op = PyObject_NEW(PyXReadlinesObject, &XReadlinesObject_Type);
	if(op == NULL) return NULL;
	Py_XINCREF(file);
	op->file = file;
	op->lines = NULL;
	op->lineno = op->lineslen = 0;
	return op;
}

static PyObject *
xreadlines(PyObject *self, PyObject *args) {
	PyObject *file;
	PyXReadlinesObject *ret;

	if (!PyArg_ParseTuple(args, "O:xreadlines", &file))
		return NULL;
	ret = newreadlinesobject(file);
	Py_INCREF(ret);
	return (PyObject*)ret;
}

static PyObject*
xreadlines_item(PyXReadlinesObject *a, int i) {
	PyObject *ret;
	if(a->lineno >= a->lineslen) {
		Py_XDECREF(a->lines);
		a->lines = PyObject_CallMethod(a->file, "readlines", "(i)", BIGCHUNK);
/*
		if (a->lineslen == 0) {
			static PyObject *indexerr = NULL;
			if (indexerr == NULL)
				indexerr = PyString_FromString(
					"end of file");
			PyErr_SetObject(PyExc_IndexError, indexerr)
			return NULL;
		}
*/
		a->lineno = 0;
		a->lineslen = PyList_Size(a->lines);
	}
	ret = PyList_GetItem(a->lines, a->lineno++);
	Py_XINCREF(ret);
	return ret;
}

static PySequenceMethods xreadlines_as_sequence = {
	0, /*sq_length*/
	0, /*sq_concat*/
	0, /*sq_repeat*/
	(intargfunc)xreadlines_item, /*sq_item*/
};

static PyTypeObject XReadlinesObject_Type = {
	PyObject_HEAD_INIT(&PyType_Type)
	0,
	"xreadlines",
	sizeof(PyXReadlinesObject) + PyGC_HEAD_SIZE,
	0,
	(destructor)xreadlines_dealloc, /*tp_dealloc*/
	0, /*tp_print*/
	0, /*tp_getattr*/
	0, /*tp_setattr*/
	0, /*tp_compare*/
	0, /*tp_repr*/
	0, /*tp_as_number*/
	&xreadlines_as_sequence, /*tp_as_sequence*/
	0,		/*tp_as_mapping*/
	0,		/*tp_hash*/
	0,		/*tp_call*/
	0,		/*tp_str*/
	0,		/*tp_getattro*/
	0,		/*tp_setattro*/
	0,		/*tp_as_buffer*/
	Py_TPFLAGS_DEFAULT | Py_TPFLAGS_GC,	/*tp_flags*/
 	0,		/* tp_doc */
};

static PyMethodDef xreadlines_methods[] = {
	{"xreadlines", xreadlines, METH_VARARGS},
	{NULL, NULL}
};

void
initxreadlines(void)
{
	PyObject *m;

	m = Py_InitModule("xreadlines", xreadlines_methods);
}
/* xreadlinesmodule.c ends here */


From trotts at llnl.gov  Sun Dec 31 17:53:38 2000
From: trotts at llnl.gov (Issac Trotts)
Date: Sun, 31 Dec 2000 14:53:38 -0800
Subject: sip documentation
Message-ID: <20001231145338.A27483@llnl.gov>

Does anyone know where I can find some documentation on how to use sip?
I've looked all over and can't seem to find any.

Thanks,
Issac




From MarkH at ActiveState.com  Fri Dec 29 20:24:07 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sat, 30 Dec 2000 01:24:07 GMT
Subject: msxml3 and IStream interface
References: <QV_26.228$Io4.17256@news.siol.net>
Message-ID: <3A4D3814.6090201@ActiveState.com>

jj wrote:

> I want to use MSXML3 library for writing xml in file, but I dont know how to
> implement IStream interface in acceptable way. In following source is
> stripped Stream class from python examples. Too prove that all other code is
> working there is working example of saving in COM structured document file.
> I would also appreciate comments about my coding, especially how to get
> another interface pointer from object. I found only working solution using
> king of import typelib with makepy.py utility.

You are certainly on the right track.  Note that IStream is somewhat 
special, in that the win32com package knows about this explicitly, so 
there is no need for makepy support for this interface.  You _do_ need 
makepy support for the XML lib, as you have done.  (you probably knew 
all that - just being clear)

>  def Read(self, amount): return 0

Read will be expecing a string to come back.  This will be raising an 
"internal exception", which you may only see when using the special 
debugging features.

*sigh*.  Looking at the code, you wont even get a decent error message 
:-(  Probably "TypeError: len() of unsized object" rather than a more 
useful "must be a string"

>  def Write(self, data): return 0

This should be

def Write(self, data): return len(data)

for a workable stub implementation.


> 
> class SaxWriter: #wrapper around xmlwriter
>  def __init__(self):
>   self.elements = []
>   self.wr = win32com.client.Dispatch("Msxml2.MXXMLWriter")
>   # !!! reason for importing typelibrary, could be done without importing
> typelib !!!!
>   self.ch =msxml_TLB.IVBSAXContentHandler(self.wr)
>   self.atrs = win32com.client.Dispatch("Msxml2.SAXAttributes")

Dig up some recent posts from myself and Alex Martelli on "makepy" and 
"gencache", and you should find answers to cleaning this up.

> def test():
>  s = SaxWriter()
>  st = Stream()  # !!!! its not working

You need to "wrap" the object.

Try:

   st = win32com.server.util.wrap(Stream())

And to see any internal errors, you can try:

   useDispatcher = win32com.server.dispatcher.DefaultDebugDispatcher
   st = win32com.server.util.wrap(Stream(), useDispatcher=useDispatcher)

then open the "Trace collector debugging tool" from Pythonwin's tools 
menu to see debug output.

For more info, you've really gotta buy my and Andy's book, or trawl deja.

Mark.



From josh at open.com  Thu Dec  7 13:55:49 2000
From: josh at open.com (Joshua Muskovitz)
Date: Thu, 7 Dec 2000 13:55:49 -0500
Subject: Attention NYC Shoppers & Visitors - Beware of Rip-off SY Stores!!! ...... kHP2pr9I
References: <66FF28553F3.hlfeano@wmwnkhkpvfv.gov> <3a2f1968_4@corp.newsfeeds.com> <3a2f30aa$1@news.iprimus.com.au>
Message-ID: <3a2fdb88_4@corp.newsfeeds.com>

> The whole thing seems like your common, garden variety racial vilification
> to me.

Well, duh.  I was sole interested in the cipher at the bottom.  It doesn't
look like a simple substitution cipher to me, and so I was asking because
maybe someone else saw an easy way to decipher it.

-- j




-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----


From james at daa.com.au  Mon Dec  4 19:48:30 2000
From: james at daa.com.au (James Henstridge)
Date: Tue, 5 Dec 2000 08:48:30 +0800 (WST)
Subject: [pygtk] Python and Orbit Examples
In-Reply-To: <200012040748.eB47mEh22820@ZOOstation.cc>
Message-ID: <Pine.LNX.4.21.0012050828550.24722-100000@tupelo.daa.com.au>

On 3 Dec 2000, Rob Brown-Bayliss wrote:

> Hello.
> 
> Has anyone got some simple, usefull examples for a beginer to Python and
> Corba in how to make two python apps on two different machines say hello
> to each other, preferably with examples that works with pyGnome

There are a few examples included in the orbit-python package (don't
bother looking at PyORBit), which you could use as a starting point.  As
for getting it to work with gnome-python, that is a bit more difficult.

I haven't written wrappers for either libgnorba or oaf.  One of these is
needed to allow ORBit to run in the same event loop as the GUI part of
your program.

James.




From gregj at pobox.com  Sat Dec 16 02:16:12 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Sat, 16 Dec 2000 07:16:12 GMT
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <915jcf0c6o@news2.newsguy.com> <9162p3$7t2$1@animus.fel.iae.nl> <MBWZ5.334$h67.37748@newsread1.prod.itd.earthlink.net> <3A397C74.6C8906E1@cosc.canterbury.ac.nz> <91d5520m6j@news1.newsguy.com>
Message-ID: <0FE_5.153194$U46.4997180@news1.sttls1.wa.home.com>

I worked on a huge C/C++ mutant project a few years ago. One of the
programmers thought this was clever:

  // buried in some nested #include file
  #define FOREVER true
  ...
  while (FOREVER) { ... }

Not to be outdone, another programmer introduced this abomination:

  #define FOR_EVER for ( ; ; )
  ..
  FOR_EVER { ... }

Clearly a case of two programmers confusing "idiom" with "idiot." Of course
when I asked why on earth anyone would do this both programmers argued that
their way was more clear than the standard idiom (more clear to whom?). I
can't see how either of their macros is more clear than the standard C
idioms. I'm convinced that this kind of thing is just attention-getting and
showing off. Real programmers don't have any problem understanding while (1)
{ ... }.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From raymond at ibercom.com  Wed Dec 20 08:14:33 2000
From: raymond at ibercom.com (Raymond Biurrun)
Date: Wed, 20 Dec 2000 14:14:33 +0100
Subject: Python interpreter in python?
References: <mailman.977254923.336.python-list@python.org>
Message-ID: <91qas9$i10$1@talia.mad.ttd.net>

"Steven D. Majewski" <sdm7g at virginia.edu> escribi? en el mensaje
news:mailman.977254923.336.python-list at python.org...
>
> That is absolutely untrue!
>
> Lisp and Scheme are frequently implemented in themselves -- a
> 'metacircular interpreter' is a fairly typical student exercise.
>

Please excuse my ignorance but what's a "metacircular interpreter" and where
can I find info about them.

Regard -- Raymond




From nde at comp.leeds.ac.uk  Fri Dec 15 07:10:00 2000
From: nde at comp.leeds.ac.uk (Nick Efford)
Date: Fri, 15 Dec 2000 12:10:00 GMT
Subject: compiled or interpreted?
References: <w8jZ5.74702$IP1.2487256@news1.giganews.com> <usnnt3oj4.fsf@cs.uu.nl>
Message-ID: <G5Lz4o.EyA@leeds.ac.uk>

On 12 Dec 2000 11:46:39 +0100, piet at cs.uu.nl <piet at cs.uu.nl> wrote:

> And Python doesn't have a JIT, like Java.

Has the idea of developing a JIT for Python ever been considered?
Or is there some technical reason for it not being possible?

Personally, I find Python to be fast enough for most things, and
when it isn't a module can be written easily enough in C to do the
business - but a JIT would be interesting.


Nick



From duncan at rcp.co.uk  Wed Dec 20 09:45:30 2000
From: duncan at rcp.co.uk (Duncan Booth)
Date: Wed, 20 Dec 2000 14:45:30 +0000 (UTC)
Subject: Need Help sorting alphabetically.
References: <3A3F1A91.A7ED9182@araidesign.com>
Message-ID: <90109F230duncanrcpcouk@194.238.50.13>

chris at araidesign.com (Chris Arai) wrote in
<3A3F1A91.A7ED9182 at araidesign.com>: 

>Hi,
>
>This seems like it should be obvious, but it is evading me!
>
>I would like to sort alphabetically (I'm not sure that is the correct
>term) so that strings of alpha numerics sort alphabetically instead of
>by ASCII order.
>
>eg:
>>>> m=['Pg95_100.GIF',
>'Pg9_100.GIF','Pg95A_100.gif','Pg9A_100.gif','Pg10_100.gif']
>>>> m.sort()
>>>> m
>['Pg10_100.gif', 'Pg95A_100.gif', 'Pg95_100.GIF', 'Pg9A_100.gif',
>'Pg9_100.GIF']
>
>
>I would like the order to be
>['Pg9_100.GIF', 'Pg9A_100.gif', 'Pg10_100.gif, 'Pg95_100.GIF',
>Pg95A_100.gif' ]
>
>What is the obvious answer??
>
Try this:
---- begin cut ---
import re
from string import lower

def make_key(value, pattern = re.compile('(\d+)|(\D+)')):
    vals = pattern.split(value)
    res = []
    for i in range(1, len(vals), 3):
        if vals[i]:
            res.append(int(vals[i]))
        else:
            res.append(lower(vals[i+1]))
    return res

def cleversort(list):
    l1, res = [], []
    for v in list: l1.append((make_key(v), v))
    l1.sort()
    for k, v in l1: res.append(v)
    return res

m=['Pg95_100.GIF',
'Pg9_100.GIF','Pg95A_100.gif','Pg9A_100.gif','Pg10_100.gif'] 

print cleversort(m)

----- end cut ----
The output is:
['Pg9_100.GIF', 'Pg9A_100.gif', 'Pg10_100.gif', 'Pg95_100.GIF',
'Pg95A_100.gif'] as you requested.

This sorts each sequence of digits as its numeric value, and each
sequence of non digits as its lowercase equivalent (which I think is what
you were asking for). 


From aleaxit at yahoo.com  Sat Dec  9 18:12:20 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 10 Dec 2000 00:12:20 +0100
Subject: using execfile() in class/instance methods
References: <slrn932qna.rtk.wcuddy@lserv.ja10629.home> <90tb1302s27@news2.newsguy.com> <slrn934t6r.s9l.wcuddy@lserv.ja10629.home>
Message-ID: <90ueaf02n20@news2.newsguy.com>

"Wayne Cuddy" <wcuddy at lserv.ja10629.home> wrote in message
news:slrn934t6r.s9l.wcuddy at lserv.ja10629.home...
    [snip]
> Your Example Worked But I Want Multiple Instances Of This Class And Be
> Able To Pass The Filename As An Arg.

So you don't want a class-attribute but an instance-attribute, OK, then you
do need to pass the appropriate dictionary to execfile (no need for a
locals/globals distinction here):

class Ba:
    def __init__(self, filename):
        execfile('ba.txt', self.__dict__)
    def bu(self):
        print self.bn['key']

ba = Ba('ba.txt')

ba.bu()

OK so far?

> I Also Would Prefer That Bn Be A
> Local Variable Of The Function Containing The Execfile() Call.  This

The locals() dictionary is not guaranteed to be usable this way -- it
will surely work for reading, but modifications to it will not
necessarily be reflected in your local-variables (there's an important
optimization behind this restriction, so, suffer it...!).

Just pass any old dictionary and use it...!

> Would Allow Me To Either Return It Or Bind It To Any Variable Name Of
> The Instance.  So I Want To Do Something Like This:
>
> Class Bla:
>       ...
>       ...
>       Def F1(Self,Filename):
>   Execfile(Filename)
>   Self.Dictfromdisk = Bn
>
>       Def F2(Self):
>   Print Bn

Didn't you just say that you wanted bn to be LOCAL to the function
doing execfile?  If you got your wish, then here it would be local
to f1, thus completely unknown in f2.

The only way f2 can access bn just like that if it is a GLOBAL
variable, which you can obtain by using
    execfile(filename,globals())
in f1, but is a perfectly horrid idea IMHO.

Apart from this weirdness in f2, f1 can be implemented easily:

    def f1(self, filename):
        a={}
        execfile(filename, a)
        self.dictFromDisk = a['bn']

for example.


Alex





From grey at despair.rpglink.com  Sat Dec 30 15:28:30 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Sat, 30 Dec 2000 20:28:30 -0000
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <3A4D6A05.EEC97FA6@uswest.net> <slrn94qqvg.mgg.grey@teleute.rpglink.com> <ije36.96813$x6.43604873@news2.rdc2.tx.home.com> <slrn94qt4f.mgg.grey@teleute.rpglink.com> <CMo36.97857$x6.44474719@news2.rdc2.tx.home.com> <slrn94s9qg.134.grey@teleute.rpglink.com> <gdq36.2696$Qb7.302774@newsb.telia.net>
Message-ID: <slrn94shbe.150.grey@teleute.rpglink.com>

On Sat, 30 Dec 2000 18:56:12 GMT, Fredrik Lundh <fredrik at effbot.org> wrote:
>from what world do you come from where initializing variables to arbitrary
>values just to get your program flow right is a good thing?

    I have never heard of a single programming book or course that says, "Oh,
don't init variables."  In this case I feel you are squarely in the wrong
with, oh, the entire CS community.

>next you gonna tell us that it's better to use flags than to use break and/or
>exceptions...

    No, but I don't think that when you can place the test into the while that
one should use a while 1:break construct.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From katz at Glue.umd.edu  Tue Dec 12 01:44:16 2000
From: katz at Glue.umd.edu (Roy Katz)
Date: Tue, 12 Dec 2000 01:44:16 -0500
Subject: Inheriting the @ sign from Ruby
Message-ID: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu>

Hello, 



I was reading a Dr. Dobb's recently and it mentioned that Ruby uses the @
symbol to denote instance members of a class.  This appeals to me because
I can comprehend it faster than seeing 'self' pasted in front of a
name, especially when the statement is buried in dense code.  

Why not bind @ to the first argument of method definitions? Consider:


  class X:

    def __init__( self, x ):

      @bind(x)     # binds to 'self', because 
                   # 'self' is the first argument to __init__

    def bind(myself, x):
      @x = x   # equivalent to myself.x



What faults or strengths do you see with this?


Roey



From arhiv at freemail.org.mk  Wed Dec  6 08:19:40 2000
From: arhiv at freemail.org.mk (Damjan)
Date: Wed, 6 Dec 2000 14:19:40 +0100 (CET)
Subject: Creating Python objects out of a memory pool
Message-ID: <Pine.LNX.4.30.0012061415500.481-100000@localhost.localdomain>

Is it posible to create Python objects in a known space in the memory
(like in a shared memory).
I want some some objects created in python to be in the shared memory
pool, while others to be allocated as usual. Is there a way to hook in to
the Python memory allocation from Python itself or I'll haveto use the C
API?

Thanks.

-- 
Damjan Georgievski		|           ?????? ???????????
Skopje, Macedonia		|           ??????, ??????????




From kmbotha at netinfo.ubc.ca  Thu Dec 28 05:43:03 2000
From: kmbotha at netinfo.ubc.ca (Kris Botha)
Date: Thu, 28 Dec 2000 10:43:03 GMT
Subject: createfilehandler on Win32
Message-ID: <3a4b186a.40583796@news.interchange.ubc.ca>

Hi, 

I'm a newbie - I hope someone can point me in the right direction
here.

I'd like to have a GUI for a "client socket" displaying incoming
messages from a server. Now, I know this is done quite elegantly using
the createfilehandler method in Tkinter, but this method is not
available for win32 apps. 

My question  -  is there a simple way to accomplish this task (i.e.
program a client socket, 'waiting' for messages) without using the
createfilehandler method?


cheers



~~====----..----====~~
        Kris Botha



From edcjones at erols.com  Fri Dec  8 22:00:44 2000
From: edcjones at erols.com (Edward C. Jones)
Date: Fri, 08 Dec 2000 22:00:44 -0500
Subject: Python wrapper for C++ STL?
References: <3A31760E.D157EB54@erols.com> <C0gY5.126062$U46.4232859@news1.sttls1.wa.home.com> <90s3fk$j8p$1@nnrp1.deja.com>
Message-ID: <3A31A05C.6A853969@erols.com>

shindich at my-deja.com wrote:

> With all due respect, what would be wrong with a definition like
> std::map<PyObject*, PyObject*>? Obviously there is no type safety here.
> But Python is not type safe.

This seems correct. It would give me a "sorted associative container", a
dictionary I could loop through.

Ed Jones




From rob at ZOOstation.cc  Mon Dec  4 02:48:13 2000
From: rob at ZOOstation.cc (Rob Brown-Bayliss)
Date: 03 Dec 2000 19:48:13 -1200
Subject: Python and Orbit Examples
Message-ID: <200012040748.eB47mEh22820@ZOOstation.cc>

Hello.

Has anyone got some simple, usefull examples for a beginer to Python and
Corba in how to make two python apps on two different machines say hello
to each other, preferably with examples that works with pyGnome

Thanks

-- 

  Rob Brown-Bayliss
 ---======o======---
  www.ZOOstation.cc



From paul.magwene at yale.edu  Tue Dec 12 18:29:03 2000
From: paul.magwene at yale.edu (Paul Magwene)
Date: Tue, 12 Dec 2000 18:29:03 -0500
Subject: Simple question/problem
References: <916b28$9lf$1@nnrp1.deja.com>
Message-ID: <3A36B4BF.80323E6F@yale.edu>

chicago_george at my-deja.com wrote:
> 
> Hello, I'm new to Python.. When I try to import java classes, I get an
> error msg saying 'no module named java'
> For example;
> 
> >>> from java import awt
> ImportError: no module named Java
> 
> Why is this happening?  Anyone???
> 

I guess I'll ask the obvious - are you using Python (i.e. CPython) or
JPython/Jython?



-- 

Paul Magwene
paul.magwene at yale.edu


From tim.one at home.com  Tue Dec 26 21:56:07 2000
From: tim.one at home.com (Tim Peters)
Date: Wed, 27 Dec 2000 02:56:07 GMT
Subject: One Python 2.1 idea
References: <mailman.977771949.17911.python-list@python.org> <9292tn$jka$1@nnrp1.deja.com> <929n9c013m3@news1.newsguy.com> <92ab9u$dou$1@nnrp1.deja.com> <92ap0i22b6r@news1.newsguy.com> <slrn94ie9k.anf.neelk@alum.mit.edu>
Message-ID: <bTc26.210688$RP.38759664@news1.rdc1.md.home.com>

[Neelakantan Krishnaswami <neelk at alum.mit.edu
 news:slrn94ie9k.anf.neelk at alum.mit.edu]
> ...
> Fortunately, I don't have to put my money where my mouth is, since
> someone has already done it for me. :) Vladimir Marangozov has
> modified the Python interpreter (as of 1.5) to use the GNU computed
> labels extension and implemented direct threading.
>
>   http://starship.python.net/crew/vlad/archive/threaded_code/
>
> Bringing that up to date with 2.0 would be an interesting experiment.

The extreme simplicity of the patch suggests it's not doing much -- and that
suspicion is intensified due to Vladimir's utter lack of quantitative claims
for it <wink>.

The page Vlad references:

    http://www.complang.tuwien.ac.at/forth/threaded-code.html

identifies Python's current mechanism as "Switch Threading", and notes:

    The result has the same advantages as token threaded code; the
    disadvantage over token threaded code is the lower speed due to
    the range check generated by most (all?) C compilers for the switch

Other than that, the mechanisms appear identical by inspection:  both use
the opcode as an index into a table of addresses (in the current "switch"
case, that table is constructed by the compiler; in Vlad's case, by hand),
then jump to the address so loaded.  Don't know how lame gcc is <wink>, but
under MSVC for the Pentium the range check (to jump to the "default:" case
if the index would exceed the table size) amounts to a register move with
decrement, compare to immediate constant, and a non-taken jump.  If that is
indeed all the savings there is to be gotten via this technique, it's
marginal at best.

Note that at the 1998 Python Conference, John Aycock reported on an extreme
technique that can eliminate *all* eval-loop overhead, including instruction
decoding and dispatch costs; look for "Converting Python Virtual Machine
Code to C" at

    http://www.foretec.com/python/workshops/1998-11/proceedings.html

The results were inconclusive (detecting a pattern here <wink>?), but it
wasn't encouraging that pystone only got a 10% speed boost.  That did and
does seem anomalously small, but so long as "+" has to do pages of analysis
at runtime, the time to *find* the BINARY_ADD code really isn't important.
Threading techniques grew up in Forth, where the implementation of a typical
opcode consists of just a handful of machine instructions.

god-forbid-anyone-profile-the-code-and-find-out-where-it's-really-
    spending-time<wink>-ly y'rs  - tim





From erno-news at erno.iki.fi  Wed Dec 13 03:00:51 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 13 Dec 2000 10:00:51 +0200
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <kuzoi15d60.fsf@lasipalatsi.fi> <Pine.BSF.4.10.10012122310510.48551-100000@tcfreenet.org>
Message-ID: <kupuiw698s.fsf@lasipalatsi.fi>

In article <Pine.BSF.4.10.10012122310510.48551-100000 at tcfreenet.org>,
Aaron Jon Reichow <reichowa at tcfreenet.org> writes:

| On 13 Dec 2000, Erno Kuusela wrote:
|| the thing that struck me as a little foreign was the apparent
|| isolation of the smalltalk image from the rest of the operating system
|| and the filesystem. i want to be able to grep code!

| While I'm not a proponent of Smalltalk isolationism, grep is a useless
| tool (more or less), simply an artifact, ye, a symptom of the
| sickness known as file-based development.

point was that it's valuable to me to be able to manipulate stuff via
the filesystem. version control, editors, access control, locking,
backups, and all the rest of the universe works with files (same
critique goes for zope).

| According to some benchmarks I've seen, Squeak is faster. ;P  (although it
| doesn't feel like it sometimes, with Morphic)

python needs to steal some implementation tricks from it then :)

  -- erno


From jasonic at nomadicsltd.com  Sat Dec 30 13:44:35 2000
From: jasonic at nomadicsltd.com (Jason Cunliffe)
Date: Sat, 30 Dec 2000 13:44:35 -0500
Subject: how to sort a dictionary by the values ?
References: <Pine.GSO.4.10.10012292313350.21787-100000@ux11.cso.uiuc.edu>
Message-ID: <t4sb3drsgvt142@corp.supernews.com>

Depending upon your needs, check out Wolfgang Grafen's wonderful SEQDICT
modules:

http://home.germany.net/100-366919/Python/Modules/Modules.html

<quote>

seqdict - single value dictionary
msedict - multiple value dictionary

The Python built-in dictionary has no reliable sequence order because of
performance reasons.

The two new dictionaries introduced here keep their entries in sequence.

* seqdict keeps one value for one key
* mseqdict keeps multiple values per key and is the closest emulation of
areal world dictionary.

As the sequential dictionary is a combination of a list and a dictionary you
can do most operations defined with lists as well and much more.
Now you can slice, add, sort with dictionaries. The functional operations
map, filter, reduce are implemented as well. Use a dictionary like a stack
with push and pop. Split it, reverse it and swap keys and values.
A small tutorial is provided as HTML and Postscript Source and is subject to
become improved some other day.
I think both dictionaries are candidates for the python standard library.
Use the seqdicts to emulate a printed dictionary. Use them for workinse them
for working with csv-tables. Keep parsed code in a seqdict, modify it and
write it back ...
</quote>

have fun..
- Jason
___________________________________________________________
Jason CUNLIFFE = NOMADICS['Interactive Art and Technology']
hwan-jo yu <hwanjoyu at students.uiuc.edu> wrote in message
news:Pine.GSO.4.10.10012292313350.21787-100000 at ux11.cso.uiuc.edu...
> Hi,
> If I sort the items() of dictionary, it is sorted by first the key and
> second the value.
> How can I sort it by the values ?
> Thanks in advance.
>
>




From lale at fotonation.com  Wed Dec  6 11:34:56 2000
From: lale at fotonation.com (Ilariu Raducan)
Date: Wed, 6 Dec 2000 16:34:56 -0000
Subject: CGI POST binary file upload?
References: <90j8j0$cmt$1@nnrp1.deja.com> <p1dX5.3468$Er5.2143@news.indigo.ie> <zodX5.1943$O%.340817@newsc.telia.net>
Message-ID: <kJuX5.3700$Er5.2139@news.indigo.ie>

Hmm..
It's true mode='rb' should be the default parameter in the function call
only when the OS is windows.
I think the module should be modified to work on every OS.. but not by
myself
But I tried the SocketServer with the folowing code and it's working:

import SocketServer, sys,traceback,re

class ReqHandler (SocketServer.StreamRequestHandler):
    def handle(self):
        print 'Starting .. . ..............'
        f = open ('log.log','wb')
        try:
            yf = self.request.makefile()
            line = yf.readline()
            while(line):
                print line
                for xx in line:
                    print ord(xx),
                if line == '\r\n':
                    print "break"
                    break
                if line[:15]=='Content-Length:':
                    lentemp = re.split('\s+',line)
                    length = int(lentemp[1])-1
                    print length
                line = yf.readline()

            line = yf.read(length)

        except:
            traceback.print_exc()

        print line
        print "**"
        f.write(line)
        f.close()
        yf.flush()
        print "**"
        yf.write("HTTP/1.0 200 OK Message\r\n\r\n")
        yf.flush()
        print 'End  .. . . .. . . .. . .. . .. .'

x=SocketServer.TCPServer(('',9080),ReqHandler)

try:
    x.serve_forever()
except KeyboardInterrupt:
    sys.exit()
except:
    pass

 Are you sure is broking the the SocketServer module?

Regards,
Lale


"Fredrik Lundh" <fredrik at effbot.org> wrote in message
news:zodX5.1943$O%.340817 at newsc.telia.net...
> Ilariu Raducan wrote:
> > In class _sockobject in the file socket.py change the 'r' mode with 'rb'
> >
> >   def makefile(self, mode='r', bufsize=-1):
> >         #MODIFIED   return _fileobject(self._sock, 'rb', bufsize)
> >         return _fileobject(self._sock, 'rb', bufsize)
>
> please.  what's wrong with changing the *call* to the
> function, instead of the function itself?
>
> btw, your "fix" breaks SocketServer.py, and god knows
> how many other programs...
>
> </F>
>
>




From moshez at zadka.site.co.il  Mon Dec 25 10:02:50 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Mon, 25 Dec 2000 17:02:50 +0200 (IST)
Subject: Some Python 2.1 ideas
In-Reply-To: <925tl301ub0@news1.newsguy.com>
References: <925tl301ub0@news1.newsguy.com>, <mailman.977590932.19335.python-list@python.org> <Cw616.23054$qN6.3308205@typhoon.we.rr.com> <3A452881.7686E717@uswest.net> <%vb16.187233$U46.5801214@news1.sttls1.wa.home.com> <3A464D5D.B541C04E@uswest.net> <mailman.977688488.25438.python-list@python.org>
Message-ID: <20001225150250.6FA06A84F@darjeeling.zadka.site.co.il>

On Sun, 24 Dec 2000 23:30:30 +0100, "Alex Martelli" <aleaxit at yahoo.com> wrote:

> A mixin-class (implementing, e.g., readline_chomp in terms of .readline
> and chomp, in this case) could also be supplied to ease the task of
> 'emulating convenience-methods'.  Not crucial in this case (the proposed
> methods appear somewhat marginal), but an approach that might help
> in other cases, I think.

I think in that case a function which does the "readline_chomp" is better:
you don't need to change many object's declerations, mix them in, not
to mention that it doesn't help built-in types at all. Frankly, I fail
to see why people are so afraid of functions: generic functions *enhance*
polymorphism.

-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From joseamartin at venmex.com  Mon Dec 11 09:15:06 2000
From: joseamartin at venmex.com (Jose Martin)
Date: Mon, 11 Dec 2000 10:15:06 -0400
Subject: questions 'bout snack module...
References: <90vo12$lu1$1@news1-2.kornet.net>
Message-ID: <t39ornknsdj5c7@corp.supernews.com>

Real Time using UDP ?, I dont think so ,:-O

I think that snack currently have no speed for your needs.





?????? <skidrow at nownuri.net> escribi? en el mensaje de noticias
90vo12$lu1$1 at news1-2.kornet.net...
> I get to know record/play with snack module,
>
> but what I wanna do is get sound data and send it through network in
> real-time...
>
> how can i do that?..(should I use Thread or something?..) plz give me
brief
> example...
>
> p.s) oh, I think I must use UDP instead of TCP/IP
>
> but when I use SOCK_DGRAM instead of SOCK_STREAM, I get "socket error"
>
> when the sock listens...
>
> if possible, plz give me UDP socket program example also... :)
>
>




From MarkH at ActiveState.com  Thu Dec  7 17:40:23 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Thu, 07 Dec 2000 22:40:23 GMT
Subject: [Pythonwin] Pywin encoding (again)
References: <0sNX5.168$M47.3490@m2newsread.uni2.es>
Message-ID: <3A301090.3060304@ActiveState.com>

Arturo P?rez Mulas wrote:

>     - Is it possible to write Python scripts under Pythonwin using 
> non-ascii characters (latin-1 or cp1252 in my case) in comments, 
> docstrings, etc?

Not with any of the relased win32all versions :-(

There have been a few patches posted here just after 135 was released, 
and by all accounts they work fine.

So the next win32all/ActivePython should have it fixed.

Mark.



From sdm7g at virginia.edu  Mon Dec 11 14:08:59 2000
From: sdm7g at virginia.edu (Steven D. Majewski)
Date: Mon, 11 Dec 2000 14:08:59 -0500 (EST)
Subject: variable variables
In-Reply-To: <Pine.LNX.4.21.0012111127460.5044-100000@relax.dk>
Message-ID: <Pine.NXT.4.21.0012111403190.278-100000@localhost.virginia.edu>


On Mon, 11 Dec 2000 pyth at relax.dk wrote:
> 
> Sorry Mikael, for using your time... I should have been more specific...
> What i was thinking of, was something like this:
> 
> test1 = 'test text 1';
> test2 = 'test text 2';
> 
> var1 = 1
> var2 = 2
> 
> print test{var1};
> 
> This should print: 'test text 1'
> This example doesn't work, but is close how you would do it in php :o)
> 

Why not just use an array:

test = [ 'test text 1', 'test text 2' ]

for var in 0,1: 
	print test[var]


Or if you want to index by something other than an int, a dictionary:

test = { 'one':'test text 1' , 'two':'test text 2' }
for var in 'one','two':
	print test[var]


-- Steve Majewski <sdm7g at Virginia.EDU>





From tim.hochberg at ieee.org  Wed Dec  6 11:11:11 2000
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Wed, 06 Dec 2000 16:11:11 GMT
Subject: Calculating factorial
References: <3a2a9496.647995@uutiset.nic.fi> <3A2BCC1F.F308E511@javanet.com> <ya0vgsx3h2f.fsf@levding.ifi.uio.no>
Message-ID: <zytX5.54052$15.11439984@news1.rdc1.az.home.com>

"Roger Hansen" <rogerha at ifi.uio.no> wrote in message
news:ya0vgsx3h2f.fsf at levding.ifi.uio.no...

[SNIP]

> Agree! it's a nice version. But I think the recursive version is even
> clearer.
>
> >>> def fact(n):
> ...     if n > 1:
> ...         return n*fact(n-1)
> ...     return 1
> ...
> >>> fact(4)
> 24
>
> You can change line 3 with
> ...         return long(n)*fact(n-1)
>
> and you have long integers

Alternatively, change "return 1" to "return 1L". Unfortunately, the
recursive version has the drawback that it run's into python's recursion
limit (at around 1000 on my machine).

One more entry that I like is:

def factorial(n):
    result = 1L
    while n:
        result *= n
        n -= 1
    return result

It's not any faster than the other choices, but it has the dubious,
theoretical distinction of using less memory than all but the xrange version
and in theory could handle numbers above sys.maxint (crude estimates
indicate that sys.maxint! would take something around a year to compute on
my machine, discounting memory issues).

-tim




From fredrik at effbot.org  Sun Dec 10 17:26:11 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sun, 10 Dec 2000 22:26:11 GMT
Subject: C API mismatch with 2.0
References: <Pine.SOL.4.30.0012102209040.29674-100000@mimosa.csv.warwick.ac.uk>
Message-ID: <7qTY5.3054$O%.484335@newsc.telia.net>

John J. Lee wrote:
> I compiled Michael Miller's _tableio.c (the C extension bit of TableIO.py)
> on windows using cygwin (with Robert Kern's help), and got the following
> warning about the C API.  I used the python20.lib from the same
> distribution of python 2.0 that I'm trying to import TableIO into.
>
> Can somebody tell me what's going on?  Where did this API version number
> come from for _tableio?

API version 1007 is 1.5.2, version 1009 is 2.0.

looks like you failed to recompile (or install) the module.  to
see which file you're loading, use:

>>> import _tableio
>>> print _tableio.__file__

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->






From pbjorn at uswest.net  Wed Dec 20 15:42:43 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Wed, 20 Dec 2000 13:42:43 -0700
Subject: ANNOUNCE: soundex.py
References: <mailman.977338109.32467.python-list@python.org>
Message-ID: <3A4119C3.2E5E54E1@uswest.net>

Thanks Skip!  I was actually using the soundex module in one of my less
frequently used apps so I hadn't realized that it was gone yet :-)

Could I get you to put an explicit licence on it (it would be much more usable
to me if it had a licence and if the licence wasn't GPL <wink>).

-- bjorn

Skip Montanaro wrote:

> Python 2.0 no longer ships with a soundex module.  Sometime ago, Tim Peters
> and Fred Drake each cooked up replacements written in Python.  I merged them
> together into a single module which is available from
>
>     http://musi-cal.mojam.com/~skip/python/
>
> If you have any questions or comments on the module, please send them my
> way.
>
> --
> Skip Montanaro (skip at mojam.com)
> Support the Mojam.com Affiliates Program: http://www.mojam.com/affl/
> (847)971-7098



From sgm103 at cs.york.ac.uk  Tue Dec  5 08:27:10 2000
From: sgm103 at cs.york.ac.uk (Stuart)
Date: Tue, 5 Dec 2000 13:27:10 -0000
Subject: How do packages and .pth files work under Windows (Py 2.0)
References: <90i4uc$lin$1@news.vanderbilt.edu>
Message-ID: <z26X5.15943$R77.1248181@nnrp4.clara.net>

> It would help me enormously if there were a way to get Python to print
> out a list of where it looks when it tries to resolve an "import". Is
> this possible?

I'm not certain, but i think this does the trick:

import sys
print sys.path

regards,

Stuart





From aahz at panix.com  Sat Dec 30 22:47:05 2000
From: aahz at panix.com (Aahz Maruch)
Date: 30 Dec 2000 19:47:05 -0800
Subject: Lazy Python usage (hack request)
References: <Pine.LNX.4.10.10012301443330.13756-100000@kev.ioc.ee> <92l4tn$raf$1@panix3.panix.com> <92lemh$5or$1@news.nuri.net>
Message-ID: <92ma7p$t5d$1@panix3.panix.com>

In article <92lemh$5or$1 at news.nuri.net>,
June Kim <junaftnoon at nospamplzyahoo.com> wrote:
>"Aahz Maruch" <aahz at panix.com> wrote :
>> 
>> Well, if you're going to stick to it not being an option to make any
>> changes, the answer is "no".  However, you can make a little hack that
>> *does* work:
>> 
>> >>> from lazy import lazy
>> >>> a = lazy.b
>> No name 'b' defined, using default 777
>> >>> print a
>> 777
>> 
>> Inside lazy.py, you define a class called lazy with a __getitem__()
>> method.
>
>Shouldn't the method be __getattr__() instead?

Um, yeah.  Sorry, my brain's a bit faded on this, the second-to-last day
of the first year of the 21st century.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

Ninth Virtual Anniversary: Tomorrow!


From josh at open.com  Sat Dec 16 18:50:03 2000
From: josh at open.com (Joshua Muskovitz)
Date: Sat, 16 Dec 2000 18:50:03 -0500
Subject: determining how much data is on a socket
References: <91gu7q$6pj$1@bob.news.rcn.net>
Message-ID: <3a3bfe0c_3@corp.newsfeeds.com>

You could try either a non-blocking socket or else use the select module to
timeout on the read.  Or you could use timeoutsocket...

-- josh





-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----


From schliep at Octopussy.MI.Uni-Koeln.DE  Fri Dec 15 10:41:11 2000
From: schliep at Octopussy.MI.Uni-Koeln.DE (Alexander Schliep)
Date: 15 Dec 2000 16:41:11 +0100
Subject: directed and undirected graph
References: <91bbli$26l8$1@news.tht.net>
Message-ID: <tq9y9xhitew.fsf@Octopussy.MI.Uni-Koeln.DE>

"Gil Hauer" <gilh at sympatico.ca> writes:

> Does anybody have a module that implements directed and undirected graph
> classes? I searched and searched but have had no luck.

There is on in Gato (http://www.zpr.uni-koeln.de/~gato). 

Yours,
Alexander

-- 
Alexander Schliep                    schliep at zpr.uni-koeln.de
ZPR/ZAIK                             Tel: +49-221-470-6040 (w) 
University of Cologne                FAX: +49-221-470-5160
Weyertal 80                          http://www.zpr.uni-koeln.de/~schliep
50931 Cologne, Germany               


From emile at fenx.com  Sun Dec 31 00:41:09 2000
From: emile at fenx.com (Emile van Sebille)
Date: Sat, 30 Dec 2000 21:41:09 -0800
Subject: Probs with classes: cannot instantiate
References: <20001230.23424514@snafu.btx.dtag.de>
Message-ID: <92mgud$7i4bk$1@ID-11957.news.dfncis.de>

Looks like you'll want to either:

from DirParser import DirParser

-- or --

x = DirParser.DirParser()

--

Emile van Sebille
emile at fenx.com
-------------------


"Markus Vogl" <vogl at mavo.de> wrote in message
news:20001230.23424514 at snafu.btx.dtag.de...
Hi all!

Sorry for asking a [stupid] question:

Building and instantiating a class in the interpreter works
pretty well.
Nevertheless in the file modus it produces error messages
like:
"call of non function"

Example:

This is the interpreter stuff:
[d:\]python
Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)]
on win32
>>> class DirParser:
...     def display(self):
...             print "This is the display method!"
...
>>> x = DirParser()
>>> x.display()
This is the display method!
>>>

This is the file [ DirParser.py] stuff:
class DirParser:
def display(self):
print "This is the display method."

To test this i use classtester.py in the same directory
[which is in
path]:

import DirParser
x = DirParser()
x.display()

The error messages:
x = DirParser()
...
TypeError: call of non-function ( type module )
...

Any help will be appreciated!!
Thank you!!!

TIA
Markus
















From pyth at relax.dk  Thu Dec  7 11:11:26 2000
From: pyth at relax.dk (pyth at relax.dk)
Date: Thu, 7 Dec 2000 17:11:26 +0100 (CET)
Subject: variable variables :o)
Message-ID: <Pine.LNX.4.21.0012071707460.22360-100000@relax.dk>

Hey :o)

Just a simple question:

Is it possible to build variable variables, and/or variable variable
variables :o) eg. in php you something that looks close to this:

$$varvar
or
$var{$var[2]}bla{$blabla}[1]

anyone? :o)

Thanx :o)

/Dan Larsen




From SBrunning at trisystems.co.uk  Tue Dec  5 05:21:28 2000
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Tue, 5 Dec 2000 10:21:28 -0000 
Subject: problems with installer 20_3i
Message-ID: <31575A892FF6D1118F5800600846864D5B1542@intrepid>

> From:	gmcm at hypernet.com [SMTP:gmcm at hypernet.com]
> Barry Scott just pointed out to me that the Python 2.0 imputil 
> does not by default install any method of importing builtins (even though 
> it installs, by default, a file system importer).
> 
> Untested: at the top level of archive_rt (after the imports), put:
> 
> sys.path.append(imputil.BuiltinImporter())
> 
> archive_rt will be imported by the C code long before your script runs, so
> 
> builtins should be available.
 
Gordon,
Thanks for looking into the problem. I'm afraid though, that this fix
doesn't work for me.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk
A man may be a fool and not know it, but not if he is married. - H. L.
Mencken




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From garryknight at bigfoot.com  Sat Dec 23 20:29:59 2000
From: garryknight at bigfoot.com (Garry Knight)
Date: Sun, 24 Dec 2000 01:29:59 +0000
Subject: coding convention conversion
References: <mailman.977516469.1157.python-list@python.org> <920v4s$oqj$2@taliesin.netcom.net.uk> <slrn949c13.40l.grey@teleute.rpglink.com>
Message-ID: <923jin$q2l$1@taliesin.netcom.net.uk>

In article <slrn949c13.40l.grey at teleute.rpglink.com> "Steve Lamb"
<grey at despair.rpglink.com> wrote:

> On Sat, 23 Dec 2000 01:29:00 +0000, Garry Knight
> <garryknight at bigfoot.com> wrote:

>>#!/usr/bin/perl
> 
>     Anyone else find it ironic that someone is solving a problem using
>     Perl in a Python group when the comparable Python would be an
>     excellent exercise to perform?  :)

IMHO some languages seem to fit some tasks better than others. Perl is
designed for just this kind of job. A Python solution would, however,
be far more readable: I have to look very closely at my Perl code to
double-check it. On the other hand, the Python solution would probably
be a fair bit longer. Then again, I'd never use a language like C or
Pascal for a text match/substitution job like this since it would
involve a lot more work than the Perl code I ended up producing.

I look forward to seeing an equivalent Python solution if someone cares
to produce one, since I'm a relative newcomer to the language.

It's-horses-for-courses-ly y'rs - Garry

-- 
Garry Knight
garryknight at bigfoot.com


From piotr at itam.zabrze.pl  Wed Dec  6 05:15:05 2000
From: piotr at itam.zabrze.pl (Piotr Kuchta)
Date: Wed, 6 Dec 2000 11:15:05 +0100
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <3A2DD2CA.14A5D7E2@engcorp.com>
Message-ID: <90l3j9$4bn$1@kapsel.intranet>

Peter Hansen <peter at engcorp.com> wrote:

> Python is actually a scripting language derived from the artificial
> (human) language Esperanto, which was created by Ludwig Zamenhof over

Polish (L. Zamenhof) origin of Python! That's nice ;-)))

> The headquarters of the World Esperanto Association is headquartered in
> Rotterdam, Netherlands.  As it happens Python's creator, Guido van
> Rossum, is also from the Netherlands, where his exposure to Esperanto
> (still widely spoken in that country) led him to attempt to bring the

I didn't noticed people speaking Esperanto in Amsterdam ;), but that's true
that Dutchmen are very talented: common Dutchman speaks well two foreign
languages, English and German.

Best regards,
Piotrek




From peter at engcorp.com  Wed Dec  6 00:46:50 2000
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 06 Dec 2000 00:46:50 -0500
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com>
Message-ID: <3A2DD2CA.14A5D7E2@engcorp.com>

Nigel wrote:
> 
>  I am doing some research and I would be grateful if anybody could give
> me some history on the origins of Python.
> 
> --
> Worker
> Disk Drive Computer Books
> http://disk.vstorecomputers.com

Sounds like an interesting and useful endeavor.  This might give you a
good start:

Python is actually a scripting language derived from the artificial
(human) language Esperanto, which was created by Ludwig Zamenhof over
100 years ago and first published in 1887.  (Esperanto was designed to
be extraordinarily easy to learn *relative* to existing "natural"
languages and has grown to have a following of hundreds of thousands of
fluent speakers.)

The headquarters of the World Esperanto Association is headquartered in
Rotterdam, Netherlands.  As it happens Python's creator, Guido van
Rossum, is also from the Netherlands, where his exposure to Esperanto
(still widely spoken in that country) led him to attempt to bring the
same clarity of design and ease of learning to the computer world.  His
efforts culminated in the 1991 release of Python, precisely 104 years to
the day after Esperanto was first published!

Similar efforts have been made in the past, and even more recently,
including an attempt to derive a computer language from Latin, on the
assumption that if it worked with Esperanto it must surely be just as
effective to use such a flexible language as Latin.  The result of that
effort was a language called Perl.  See, for example,
http://www.csse.monash.edu.au/~damian/papers/HTML/Perligata.html for
further historical information about this.

As others have suggested, a quick search with any search engine would
turn up similar information, but I'm happy to give you a head start. 
Good luck with your project!

Cheers! ;)


From fidelman at world.std.com  Wed Dec  6 20:32:57 2000
From: fidelman at world.std.com (Miles R. Fidelman)
Date: Thu, 7 Dec 2000 01:32:57 GMT
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu>
Message-ID: <G56CB0.K2u@world.std.com>

Nathan Gray (nospam at caltech.edu) wrote:
: What a rude response to a polite and legitimate question.
<snip>
: It's really upsetting how often I've seen this sort of response to a polite
: question on this newsgroup lately.  Tell me, what's so terrible with asking a
: group of experts for pointers when researching a subject?  It's a perfectly

Just to provide some perspective: I run a non-profit that focuses on on
telecom. policy issues. It's amazing how many times I get emails from
people asking for answers to incredibly basic questions - often ones that
are answered on the web site where they found my email address in the
first place.  

It's one thing to get questions from somebody who's done some basic
homework first, it's quite another to get questions that indicate that
someone hasn't done even the most rudimentary research on their
own. Personally, I find that particularly irsome when it comes from a
graduate student or author who could be expected to have some research
skills under their belts.

I always figure that people's time is a resource to use AFTER one has done
one's homework and is now stuck.

<snip>
: If you weren't going to help why did you even bother to reply?  One of the
: great merits of the Python community has always been the willingness of people
: on this newsgroup to help others with their questions.    Since you're so
: savvy why not just help the guy out instead of fueling up the flamethrower?

Now this I agree with. The amount of bandwidth consumed by jumping on
people is absurd :-)

Miles Fidelman


From dsh8290 at rit.edu  Thu Dec 28 17:24:01 2000
From: dsh8290 at rit.edu (D-Man)
Date: Thu, 28 Dec 2000 17:24:01 -0500
Subject: Python for prototyping of the user interface
In-Reply-To: <20001228.17165633@mis.configured.host>; from vbusch@gmx.de on Thu, Dec 28, 2000 at 05:16:56PM +0000
References: <20001228.17165633@mis.configured.host>
Message-ID: <20001228172401.A15784@westsidecnc.rh.rit.edu>

Python doesn't have it's own built-in gui toolkit.  The interpreter
often comes with tkinter, a binding for tk with it.  There are also
bindings for GTK, Qt, MFC, wxWindows and probably others too.

I happen to like GTK.  If you want to use GTK, there is a gui builder
called glade.  (check www.gtk.org for a link)  In addition, there is a
library (with python bindings) called libglade.  libglade reads the
saved project format from glade and builds the gui at runtime.  Gtk
does have notebooks and tables.  In fact, the table is one of the
central widgets in GTK.

wxWindows is another rather intriguing toolkit.  It is essentially an
interface specification that is then implemented in terms of the
native toolkit.  For example, in Unix systems there is wxGTK and
wxMotif while for MS Windows there is wxMFC.  The wx widgets look
like GTK or MFC since underneath the interface is a GTK or MFC widget.
wxWindows also has a gui builder, but it's not free.  (I don't think
it is very expensive though).

This is just a brainstorm but if you use Jython,
you may have access to Java's SWING toolkit.

If you provide more information such as which platform and which
toolkit you are/want to use we can give you more information.

HTH,
-D

On Thu, Dec 28, 2000 at 05:16:56PM +0000, Volkmar Busch wrote:
> Hi,
> 
> I took my first steps in using Python, and I really like the language. I 
> read a lot about using Python for prototyping. As a basis for discussion 
> with our users we use GUI prototypes. Building complex user interfaces in 
> Python to me seems quite complicated, because
> 
> - I am missing a Gui builder
> - I am missing controls like notebooks and tables
> 
> So I'd like to asked the experienced Python users how they solve this 
> problem.
> 
> Thanks in advance,
> Volkmar



From a_olme at hotmail.com  Sun Dec 31 09:29:04 2000
From: a_olme at hotmail.com (Anders Olme)
Date: Sun, 31 Dec 2000 15:29:04 +0100
Subject: array of objects
References: <92n1s7$tq9$1@nnrp1.deja.com> <siE36.2779$Qb7.318673@newsb.telia.net>
Message-ID: <92nfrh$7h7h2$1@ID-48779.news.dfncis.de>

ok. thanks!

//Anders Olme




From fredrik at effbot.org  Sat Dec 16 08:03:53 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sat, 16 Dec 2000 13:03:53 GMT
Subject: __del__() in a Tkinter app
References: <MlI_5.53768$eT4.4433606@nnrp3.clara.net>
Message-ID: <ZKJ_5.110$u17.27823@newsb.telia.net>

"Richard" wrote:
> I have tried adding a __del__() method to a Tkinter application so it can
> remove temporary files when the program exits.

Don't ever do things like that in a __del__ method.   Python isn't
C++, and doesn't make any guarantees that __del__ methods will
ever be called.

To properly shut down a Tkinter application, bind the "<Destroy>"
event on the root/toplevel window.

You can also overload the destroy method:

    class myGUI(Tk):

        def __init__(self, ...):
            Tk.__init__(self)
            # not really needed in 2.0 and later, but it doesn't hurt
            self.protocol("WM_DELETE_WINDOW", self.destroy)

        def destroy(self):
            Tk.destroy(self)
            ... handle shutdown

</F>




From cg at schlund.de  Fri Dec 15 12:28:40 2000
From: cg at schlund.de (Carsten Gaebler)
Date: Fri, 15 Dec 2000 18:28:40 +0100
Subject: Compiling Python 2.0 with Berkeley DB 3.0.55?
References: <3A39DF89.D76B6415@schlund.de> <u4s_5.168$SKV4.36635003@news.randori.com>
Message-ID: <3A3A54C8.E85CB507@schlund.de>

Robin Dunn wrote:
> > In our company we have a Linux server with the Berkeley DB 3.0.55
> > installed and some DB-3.0.55 files. I compiled Python 2.0 out of the box
> > on that machine and the dbhash module worked fine for the DB files. But
> > on my own workstation I get the error message "Invalid argument" from
> > dbhash.open() when I try to open a DB-3.0.55 file. I compiled Berkeley
> > DB 3.0.55 and replaced the db*.h files in /usr/include/ and libdb.a in
> > /usr/lib/ but it just won't work. Any hints?
> >
> 
> Is this is on an existing file created on the other system?

Yes.

> Can you create and use new files with dbhash.open()? 

Yes. But if I run db_dump on these files, it says "unexpected file type
or format", regardless which db_dump version I use (2.0.77, 3.0.55,
3.1.17)

> If so the file may have become corrupted or something.

I transferred it via scp, so I don't expect any corruption.

>  Try running db_verify (one of the utilities built
> when you compiled DB) on the file and see what it tells you.

There is no db_verify among my Berkeley DB utilities.

> P.S. In case you are interested, the next generation of Berkeley DB wrappers
> is available at http://pybsddb.sourceforge.net/

Yes, I've seen that, but for now I'd like to get that "basic stuff"
working.


cg.


From Blair.Lowe at compeng.net  Thu Dec 21 19:31:48 2000
From: Blair.Lowe at compeng.net (Blair Lowe)
Date: Thu, 21 Dec 2000 17:31:48 -0700
Subject: while true: !!!
In-Reply-To: <Kjl%5.2$AH6.83@newsc.telia.net>
References: <mailman.977127721.4398.python-list@python.org>
 <Kjl%5.2$AH6.83@newsc.telia.net>
Message-ID: <v0422080bb66846e9440a@[192.168.11.33]>

Hey!

Happy Holidays.

<and rave>
I wrote TWO articles in this thread, and you have written three, yet 
you ("Fredrik Lundh" ) bitch about me (at least there was a <rant> 
tag, thanks for that;) ) wasting bandwidth and flame me, just to try 
to get in your last word.

If you would like to kill a thread, then learn to send more friendly 
messages. Personally I don't care if there are 200 more messages on 
this issue. My email machine sorts this stuff with procmail (eudora 
pro has some good filtering too), so I can take it or leave it. I 
recommend that others sort their email as well to reduce thread 
stress.
</ and rave>.

Thanks to "Delaney, Timothy" for his gracefull reply, and apologise 
for grilling him so.

<and more rant and rave>
The fact is that I DO care about this issue, and that is why I read 
the mailing list, and I don't think that I should be delegated to the 
pep just because I put in my 2 cents. I would love to do that but I 
do not have 1 to 2 hours to get into the source forge stuff right 
now. I would like to do it, though, and I may yet. Why don't YOU do 
it?

or even better, fix my compilation problem with "compiling 
BeOpen-Python-2.0-1.src.rpm" and stop reading this thread.
</ and more rant and rave>

One thing that we have learned here is that like perl, there is more 
than one way of doing things ... and here is another "english" way of 
putting the statement.

Just because I think that what has become a standard from repeated 
use, and lack of rethinking does NOT mean that I am from another 
planet.

WHO says that while 1 is a "standard python idiom"? Certainly not in 
my company! Go ahead and use while 1: I will know what you mean, and 
therefor this is perhaps a useless discussion as you say, but 
READABILITY is crucial in any computing project, and that is one of 
the reasons python was invented.


I do agree with your position that we "don't say it in English", so ...


Along the lines of:
At 07:16 +0000 2000/12/16, Greg Jorgensen wrote:
>I worked on a huge C/C++ mutant project a few years ago. One of the
>programmers thought this was clever:
>
>   // buried in some nested #include file
>   #define FOREVER true
>   ...
>   while (FOREVER) { ... }

etc.


finished = 0
...
while not finished:
    bla bla bla.


OK. I'll try not to contribute to this thread anymore (but no promises).

Blair.



At 10:05 +0000 2000/12/18, Fredrik Lundh wrote:
>Blair Lowe wrote:
>
><rant>
>
>>  >The idiom is actually
>>  >
>>  >while 1:
>>  >
>>  > do something
>>  >     if condition:
>>  >        break
>>  >
>>  > do something else
>>  >
>>  >Basically, it's for dealing with situations where you can't tell at the
>>  >beginning or end of a loop if the loop is completed - only somewhere
>>  >in the middle. The canonical example is
>>
>
>  > This is exactly what a python program should be NOT doing.
>
>bullshit.
>
>the "while 1" loop IS the standard Python idiom.
>
>no matter what you think, that won't change.  get over it.
>
>>  "while 1" is one of the most cryptic statements in Computer
>  > Science.
>
>what planet do you live on?
>
>>  It is cheap to define true and false once, or (my favourite) define
>>  ever (;;). Although the definition is cryptic, the resulting code is
>>  very readable.
>
>really?  in what english language does
>
>     "while true"
>
>make any more sense than
>
>     "while 1"
>
>maybe you meant:
>
>     "while true is true"
>
>which, of course, is just plain silly.
>
>:::
>
>can we please kill this thread now?
>
>instead of repeating the same stupid opinions over and over
>again, do something useful.  there are lots of PEPs that could
>need your input:
>
>     http://python.sourceforge.net/peps/
>
>if you find something to complain about in there, post to this
>forum (but not to this thread, please).
>
>(and if you really think that "while 1" is the biggest problem on the
>planet right now, feel free to contribute a "how to write potentially
>endless loops" PEP.  just make sure to include all points of view...)
>
></rant>
>
></F>
>
>
>--
>http://www.python.org/mailman/listinfo/python-list

Computer Engineering Inc. http://www.compeng.net
Phone: 780 499 5687 (9 - 5 MST) Fax:   780 435 0693 (24 Hours)
All email advice that is provided for free is without warrantee: use 
at your own risk. 



From andy at reportlab.com  Tue Dec 12 22:42:40 2000
From: andy at reportlab.com (Andy Robinson)
Date: Wed, 13 Dec 2000 03:42:40 +0000
Subject: Dijkstra's Shortest Path algorithm
References: <Pine.GSO.4.21.0012090042370.15110-100000@y.glue.umd.edu>
Message-ID: <dvrd3t40u9l7atgdro59m5ttg7dsfv7c4i@4ax.com>

On Sat, 9 Dec 2000 00:48:00 -0500, Roy Katz <katz at Glue.umd.edu> wrote:

>Hello!
>
>
>For a recent assignment in my computer science class, we coded up a
>graph which implements the Shortest Path algorithm (it's a model of
>the Internet, with 'web pages' as nodes, and the latency between them
>as edges).  It took a good two or three days for me in C++, and
>I was wondering if Python (or Numerical Python) has anything like this
>(a Graph module??).  If there isn't, perhaps I should post my class at
>SourceForge and try to make an extension out of it :)
>
Yes, there is.

Aaron Watters kjBuckets collections library include a graph type
written in C, which has this and several other useful algorithms.
www.chordate.com

- Andy Robinson


From spahievi at vega.bg  Mon Dec 11 15:19:05 2000
From: spahievi at vega.bg (Niki Spahiev)
Date: Mon, 11 Dec 2000 22:19:05 +0200
Subject: creating an odbc link with python
In-Reply-To: <31575A892FF6D1118F5800600846864D5B15A5@intrepid>
References: <31575A892FF6D1118F5800600846864D5B15A5@intrepid>
Message-ID: <1843074248.20001211221905@vega.bg>

11.12.2000, 11:35:41, Simon Brunning wrote:

>> From: Niki Spahiev [SMTP:spahievi at vega.bg]
>>     from odbc_installer import config_data_source, ODBC_ADD_SYS_DSN
 
SB> What's this odbc_installer thing? I don't seem to have it, and nor can I
SB> find it - a couple of references to *using* it on deja, but that's it.

It is part of dynwin package from Sam Rushing www.nightmare.com


-- 
Best regards,
 Niki Spahiev





From sholden at holdenweb.com  Fri Dec 15 09:48:53 2000
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 15 Dec 2000 09:48:53 -0500
Subject: Python for REAL Projects ?
References: <3a380d1e.409238@bendigo.news.origin.net.au>
Message-ID: <3eq_5.5732$n51.113513@e420r-atl1.usenetserver.com>

<dragon at hitech.net.au> wrote in message
news:3a380d1e.409238 at bendigo.news.origin.net.au...

[enquiry as to Python's practicality in real-world projects]

> Can this type of project be sucessfully tackled with Python + Tkinter
> + MySQL ??

Personally I wouldn't use MySQL for commercial projects, because of its lack
of support for transactions and referential integrity constraints.  The
former is slated for implementation at a later date, but the latter appears
to be regarded as a feature by MySQL users - at least according to comments
I have seen to the effect that "these can easily be maintained by the
application" (in New Riders' "MySQL").  I don't know whether this is typical
of the MySQL community.

While it may be true, it has the unfortunate effect of requiring application
code to change as referential integrity constraints are added to the logical
schema, defeating the intent to keep programs and data model independent.

I know that many systems have been built around MySQL, and I'll probably get
flamed for saying this by MySQL advocates, but my impression is that MySQL
is a toy system which will not compete in a true multi-user production
environment with commercial products. Not that there's anything wrong with
that, and if you accept its limitations MySQL can be effective.

Fortunately you aren't limited to MySQL as a back-end choice, as Python has
some really good ODBC interfaces which allow Oracle, SQL Server, Sybase and
many other good relational products to be used as a back-end (including,
quite possibly, the Interbase you already use, though I haven't personally
seen ODBC drivers for it).

For better look-and-feel compatibility with the implementation platform you
may wish to consider wxWindows as a GUI, since Tkinter imposes its own style
which is subtly different from any supported platform.

Python itself, however, is a great language for commercial projects.

regards
 Steve






From qrczak at knm.org.pl  Tue Dec 19 14:11:17 2000
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 19 Dec 2000 19:11:17 GMT
Subject: while true: !!!
References: <mailman.977127721.4398.python-list@python.org> <DnD%5.166633$U46.5342944@news1.sttls1.wa.home.com>
Message-ID: <slrn93vcml.nh7.qrczak@qrnik.zagroda>

Tue, 19 Dec 2000 06:38:27 GMT, Greg Jorgensen <gregj at pobox.com> pisze:

> I remember a C++ programmer I worked with who though that true and
> false were keywords until he discovered this (typical) definition
> in a header file:
> 
>     enum { false = 0, true = -1 };

bool, false and true are keywords in C++ and don't require #inclusion
of any header. The mentioned compiler is not conforming to ISO/ANSI C++
(perhaps because it's old).

In C99 _Bool is a keyword, and bool, false and true are macros defined
in <stdbool.h>.

True converted to an integer is 1, not -1.

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


From a_olme at hotmail.com  Sun Dec 31 05:30:31 2000
From: a_olme at hotmail.com (a_olme at hotmail.com)
Date: Sun, 31 Dec 2000 10:30:31 GMT
Subject: array of objects
Message-ID: <92n1s7$tq9$1@nnrp1.deja.com>

Hello!

How do I create a array of object from my own classes?

If I have a class like this:

class test:
	def dosomething(self):
		return 1234

obj= test()

obj= [test(),test()]

The last line works but is there a more elegant way?

//Anders Olme


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


From moshez at zadka.site.co.il  Fri Dec 29 02:07:21 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Fri, 29 Dec 2000 09:07:21 +0200 (IST)
Subject: Interesting behaviour of the assignment
In-Reply-To: <92fhsr01u88@news1.newsguy.com>
References: <92fhsr01u88@news1.newsguy.com>, <92f4gk$qk6$2@news.nuri.net>
Message-ID: <20001229070721.86D9BA84F@darjeeling.zadka.site.co.il>

On Thu, 28 Dec 2000, "Alex Martelli" <aleaxit at yahoo.com> wrote:

> (Actually, in the case of None, the semantics of is and
> == are identical, but 'is' can be a tiny little bit faster).

No.

>>> class FoolAlex:
...         def __cmp__(self, other):
...             return cmp(None, other)
...
>>> FoolAlex() == None
1
>>> FoolAlex() is None
0

Which is why, in code I write, I always use "x is None" -- faster
and more reliable.
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From jasonic at nomadicsltd.com  Tue Dec  5 15:18:32 2000
From: jasonic at nomadicsltd.com (Jason Cunliffe)
Date: Tue, 5 Dec 2000 15:18:32 -0500
Subject: Smalltalk/ROSE-like prog. environment?
References: <90j2gg$kpf$1@news.rz.uni-karlsruhe.de>
Message-ID: <t2qj8hii4fh3b4@corp.supernews.com>

Bjoern

Yes. There are three free cross-platform tools I recommend you to look at
for deep and enduring inspiration:

1.  Boa Constructor
wxPython visual IDE in Python. It gets better all the time.
http://sourceforge.net/projects/Boa-Constructor/

2. KeyKit
object oriented midi programming toolkit with very interesting interface
also written in KeyKit's own oops script language. Brilliant fun stuff
http://nosuch.com/keykit/

3. PD [Pure Data]
object oriented Midi programming tool.. some very interesting ideas also
about how interfaces, data, object and message passing can and should work.
you could use PD for example to rapidly mockup the sort of system you would
like to see and then go about adapting it for python use. you might even be
able to use PD source for your widget set.
Hard to describe, but potent as soon as you use it. Simplicity of the
interface is part of its genius.
http://iem.mhsg.ac.at/pd/   [be sure to get GEL as well!]

good luck and please keep me posted on this topic

- Jason
___________________________________________________________
Jason CUNLIFFE = NOMADICS['Interactive Art and Technology']
Bjoern Giesler <Bjoern.Giesler at gmx.de> wrote in message
news:90j2gg$kpf$1 at news.rz.uni-karlsruhe.de...
> Hi,
>
> I have a dream... :-)
>
> I'd like to have a programming environment that allows you to do graphical
> OO design (UML, ROSE-like) with access to all classes a project consists
of
> (Smalltalk-like). This thing would allow the programmer to graphically
> design class interfaces and interactions, and edit method source code on a
> per-method basis. I.e. no editing-around in large .py files but working
just
> on the job at hand.
>
> This would have a couple of other nifty features: A test mode in which an
> exception is presented to the user in a dialog box that allows
modification
> of the code _while the program is running_ (this could be done with
> continuations, I believe) and stuff like that.
>
> This is a very hazy notion that has been floating around in my head for a
> while. I believe that Python would be the ideal language for this kind of
> application. Does anybody know of such an environment?
>
> If not, I'd like to give it a shot. Anybody have any ideas or want to
> collaborate with me?
>
> Regards,
> --Bj?rn




From joconnor at cybermesa.com  Tue Dec 12 22:23:02 2000
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Wed, 13 Dec 2000 03:23:02 +0000
Subject: [Q] Are Exceptions used that much in practice?
 
Message-ID: <3A36EB96.1EF845C9@cybermesa.com>

> > I believe I've been misunderstood.  The problem is not "can I write code
> > without handling exceptions" but rather "are exceptions used
> > voluntarily, rather than by compulsion".
>
> The Python interpreter will throw exceptions.  You are not obligated to
> handle them.
-- 

I think his point is that in Java, when you declare a function , you can
declare if it throws a particular exception


	void myFunction() throws MyException {

	}


This requires (the compiler and language spec enforce) that any calls to
myFunction be wrapped in an excetpion handler that will catch at least
MyException exceptions

No, Python does not enforce this

Take care,
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

"God himself plays the bass strings first when He tunes the soul"



From support at internetdiscovery.com  Sat Dec  9 18:40:15 2000
From: support at internetdiscovery.com (Mike Clarkson)
Date: Sat, 09 Dec 2000 23:40:15 GMT
Subject: ANNOUCE: Tix 8.1.1 on http://tix.sourceforge.net
Message-ID: <3a32c0b3.10938890@24.0.228.33>

This is a minor Bugfix release for Tix.  Tix is a powerful high-level
widget set that expands the capabilities of your Tk/Tcl and Python
applications. Using Tix together with Tk will greatly enhance the
appearance and functionality of your application. 

I've also added an except from a paper submitted to the Python
conference on the Python Tix wrappers to the website
http://tix.sourceforge.net.
 
Changes for 8.1.1:

1) Added tixDebug procedure. To activate debugging messages, use
                                         tix configure -debug
2) Minor bug fixes.
3) Bulletproofed ComboBox.tcl against a bug showing up under Python.
4) Fixed Windows to properly handle TCL_STORAGE_CLASS with
USE_TCL_STUBS
 5) Fixed TkConsole declarations in tixWinMain.c
 6) Added winDumpExts.c for 8.3 - should change to use Tcl 8.3's
method.
 7) Rewrote the patches to Tkinter.py so it will work with or without
Tix.
 8) Updated README files to say that Tix should now build under 8.3,
 but it's not "supported" in that it hasn't been tested on it yet.

I'll submit the patches to TkInter to the Python patches on
sourceforge so that the Tkinter core can support Tix automatically.

Mike.


From fredrik at effbot.org  Wed Dec 20 14:58:41 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 20 Dec 2000 19:58:41 GMT
Subject: Import problem
References: <91r1j3$94h$1@bob.news.rcn.net>
Message-ID: <Rb806.721$Qb7.55951@newsb.telia.net>

David Allen wrote:
> What else can cause this?  How is it that you can
> from modulename import * and then NOT have access
> to the things in that namespace?

when Python imports a module, it creates an empty
module namespace.  It then executes the module code
in that namespace.

if the module is already present in the sys.modules cache,
Python doesn't load it again.  Instead, it returns a reference
to the existing module.

"from X import *" imports the module X, and creates
references in the current name space to all public symbols
defined by that module *at that time*.

in your case, you end up calling "from X import *" before
module X is fully populated.

> Any help on figuring this out would be much appreciated.

import rule #4: only use "from import" if you know exactly
what you're doing.

(corollary: if you're using "from import", and get an error
you don't understand, you shouldn't be using it ;-)

for more info, see:

    http://effbot.org/guides/import-confusion.htm

if you have trouble understanding names and references,
this might help:

    http://effbot.org/guides/python-objects.htm

</F>




From tim.hochberg at ieee.org  Tue Dec  5 11:10:03 2000
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Tue, 05 Dec 2000 16:10:03 GMT
Subject: Calculating factorial
References: <3a2a9496.647995@uutiset.nic.fi> <3a2c8c2a_3@corp.newsfeeds.com> <90ib2e0216k@news2.newsguy.com>
Message-ID: <vr8X5.51298$15.10971944@news1.rdc1.az.home.com>

"Alex Martelli" <aleaxit at yahoo.com> wrote in message
news:90ib2e0216k at news2.newsguy.com...
> "Joshua Muskovitz" <josh at open.com> wrote in message
> news:3a2c8c2a_3 at corp.newsfeeds.com...
> > What about using xrange() instead of range() to support *really* large
> > numbers?
>
> Of marginal usefulness in the context of computing factorials.

I agree that in practice, but to be nitpicky, in theory it does matter.

> The memory space taken by a number X grows, roughly, with the
> logarithm of X.  From Stirling's formula, we know that
>     log(N!) about= N log(N) - N
> for large N; for N large enough, this means that N! takes up
> memory space growing as O(N logN).

Agreed.

> Since range(N) takes up memory space growing as O(N), for a
> large enough N this memory consumption becomes irrelevant
> compared to the memory which N! itself takes up.

This is not true. If range supported long integers, for _really_ large N,
range(N) would take up O(sum(log i))=O(log(N!)) memory space. So it would
take up the same memory space as N!. However, since N is limited to
sys.maxint, and since the integers in range always take log(sys.maxint)
space, range(N) actually takes up O(N log sys.maxint) memory which is always
larger than the O(N log N) take by N!.

> There _may_ be a value or two of N such that you have just
> enough memory to compute N! if you carefully husband every
> little bit, which is why the usefulness of xrange is
> "marginal" rather than "non-existent".  But I doubt you'd
> want to wait for the computation times needed for such
> calculations on a typical machine of today (with a few
> hundreds megabytes and several hundreds Meaningless
> Instructions Per Second), particularly with the multiplication
> algorithm that is used by Python's builtin 'long integer'
> numbers.

This is why I agree that range is most likely best in practice.

[SNIP]

-tim





From slhath at home.com  Tue Dec 26 18:09:13 2000
From: slhath at home.com (Scott Hathaway)
Date: Tue, 26 Dec 2000 23:09:13 GMT
Subject: excel client in python
Message-ID: <ty926.92981$yR4.2557991@news1.rdc1.tx.home.com>

Hello,

I have the following vba code from Excel that I want to convert to python
code to manipulate excel from python.  Can anyone help me translate?

  Workbooks.Open Filename="C:\test.html"
  ActiveWorkbook.SaveAs Filename:="C:\test2.xls", FileFormat:=xlNormal, _
        Password:="", WriteResPassword:="", ReadOnlyRecommended:=False, _
        CreateBackup:=False

Here is what I have tried (among many other things):

import win32com.client
myExcel = win32com.client.Dispatch('Excel.Application')

myExcel.Workbooks.Open('c:\\test.html')
myExcel.ActiveWorkbook.SaveAs('c:\\test3.xls', 'xlNormal', '', '', 0, 0)

Here is the error I get:

C:\>python excel.py
Traceback (most recent call last):
  File "excel.py", line 7, in ?
    myExcel.ActiveWorkbook.SaveAs('c:\\test3.xls', 'xlNormal', '', '', 0, 0)
  File "<COMObject <unknown>>", line 2, in SaveAs
pywintypes.com_error: (-2147352567, 'Exception occurred.', (0, 'Microsoft
Excel'
, 'SaveAs method of Workbook class failed', 'C:\\Program Files\\Microsoft
Office
\\Office\\1033\\xlmain9.chm', 0, -2146827284), None)

Can anyone help?

Scott







From darius at bofh.net.au  Mon Dec  4 21:49:52 2000
From: darius at bofh.net.au (KevinL)
Date: Tue, 05 Dec 2000 13:49:52 +1100
Subject: Request for help: arbitrary user code
Message-ID: <200012050249.NAA06208@wobbly.bofh.net.au>

I've been working on a program for a while that executes arbitrary 
user-provided code, and I'm trying to setup a secure, reasonably fast way of 
doing such.  The code is generally provided as functions which are compiled 
and executed multiple times in a sort of fake-binding to various objects.

I'm running into a problem with the way python handles namespaces for said 
functions - that it binds them at function compilation rather than at function 
call, and that they must be dictionaries.  My first response was that 
immutable dictionaries would solve the problem - the other possibility I've 
come up with recently is that being able to change func.func_globals on the 
fly would also solve this.

Anyway, I was hoping to get some comments from better minds than my own, so 
I've setup a zope wiki page explaining the problem I'm having - it's at
http://www.bofh.net.au/mud/doc/zwiki/GlobalNamespace

I'd love to solve this problem - without a good solution, it is starting to 
look like python just isn't suited for providing a multi-user scripting 
platform, which would be depressing, as in most other ways it's near-perfect...
This one problem is prooving intractable for me, tho :(.

TIA,
KevinL




From andy at reportlab.com  Wed Dec  6 14:03:39 2000
From: andy at reportlab.com (Andy Robinson)
Date: Wed, 06 Dec 2000 19:03:39 GMT
Subject: [DAO Database in win32] What is more efficient
References: <0g8X5.152$7U6.3922@m2newsread.uni2.es> <PXlX5.117656$U46.3885672@news1.sttls1.wa.home.com> <ssnX5.41$rZ6.2186@m2newsread.uni2.es>
Message-ID: <3a2e8ce3.1084445@news.demon.co.uk>

"Arturo P?rez Mulas" <arturo_perez at wanadoo.es> wrote:

>> If you aren't using ADO you should be. DAO was replaced by ADO some time
>> ago.
>
>By the way, is mxODBC any better?
>
>Thanks
>
If you just want to do simple queries, there is no need to upgrade. I
have used both modules in production systems and had no trouble
with either.   mxODBC is best if 
(a) you want to work a lot with dates and times,
(b) if you want your Windows and Unix code to have EXACTLY the same
database interface (you can build mxODBC for many database on Unix)
or (c) if it is a high-load application and you want to be sure of
using prepared statements. 

- Andy Robinson


From tim.one at home.com  Mon Dec 25 12:28:20 2000
From: tim.one at home.com (Tim Peters)
Date: Mon, 25 Dec 2000 12:28:20 -0500
Subject: Season's greetings (was Re: Some Python 2.1 ideas)
In-Reply-To: <927suu$r5l$1@panix2.panix.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEHBIFAA.tim.one@home.com>

[Bob Alexander]
> P.S. Merry Christmas!

[Aahz Maruch]
> Bah, humbug.

Sounds like you need something to cheer you up, Santa Aahz!  Elf Timmy has
just the thing:

    http://www.mymiserablechristmas.com/

at-least-you're-not-picking-turkey-bones-out-of-your-feet-ly
    y'rs  - tim




From gherron at aw.sgi.com  Tue Dec 19 14:17:08 2000
From: gherron at aw.sgi.com (Gary Herron)
Date: Tue, 19 Dec 2000 11:17:08 -0800
Subject: pow(-1,0) = 1, -1**0 = -1, why?
References: <7lav3t4ptquheuah74s5ff10ghg54u2dhf@4ax.com>
Message-ID: <3A3FB434.8BAF69FE@aw.sgi.com>

Kirby Urner wrote:
> 
>  Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)] on win32
>  Type "copyright", "credits" or "license" for more information.
>  IDLE 0.6 -- press F1 for help
> 
>  >>> -1**0
>  -1
>  >>> pow(-1,0)
>  1
>  >>> -10**0
>  -1
>  >>> pow(-10,0)
>  1
> 
You're getting confused by operator precedence here.

The expression
  -1**0 
is interpreted as
  -(1**0)
evaluates, with no surprises, to
  -1.

On the other hand,
  (-1)**0
evaluates to
  1,
as does pow(-1,0).

-- 
Dr. Gary Herron <gherron at aw.sgi.com>
206-287-5616
Alias | Wavefront
1218 3rd Ave, Suite 800, Seattle WA 98101



From erno-news at erno.iki.fi  Wed Dec 27 18:21:20 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 28 Dec 2000 01:21:20 +0200
Subject: Reading Unicode Files
References: <3a4a3186.17549945@news.t-online.de>
Message-ID: <ku4rzpsb67.fsf@lasipalatsi.fi>

"Gerson" == Gerson Kurz <gerson.kurz at t-online.de> writes:

| The Windows 2000 registry editor creates .REG files in UNICODE by
| default. Is there an easy way to use something like readlines() for
| it, that is, a list of u-strings ?

you can use the unicode() built-in function to conver the data
into a unicode string, if you know what encoding it is in.
something like
>>> unicode(open('f.reg').read(), 'utf-8')
should work (as long as you get the codec right).

or, you can make a file object workalike that gives you unicode
data. something like

>>> import codecs
>>> encode, decode, reader, writer = codecs.lookup('utf-8')
>>> f = reader(open('f.reg'))
>>> lines = f.readlines()

except with my copy of python, f.readlines() doesn't seem
to work quite like i'd expect it to. but readline() seems to work.

  -- erno


From glenfant at equod.com.nospam  Mon Dec 18 12:46:13 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Mon, 18 Dec 2000 18:46:13 +0100
Subject: Timers in Python?
References: <3a3e42ef.16144323@news.skynet.be>
Message-ID: <91li8l$33h$1@reader1.imaginet.fr>

Have a look to "threading" package and time.sleep(...) function.
Seems that's all you need!
Important, use threading.Lock() objects if threads may use not sharable
resources.
Warning, binary distributions of Python/Unix are not always compiles with
threads options.
You may need to compile Python from sources if required.

import threading, time

Mylock = threading.Lock()

class TimeScheduled(threading.Thread):
    def __init__(self, function, delay):
        self.function = function
        self.delay = delay

    def run(self):
        self.alive = 1
        while self.alive:
            Mylock.acquire() # optional (depends whether unsharable
resources used or not)
            self.function()
            Mylock.release() # optional
            time.sleep(self.delay)

    def kill(self):
        self.alive = 0

def sayhello():
    # Very complex stuff
    print "Hello world"
...
# Every 2 minutes
scheduledHello = TimeSceduled(sayhello, 120)
scheduledHello.start()
...
scheduledHello.kill()
...

"Daniel" <Daniel.Kinnaer at AdValvas.be> a ?crit dans le message news:
3a3e42ef.16144323 at news.skynet.be...
>
> Is there an equivalent of the Delphi Timer-component in Python as
> well?  That is, I want to have two Timers : Timer1 needs to activate
> itself every 2 minutes and Timer2 needs to activate itself every 7
> minutes.
>
> In Delphi I can have 2 Timer.events  :
>
> procedure TForm1.Timer1Timer(Sender: TObject);
> begin
>   //do a special task1
> end;
>
> procedure TForm1.Timer2Timer(Sender: TObject);
> begin
>   //do a special task2
> end;
>
> How is this done in Python?
>
> Thanks for helping out.  Daniel



From robin at jessikat.fsnet.co.uk  Wed Dec 20 19:46:33 2000
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Thu, 21 Dec 2000 00:46:33 +0000
Subject: Fatal Python error: PyThreadState_Get
References: <91rcjo$2dtl$1@news2.ottawa.cyberus.ca>
Message-ID: <VSzptWApLVQ6Ewio@jessikat.fsnet.co.uk>

In article <91rcjo$2dtl$1 at news2.ottawa.cyberus.ca>, Gordon Williams
<g_will at cyberus.ca> writes
>Using
>W95
>Python 2.0
>Pythonwin build 135
>
>I am now in the process of converting some python 1.5.2 code to python 2.0.
>I now get the message below.
>
>
>Fatal Python error: PyThreadState_Get: no current thread
>
>abnormal program termination
>
>I'm sending and receiving information in one thread via a serial port.  In
>the 1.5.2 code I was using the Marshal serial module.  As that only works
>with 1.5.2, I have now converted that to use windows native stuff using Mark
>Hamond's work.
>
>There was a thread earlier this month about this same problem.  I think that
>they were also using W95.  It was unclear whether it was resolved or how it
>was resolved.
>
>Before I start tracking the problem down, is this a known problem?  What
>causes it? Any pointers on where I should look first?
>
>Regards,
>
>Gordon Williams
I had a problem with popen3 and Win95 vs 134 (as I recall), I think that
got solved by 135 so this must be another. Certainly the problem I saw
wasn't observed except on Win95/8 (Win NT and 2000 escaped).
-- 
Robin Becker


From gogo at bluedynamics.com  Wed Dec 20 19:10:52 2000
From: gogo at bluedynamics.com (Georg Gogo. BERNHARD)
Date: Thu, 21 Dec 2000 00:10:52 GMT
Subject: IP adresses are not unique enough.
References: <3a413a17.3416402@192.168.0.1> <3A41475A.5DC4ABD8@san.rr.com>
Message-ID: <3a4148ad.7150467@192.168.0.1>

Well, the thing is that we want to use the mechanism with a ms-sql
server and its "uniqueidentifier" datatype. This GUID is some sort of
standard and is used within DCOM or RPC algorithms and seems to be
very relyable. 

There is a big disadvantage if I use IP adresses since they can be
"unrouteable" as you called, which could mean that they are within a
LAN but using a firewall. The MAC adresses are unique, as I
understood. The Windows CoGreateGUID COM API function combines a
timestamp and a MAC adress. I want to do the same thing in python,
which is maybe not possible.

Thank you for your help,
Gogo.




On Wed, 20 Dec 2000 23:57:14 GMT, Darren New <dnew at san.rr.com> wrote:

>> A plain python algorithm itself would be very useful, since one could
>> use it for RPC / COM as well as for creating indexes for distributed
>> databases.
>
>Well, concatenating one of the IP addresses assigned to the machine to the
>clock would give you something that should be globally unique, if you had an
>IP address that wasn't unroutable. Generate a random number to tack on the
>end, and you're set?
>
>-- 
>Darren New / Senior MTS & Free Radical / Invisible Worlds Inc.
>San Diego, CA, USA (PST).  Cryptokeys on demand.
>Personal malapropism generator free with purchase!
>Steganography: The manual is hidden in the source code.
>Gold mining is only profitable because of the gold found.



From moshez at zadka.site.co.il  Sat Dec 23 16:27:01 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Sat, 23 Dec 2000 23:27:01 +0200 (IST)
Subject: namespaces & indentation
In-Reply-To: <92267v$q3e$1@news.mch.sbs.de>
References: <92267v$q3e$1@news.mch.sbs.de>, <mailman.977538666.16764.python-list@python.org>
Message-ID: <20001223212701.79EF0A84F@darjeeling.zadka.site.co.il>

[Tim Peters]
>1. Strive to test all paths thru each function (good advice regardless of
>language).

[Chris Schaller]
> That's right, but that's not so easy in Python (sorry, I'm a little bit
> fastidious from using FPLs) due to its dynamics.

Tim said "test", not "statically check for errors" on purpose. Even in
FPLs untested branches can have bugs, at the very least logic bugs.
Python does very little in compile-time checking, so in Python,
testing is mandatory.
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From ben.hutchings at roundpoint.com  Tue Dec 12 07:15:16 2000
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 12 Dec 2000 12:15:16 +0000
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <29lZ5.138780$U46.4638547@news1.sttls1.wa.home.com>
Message-ID: <uhf49kf8r.fsf@roundpoint.com>

"Greg Jorgensen" <gregj at pobox.com> writes:
<snip>
> A flame war broke out a few years ago in comp.lang.c++ over which
> was better: for ( i=0; i < N; i++ ) or for (i=0; i < N; ++i ). The
> result is exactly the same, and K&R use the first (postincrement)
> idiom, but it turns out that ++i is usually a tiny bit more
> efficient than i++.
<snip>

For iterators of built-in types (like int) it's unlikely to matter,
but for iterators of user-defined types the latter may be less
efficient as it's harder for the compiler to optimise the code for the
overloaded post-increment operator.

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From blosh at tsbrass.com  Wed Dec 13 15:21:39 2000
From: blosh at tsbrass.com (Brad)
Date: Wed, 13 Dec 2000 20:21:39 GMT
Subject: smtplib
Message-ID: <918lo8$6pg$1@nnrp1.deja.com>

anyone know what this means...I am running Python 1.6 on HPUX10.20

smtplib.SMTPHeloError: (504, 'DNS verification of sending machine failed: no
mail will be accepted')


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


From tim.one at home.com  Mon Dec 25 23:18:02 2000
From: tim.one at home.com (Tim Peters)
Date: Mon, 25 Dec 2000 23:18:02 -0500
Subject: One Python 2.1 idea
In-Reply-To: <928n2s02jko@news1.newsguy.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEHNIFAA.tim.one@home.com>

[posted & mailed]

[on speeding listcomps]

[Alex Martelli]
> I might be, given that I think list comprehensions *deserve* wider
> usage and I'd be loath to think of people who *might* use them...
> and won't because of a silly 5%-performance-issue (I guess I might
> 'think of it as evolution in action', but, I'm a softy!-).

Cool!

> If I'm right that the issue is allocation, then I'd need a different
> opcode at the start (a build-list variation) and one at the end to
> say 'ok, listbuilding done now, please ensure it's a normal list
> without too much extra fat'; the former would change the append
> method pointer of the list object to {either reallocate exponentially,
> or, probably better, build the list in disjoint pieces -- as nothing
> BUT 'append' will be used on the list until ok-done-now time},
> the latter would (if needed) recompact it up and restore the normal
> append method for possible future uses.
>
> Am I wildly off-base...?  If not, it seems to me that procuring
> an acceptable diff-program is going to be the largest part of
> making a patch to submit...:-).

Easy one first:  the Cygwin diff works fine on Windows; or if you grab a
copy of the Python source off of SourceForge, cvs diff works OK too.  There
are detailed instructions for setting up cvs and ssh for use with
SourceForge on Windows at:

    http://python.sourceforge.net/winssh.txt


Harder one:  Unless lists grow very large, all evidence to date has said
that one-at-a-time appending displays linear-time behavior on all platforms
examined.  It's unfortunately the platforms that Guido uses that *never*
display worse behavior than that, so it's never been possible to convince
him it can be a drag in some cases.

That said, the builtin map currently uses PyList_SetItem on a preallocated
list when it can(*), while listcomps use a bound .append method which goes
thru listobject.c's hairer ins1.  Even if resizing is free, ins1 is simply
more involved than PyList_SetItem so costs more.  And it's unclear to me why
builtin_map doesn't use the cheaper-still PyList_SET_ITEM macro.

(*) (Apologies if you already know this) Nothing that's simple, because it's
OK for sequences to lie to map and listcomps about their lengths.  For
example,

>>> class LieLow:
	def __len__(self): return 0
	def __getitem__(self, i): return "abc"[i]

>>> class LieHigh:
	def __len__(self): return 1000000
	def __getitem__(self, i): return "xyz"[i]


>>> x, y = LieLow(), LieHigh()
>>> len(x), len(y)
(0, 1000000)
>>> map(lambda i: i, x) + map(lambda i: i, y)
['a', 'b', 'c', 'x', 'y', 'z']
>>> [i for i in x] + [i for i in y]
['a', 'b', 'c', 'x', 'y', 'z']
>>>

It's actually the raising of IndexError that controls how long the results
are.  In the "map" example, the map over x falls back to using one-at-a-time
append because len(x) guessed too low about the true length, while the map
over y grossly overallocates a result list and cuts it back at the end.

In short, len() can't be trusted, so preallocation can't rely on it
(nevertheless, it's *usually* correct, and trusting it as far as is possible
pays off big for map).

Curiously, listcomps were implemented without any additions to the PVM; in
effect, [f(x) for x in s] is macro expanded to:

    anonymous1 = []
    __1__ = anonymous1.append
    for anonymous2 in s:
        __1__(f(anonymous2))
    # result is now anonymous1

One way to speed that is map-like (where var names are really invisible
temps):

    guess = len(s)
    result = [None] * guess
    i = 0
    try:
        while 1:
            elt = s[i]  # this may raise IndexError
            val = f(elt)
            if i < guess:
                result[i] = val
            else:
                result.append(val)
            i += 1
    except IndexError:
        pass
    if i < guess:
        del result[i:]

There are of course many ways to introduce opcodes to speed that.  It's
curious that map() requires that sequences implement __len__ (although
tolerates them lying), while listcomps currently don't.

Another way to speed it would be to leave the listcomp implementation alone
but make one-at-a-time append faster.  That may be more attractive!
one-at-a-time append is very common in Python programs.  Kevin O'Connor was
interested in this back when we were all too busy finishing 2.0 to play
along; unfortunately, he appears to have vanished.  A DejaNews search should
turn up a patch of his that sped listcomps by 30% for him this way.  That's
both easier and harder <0.5 wink -- see the DejaNews thread for why>.




From junaftnoon at nospamplzyahoo.com  Tue Dec 19 08:10:46 2000
From: junaftnoon at nospamplzyahoo.com (June Kim)
Date: Tue, 19 Dec 2000 22:10:46 +0900
Subject: tkinter event binding to mouse double click
References: <91n7l6$7ij$1@news.nuri.net> <91njgc$l4q$1@nnrp1.deja.com>
Message-ID: <91nmqj$ls4$1@news.nuri.net>

"John Grayson" <johngrayson at home.com> wrote in message
news:91njgc$l4q$1 at nnrp1.deja.com...
> In article <91n7l6$7ij$1 at news.nuri.net>,
>   "June Kim" <junaftnoon at nospamplzyahoo.com> wrote:
> > I'm using Python 2.0 on Win32.
> > In the code following, the binding doesn't seem to work; it can't
> detect
> > mouse double-clicking at all. (Maybe a platform-specific problem, I
> reckon)
> >
> > from Tkinter import *
> >
> > root = Tk()
> >
> > def callback(event):
> >     print "clicked at", event.x, event.y
> >
> > frame = Frame(root, width=100, height=100)
> > frame.bind("<Double-Button-1>", callback)
> > frame.pack()
> >
> > root.mainloop()
> >
> >
>
> The problem is that you have to click on the *frame* not the (empty)
> space in the frame.
>
> Try this:
>
> frame = Frame(root, bg='red', width=100, height=100)
>
>
> John Grayson

Well, it doesn't work, either. Probably it will work on linux but not on
win32.
Thanks anyway.

Best Regards,

June



From fredrik at effbot.org  Sun Dec 31 04:48:48 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sun, 31 Dec 2000 09:48:48 GMT
Subject: HTML Parser
References: <mailman.978224102.28797.python-list@python.org>
Message-ID: <4iD36.2777$Qb7.317090@newsb.telia.net>

HTML ParserDenis Voitenko wrote:
> I am trying to write an HTML parser. I am starting off with a simple
> one like so:

There's already an HTML parser in the standard library.

If you want to render HTML (e.g. to plain text), check out
the htmllib module.  If you want to process HTML (filtering
tags, etc), it's usually easier to use the sgmllib module.

For examples, search for "sgmllib" and "htmllib" on deja.com
(or check my library guide).

If you still insist on writing your own, check the "sgmllib.py"
sources.  It's harder than you think.

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From jakobsen at home.com  Fri Dec 29 23:45:53 2000
From: jakobsen at home.com (Peder Jakobsen)
Date: Sat, 30 Dec 2000 04:45:53 GMT
Subject: Newbie question ... reading modules from directories in Windows
Message-ID: <5Md36.46251$59.13170408@news3.rdc1.on.home.com>

Just a small question:

I have added a "C:\Python20\sandbox" directory to my PYTHONPATH

Now I'm trying to import modules from that directory whole using the
interactive Python prompt, but it only works if I start python.exe in the
directory where the module resides.

I thought by adding the directory where the .py files exist to the Path,
Python would be able to detect them???

Thanks in advance,

Peder Jakobsen, Ottawa




From wware at world.std.com  Mon Dec 25 23:52:51 2000
From: wware at world.std.com (Will Ware)
Date: Tue, 26 Dec 2000 04:52:51 GMT
Subject: ready for C or C++?
References: <20001225031852.04296.00003766@ng-fp1.aol.com>
Message-ID: <G65s83.AwK@world.std.com>

Xmnemonic (xmnemonic at aol.com) wrote:
> If I've learned Python and become proficient with it, am I ready to learn C or
> C++?

The venerable ancient way to learn C is with Kernighan and Ritchie's
book "The C Language". That will start with fairly elementary C programs
and develop the language in a way that seemed reasonable and systematic
to people who'd participated in its design. (Alas my historical knowledge
begins to fail me here... I think there were other designers as well.)
An alternative would be to find one of those plastic 3-ring-binder
cheat sheets, though it might serve as a pretty poor tutorial. Or you
could look around to see whatever's popular in the Learn-C-in-24-hours
category.

Another approach would be to start with Modules/xxmodule.c and
Objects/xxobject.c. The downside to that approach is that you will be
starting with fairly complicated C stuff, the upside is that it will be
well integrated into the Python world with which you are already
familiar, and you'll be starting from stuff that is working (until your
changes break it).

You might want to jump back and forth between the two approaches, making
the jump whenever you get frustrated. Remember that C is a reasonably
complicated thing so don't expect (book titles notwithstanding) that you
will pick up everything you want to know in a very short time.
-- 
import string,time,os;print string.join((lambda x:x[:10]+x[8:])(map(
lambda x:string.center("*"*(lambda x:((x<24) ### Seasons Greetings, Will Ware
*(x-3))+3)(x),24),range(1,28, 2))),"\n") ################ wware at world.std.com


From helend at accessone.com  Fri Dec  8 11:50:44 2000
From: helend at accessone.com (Helen Dawson)
Date: Fri, 08 Dec 2000 16:50:44 GMT
Subject: Python legalities
References: <3A2FC63D.652101A@accessone.com> <3A308881.7FBEBAD4@engcorp.com>
Message-ID: <3A31117F.8B4037CF@accessone.com>

It was Bruce, but I'm sharing an e-mail address for usenet browsing.

There is certainly no guarantee that commercial libraries are patent
or copyright safe. However, our lawyers tell me that when you
license commercial libraries they usually come with a contract that
says that they promise that there aren't any patent or copyright
problems, and they indemnify you against those problems.

It's not giving somebody money that is particularly magic, it's what's
in the licensing agreement. Then again, I don't know how common
this indemnification clause really is, and how effective it really is.
I'm not qualified to debate that.

I agree that many of the 'dangers' of free libraries exist in commercial
libraries, and in in-house libraries (patent problems in particular).

Peter Hansen wrote:

> Helen (or maybe it was really Bruce) Dawson wrote:
> >
> > If we license software from a third party, then part of what we
> > are paying them for is for them to promise that there aren't any
> > copyright or patent issues, and for them to take the legal heat if
> > there are any issues.
> >
> > With free software in general, and Python in particular, there is
> > nobody to indemnify us.
> >
> > Bruce Dawson
> > comments at cygnus-software.com
>
> Somehow I doubt the distinction between "free" and commercial software
> is really that clear.  Is it definitely, legally, contractually the case
> that you are completely safe licensing software for money from a third
> party?  Isn't it possible they have inadvertently included copyrighted
> material from other sources without realizing it (or deliberately), and
> the copyright owner could come after you?
>
> I'm not saying you are definitely wrong, but perhaps the legal eagles
> are reading too much benefit into the difference between licensing with
> money exchanged and licensing without money exchanged.
>
> IANAL, but some of my best friends know people who are.



From clarence at netlojix.com  Sat Dec  9 18:33:35 2000
From: clarence at netlojix.com (Clarence Gardner)
Date: Sat, 9 Dec 2000 15:33:35 -0800
Subject: Dumb Newbie Question
References: <3A327D01.9841EF98@drew.edu>
Message-ID: <976405047.195524076@news.silcom.com>

On Sat, 09 Dec 2000, Scott Weisel wrote:
>I thought I might be able to modify sys.path like this
>sys.path = sys.path + "c:\\python16\\programs 
>but I'm getting a TypeError stating "illegal argument type for built-in
>operation."

sys.path is a list -- you can join lists together with '+' like you
tried, but not a list and a string, so you want:
    sys.path = sys.path + ["c:\\python16\\programs"]

Note that, being an assignment statement, sys.path will refer to
a different object after this statement than before it.  You can also
add your element to the end of the existing list with
    sys.path.append("c:\\python16\\programs")



From quinn at dinar.ugcs.caltech.edu  Wed Dec 13 17:07:09 2000
From: quinn at dinar.ugcs.caltech.edu (Quinn Dunkan)
Date: 13 Dec 2000 22:07:09 GMT
Subject: Bug: Readline getting stuck on Linux and Solaris
References: <lxk8zqwupy2.fsf@ioasun3.epfl.ch> <90m53k$ovh$1@nnrp1.deja.com> <m3u28h48ja.fsf@atrus.jesus.cam.ac.uk> <slrn92u8al.76b.quinn@lira.ugcs.caltech.edu> <m3r93k4kqk.fsf@atrus.jesus.cam.ac.uk> <slrn932eoj.d1j.quinn@ngwee.ugcs.caltech.edu> <m3aea12o2z.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <slrn93fsoc.5pr.quinn@dinar.ugcs.caltech.edu>

On 12 Dec 2000 23:53:56 +0000, Michael Hudson <mwh21 at cam.ac.uk> wrote:
>quinn at ngwee.ugcs.caltech.edu (Quinn Dunkan) writes:
>> (gdb) bt
>> #0  0xfa3ab88 in __ioctl () at lio.c:201
>> #1  0xfa434f0 in ioctl () at nftw.c:184
>> #2  0x5ffd2a74 in rl_restart_output ()
>> #3  0x5ffdd0f4 in _rl_clean_up_for_exit ()
>> #4  0x5ffdd860 in rl_cleanup_after_signal ()
>> #5  0x5ffdd304 in _rl_erase_entire_line ()
>
>Can I see some more of this, please?

Here are some more details tracebacks (|fmt'ed to shut up slrn):

(dbx) where
>  0 __ioctl(0x0, 0x54d7, 0x5fff9d68, 0x0, 0xc, 0xc, 0x7fe35d90, 0x0)
["/xlv51/6.5.5f/work/irix/lib/libc/libc_n32_M3/sys/ioctl.s":20, 0xfaf7dbc] 1
_ioctl(0x0, 0x54d7, 0x5fff9d68, 0x0, 0xc, 0xc, 0x7fe35d90, 0x0)
["/xlv51/6.5.5f/work/irix/lib/libc/libc_n32_M3/sys/ioctlSCI.c":28, 0xfb00720]
   2 ___new_tcsetattr(0x0, 0x54d7, 0x5fff9d68, 0x0, 0xc3ac600, 0xc,
0x7fe35d90, 0x0)
["/xlv51/6.5.5f/work/irix/lib/libc/libc_n32_M3/term/new_tcsetattr.c":61,
0xfb03878]
   3 tcsetattr(__fd = 0, __act = 21519, __t = 0x5fff9d68)
["/usr/include/sys/termios.h":136, 0x5ffc62f4]
   4 set_tty_settings(tty = 0, tiop = 0x5fff9d68)
["/usr/local/src/readline-4.0/rltty.c":428, 0x5ffc6544]
   5 rl_deprep_terminal() ["/usr/local/src/readline-4.0/rltty.c":581,
0x5ffc6954]
   6 rl_cleanup_after_signal() ["/usr/local/src/readline-4.0/signals.c":361,
0x5ffd655c]
   7 rl_signal_handler(sig = 2) ["/usr/local/src/readline-4.0/signals.c":157,
0x5ffd5f84]
   8 sig_fixup_mask(0x0, 0x54d7, 0x5fff9d68, 0x0, 0xc, 0xc, 0x7fe35d90, 0x0)
["/xlv51/6.5.5f/work/eoe/lib/libpthread/libpthread_n32_M3/sig.c":446,
0xc38ebb0]
   9 _sigtramp(0x0, 0x54d7, 0x0, 0x0, 0xc, 0xc, 0x7fe35d90, 0x0)
["/xlv51/6.5.5f/work/irix/lib/libc/libc_n32_M3/signal/sigtramp.s":71,
0xfac93c0]
   10 _ksigaction(0x2, 0x7fe362f8, 0x5fff9ea0, 0x0, 0x11e1000, 0x7fe362f8,
0x7fe363c0, 0xc3a4cb8)
["/xlv51/6.5.5f/work/irix/lib/libc/libc_n32_M3/signal/ksigaction.s":23,
0xfac92ec]
   11 _SGIPT_libc_sigaction(0x2, 0x7fe362f8, 0x5fff9ea0, 0x5ffd5f20,
0x11e1000, 0x7fe362f8, 0x7fe363c0, 0xc3a4cb8)
["/xlv51/6.5.5f/work/eoe/lib/libpthread/libpthread_n32_M3/sig.c":419,
0xc38eab8]
   12 ptctl(0x2, 0x7fe362f8, 0x5fff9ea0, 0x0, 0x11e1000, 0x7fe362f8,
0x7fe363c0, 0xc3a4cb8)
["/xlv51/6.5.5f/work/eoe/lib/libpthread/libpthread_n32_M3/libcthread.c":149,
0xc387cd4]
   13 _sigaction(0x2, 0x7fe362f8, 0x5fff9ea0, 0x0, 0x11e1000, 0x7fe362f8,
0x7fe363c0, 0xc3a4cb8)
["/xlv51/6.5.5f/work/irix/lib/libc/libc_n32_M3/signal/sigaction.c":26,
0xfac98b4]
   14 rl_set_sighandler(sig = 2, handler = 0x5ffd5f20, ohandler = 0x5fff9ea0)
["/usr/local/src/readline-4.0/signals.c":242, 0x5ffd60f4]
   15 rl_maybe_set_sighandler(sig = 2, handler = 0x5ffd5f20, ohandler =
0x5fff9ea0) ["/usr/local/src/readline-4.0/signals.c":259, 0x5ffd6158]
   16 rl_set_signals() ["/usr/local/src/readline-4.0/signals.c":272,
0x5ffd61f0]
   17 rl_reset_after_signal() ["/usr/local/src/readline-4.0/signals.c":371,
0x5ffd65c4]
   18 rl_signal_handler(sig = 2) ["/usr/local/src/readline-4.0/signals.c":179,
0x5ffd5ff8]
   19 sig_fixup_mask(0x2, 0x7fe362f8, 0x5fff9ea0, 0x0, 0x11e1000, 0x7fe362f8,
0x7fe363c0, 0xc3a4cb8)
["/xlv51/6.5.5f/work/eoe/lib/libpthread/libpthread_n32_M3/sig.c":446,
0xc38ebb0]
   20 _sigtramp(0x0, 0x7fe362f8, 0x0, 0x0, 0x11e1000, 0x7fe362f8, 0x7fe363c0,
0xc3a4cb8)
["/xlv51/6.5.5f/work/irix/lib/libc/libc_n32_M3/signal/sigtramp.s":71,
0xfac93c0]
   21 _ksigaction(0x2, 0x7fe36918, 0x5fff9ea0, 0x0, 0x11e1000, 0x7fe36918,
0x7fe369e0, 0xc3a4cb8)
["/xlv51/6.5.5f/work/irix/lib/libc/libc_n32_M3/signal/ksigaction.s":23,
0xfac92ec]
   22 _SGIPT_libc_sigaction(0x2, 0x7fe36918, 0x5fff9ea0, 0x5ffd5f20,
0x11e1000, 0x7fe36918, 0x7fe369e0, 0xc3a4cb8)
["/xlv51/6.5.5f/work/eoe/lib/libpthread/libpthread_n32_M3/sig.c":419,
0xc38eab8]

[ ... 11--21 loop forever ... ]

I get this trace when I do a ^C on an Indigo2 (IRIX64 6.5), even if I do
the ^C after a 'while 1: pass'.  

Another Indigo2 with the same stats gives:
(dbx) where
>  0 _getpid(0x1, 0x7fcd30b0, 0x0, 0x0, 0x0, 0x1, 0x0, 0xc374f0c)
["/xlv47/6.5.8m/work/irix/lib/libc/libc_n32_M4/proc/getpid.s":15, 0xfabb9f8]
   1 <Unknown>() [< unknown >, 0x5ffdd33c]

when it locks after a while 1 ^C, and

>  0 __ioctl(0x0, 0x54d5, 0x7fabcb40, 0x0, 0x0, 0x0, 0x0, 0x0)
["/xlv47/6.5.8m/work/irix/lib/libc/libc_n32_M4/sys/ioctl.s":20, 0xfafebf8]
   1 _ioctl(0x0, 0x54d5, 0x7fabcb40, 0x0, 0x0, 0x0, 0x0, 0x0)
["/xlv47/6.5.8m/work/irix/lib/libc/libc_n32_M4/sys/ioctlSCI.c":28, 0xfb0789c]
   2 ___new_tcgetattr(0x0, 0x54d5, 0x7fabcb40, 0x0, 0x0, 0x0, 0x0, 0x0)
["/xlv47/6.5.8m/work/irix/lib/libc/libc_n32_M4/term/new_tcgetattr.c":26,
0xfb0aa60]
   3 <Unknown>() [< unknown >, 0x5ffd22c8]

(varies, but always begins with the ioctls) after a ^C after the prompt first
appears.

Ok, on an Indy1 (plain IRIX 6.5), I get the same sort of thing.  Which is
weird, because I coulda sworn when I tried it before, 'while 1: pass' ^C
worked right on the indy but not on the indigo.

Oh, and I used dbx because gdb isn't on all the machines, and it seems to give
more info anyway (maybe gdb doesn't agree with sgi cc compiled libs).  The
second indigo2 and the indy compiled python and readline with gcc, while the
first indigo2 uses sgi cc, so perhaps that accounts for the more complete,
consistent, and less confusing info from the first one. All of them lock up on
^C, even under 'while 1: pass'.  They all work correctly if readline is not
imported, or if I manage to hit ^C before readline is imported.  Two of them
work fine when threads aren't compiled in, and I assume the third one would
too, but I haven't tested it.

I also grepped through the preprocessed source of readline, and found
sigactions() but no signals() (well, except in the included protos).  I also
tried readline-2.2.1; no luck.


But that first traceback is nice, because it manages to implicate signals,
readline, and libpthread all at once :)

... just what we need!  Two compilers, two ABIs, 32 bit and 64 bit arches,
multiple library versions, and inconsistent errors!  Irix is such fun.

>-- 
>  In general, I'd recommend injecting LSD directly into your temples,
>  Syd-Barret-style, before mucking with Motif's resource framework.
>  The former has far lower odds of leading directly to terminal
>  insanity.                                            -- Dan Martinez

Y'know, those random quote sigs can be eerily apropos.  "terminal" insanity
indeed.


From rwd at cableinet.co.uk  Tue Dec 19 13:44:25 2000
From: rwd at cableinet.co.uk (Richard Dobson)
Date: Tue, 19 Dec 2000 18:44:25 GMT
Subject: MIDI extensions (WIN32, Mac)?
Message-ID: <3A3FACEF.DAF77229@cableinet.co.uk>

Hello,


I have just started looking at Python seriously, with a focus on music
applications; e.g. algorithmic composition. Python seems to me to be
idally suited as a first programming language for Music Technology
students who have never done programming before. So far the net trawling
I have done has failed to turn up any already-done extensions supporting
MIDI i/o. I could probably code this myself at least for WIN32 if need
be, but it is the sort of thing I feel sure somebody must already have
done. Can anyone suggest anything? Or has it indeed not been done yet?


Richard Dobson

-- 
Test your DAW with my Soundcard Attrition Page!
http://www.bath.ac.uk/~masrwd (LU: 3rd July 2000)
CDP: http://www.bath.ac.uk/~masjpf/CDP/CDP.htm (LU: 23rd February 2000)


From MarkH at ActiveState.com  Sat Dec 30 19:27:20 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sun, 31 Dec 2000 00:27:20 GMT
Subject: PythonWin prompts for win32api methods
References: <t4saigog72et1c@corp.supernews.com>
Message-ID: <3A4E7C47.30207@ActiveState.com>

Jason Cunliffe wrote:

> I have been exploring win32api methods using PythonWin.
> 
> One of the features I like most about PythonWin is the popup prompts for
> objects and methods complete with doc strings where present.
> When I import win32api it does do this. It is easy enough to flip back and
> forth to the win32 helpfiles, but5 so much nicere to do it inline.
> 
> How can I enable this feature please?

In general, you can import the module at the Pythonwin interactive 
window, and that module will then have calltips and autocomplete.

Unfortunately, most of the win32 methods _don't_ have docstrings, as the 
docs are embedded in "autoduck" format.  So in most cases, you wont get 
useful calltips.

Mark.



From thomas at xs4all.net  Tue Dec 26 08:00:32 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Tue, 26 Dec 2000 14:00:32 +0100
Subject: One Python 2.1 idea
In-Reply-To: <929n9c013m3@news1.newsguy.com>; from aleaxit@yahoo.com on Tue, Dec 26, 2000 at 10:03:41AM +0100
References: <mailman.977771949.17911.python-list@python.org> <9292tn$jka$1@nnrp1.deja.com> <929n9c013m3@news1.newsguy.com>
Message-ID: <20001226140032.C25820@xs4all.nl>

On Tue, Dec 26, 2000 at 10:03:41AM +0100, Alex Martelli wrote:
> <rturpin at my-deja.com> wrote in message news:9292tn$jka$1 at nnrp1.deja.com...
>     [snip]
> > Second, a question. Consider the invocation of a method,
> > such as:
> >
> >     x.foo()
> >
> > Does this really cause the lookup of the string "foo"
> > in x's namespace dictionary? The first time? Every time?
> 
> Yes, and yes.  No implicit caching.

Actually, it's even worse :) Methods are never stored in an instance's dict,
but are created 'on the fly' from the class's unbound method. The binding of
the method isn't too expensive, but it isn't totally free either (if
anything, it's a dict 'miss' and an aditional dict lookup in the class's
dict. More than one if you use deep or multiple inheritance (or both.))

You *can* speed it up, by doing 'x.foo = x.foo' somewhere early in your call
chain :) I experimented with this sometime during 2.0 development, but doing
this for every call (by adjusting instance_getattr2 to store bound methods
in the instance dict after creating them) proved only a marginal (and
disputable) improvement, and broke the pickle modules (because bound methods
can't be pickled.) And I'm not sure if the benifit is large enough to
actually build a framework of caching (which Python lacks entirely,
currently) and use it in the few places that might use it.

But if your code can cope with bound methods being stored in instance dicts
(that is, you don't use pickle in any form :) and you use deep or wide
inheritance, you might benifit from this 'optimization' by doing it
manually.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From seandc at worldnet.att.net  Thu Dec 14 17:07:09 2000
From: seandc at worldnet.att.net (Sean Doull-Connolly)
Date: Thu, 14 Dec 2000 22:07:09 GMT
Subject: Python and databases
References: <XFMail.001213131959.mikael@isy.liu.se>
Message-ID: <3A394453.E5453522@worldnet.att.net>

I've used this as a resource:
  Programming With Python 
  By Tim Altom with Mitch Chapman.
  Prima Publishing
  ISBN: 0-76152-334-0
  Paperback, 408 pages, October 1999

There is sample code to read xBASE databases.
I have extended the sample to query my FoxPro
rolodex using Tkinter.  The resulting code
works well in both Linux (X) and WIN98.

Sean
====

Mikael Olofsson wrote:
> 
> On 13-Dec-00 Kevin B wrote:
>  >  Are there any resources that have information about Python and database
>  >  access? I have reviewing several book such as "Programming Python" and
>  >  "Learning Python" and have not found even a chapter on Python and database
>  >  access.
> 
> Well... There is
> 
>   http://www.python.org/topics/database/
> 
> /Mikael
> 
> -----------------------------------------------------------------------
> E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
> WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
> Phone:   +46 - (0)13 - 28 1343
> Telefax: +46 - (0)13 - 28 1339
> Date:    13-Dec-00
> Time:    13:19:33
> 
>          /"\
>          \ /     ASCII Ribbon Campaign
>           X      Against HTML Mail
>          / \
> 
> This message was sent by XF-Mail.
> -----------------------------------------------------------------------


From david at concorde.ndirect.co.uk  Tue Dec 12 02:44:09 2000
From: david at concorde.ndirect.co.uk (David Odhams)
Date: Tue, 12 Dec 2000 07:44:09 -0000
Subject: Dijkstra's Shortest Path algorithm
References: <Pine.GSO.4.21.0012090042370.15110-100000@y.glue.umd.edu>
Message-ID: <3a35d6ff.0@news.netdirect.net.uk>

"Roy Katz" <katz at Glue.umd.edu> wrote in message
news:Pine.GSO.4.21.0012090042370.15110-100000 at y.glue.umd.edu...
> Hello!
>
>
> For a recent assignment in my computer science class, we coded up a
> graph which implements the Shortest Path algorithm (it's a model of
> the Internet, with 'web pages' as nodes, and the latency between them
> as edges).  It took a good two or three days for me in C++, and
> I was wondering if Python (or Numerical Python) has anything like this
> (a Graph module??).  If there isn't, perhaps I should post my class at
> SourceForge and try to make an extension out of it :)
>
>
> Thanks!
> Roey Katz
>
> computer engineering
> University of Maryland, College Park -- USA.
>
>
>
>
> viva esperanto!
>

There is a graph and set C extension called kjbuckets by Aaron Watters at
http://starship.python.net/crew/aaron_watters/kjbuckets/  . The distribution
also includes a shortest path algorithm and other examples.

Regards
Dave Odhams









From aleaxit at yahoo.com  Wed Dec 13 09:23:51 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 13 Dec 2000 15:23:51 +0100
Subject: Inheriting the @ sign from Ruby
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <915l04$l3i$1@nnrp1.deja.com>
Message-ID: <9180qi02jmt@news2.newsguy.com>

"Simon B." <sbrunning at bigfoot.com> wrote in message
news:915l04$l3i$1 at nnrp1.deja.com...
> In article <Pine.GSO.4.21.0012121027001.14542-100000 at y.glue.umd.edu>,
>   Roy Katz <katz at Glue.umd.edu> wrote:
>
> > I've been programming in Python for three years now
> > and I understand only that passing an int or float is call-by-value,
> > otherwise it is call-by-reference
>
> Isn't that Java? It *certainly* isn't Python - *everything* is by
> reference in python. Unless I'm *really* confused...

Python and Java have identical argument-passing semantics
(well, except that Java has static compile-time constraints
on the types of actual vs formal parameters, of course).

Python does have a few more "immutable-types" than Java;
besides numbers and strings, that are immutable in both
languages, Python's tuples are immutable sequences (I do
not believe Java has anything like that), and Python's
instances may easily be made practically-immutable by
raising suitable exceptions in __setattr__ (not a very
frequently used possibility, admittedly), e.g:

class AlmostImmutable:
    def __init__(self, initiallyMutable=None, **attributes):
        self.__dict__.update(attributes)
        self._lock(not initiallyMutable)
    def _lock(onoff):
        self.__dict__['_AlmostImmutable__locked'] = onoff
    def __setattr__(self, name, value):
        if self.__locked: raise AttributeError, "instance is locked"
        self.__dict__[name] = value

inheriting by this mixin class can easily make your own
class-instances 'almost-immutable' (they CAN actually
be changed, via .__dict__ manipulations or use of the
provided ._lock(0) method, but 'normal' attempts at
changing them when 'locked' will raise AttributeError).


But the fact that a given 'thing' is not mutable does not
really affect the fact that a reference to it, rather than
a copy of its value, is what gets passed (in Java as well
as in Python).  Maybe showing this mixin class might help
light dawn in this regard -- clearly, the instances of
classes inheriting from this one have two states, locked
and not locked, they're (ersatz:-) immutable in the locked
state, yet when such instances are passed as arguments the
same, identical mechanism is clearly being used, whatever
their current state may be...

However, I suspect the key issue is that people believe
an argument is 'by reference' if and only if *assigning to*
(as opposed to: calling suitable methods on) that formal
argument will somehow affect something in the caller.  Since
such effects of 'assignment' to formal arguments never occur
in either Python or Java, it may be somewhat hard to convince
people that by-reference IS happening anyway -- that the
real issue is that the semantics they have in mind for
*assignment*, rather than what they have in mind for *argument
passing*, is what is really changed!-)


Alex





From MarkH at ActiveState.com  Tue Dec  5 00:14:41 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Tue, 05 Dec 2000 05:14:41 GMT
Subject: Extending Windows Shell with Python
References: <3A2BADF5.AE73CDD@NOSPAMmaxnm.com>
Message-ID: <3A2C7874.4080106@ActiveState.com>

Itamar Shtull-Trauring wrote:

> http://msdn.microsoft.com/library/psdk/shellcc/shell/Shell_Adv/namespace.htm
> explains how to extend the Windows Explorer so that you can browse your own
> namespaces. I've seen it used, for example, to allow you to browse a FTP
> site through the Windows Explorer.
> 
> Is it possible to this in Python?  I assume it is, and I'm going to try
> doing it myself, but if someone's already done it I'd love to have some
> example code.

Not quite, but close.  Enhancements to the win32com.shell module will 
probably be required.

Funny how these things come in batches!

Mark.



From fredrik at effbot.org  Sat Dec 30 13:56:12 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sat, 30 Dec 2000 18:56:12 GMT
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <3A4D6A05.EEC97FA6@uswest.net> <slrn94qqvg.mgg.grey@teleute.rpglink.com> <ije36.96813$x6.43604873@news2.rdc2.tx.home.com> <slrn94qt4f.mgg.grey@teleute.rpglink.com> <CMo36.97857$x6.44474719@news2.rdc2.tx.home.com> <slrn94s9qg.134.grey@teleute.rpglink.com>
Message-ID: <gdq36.2696$Qb7.302774@newsb.telia.net>

Steve Lamb wrote:
>     From what world do you come from where initilizing variables is bad?

from what world do you come from where initializing
variables to arbitrary values just to get your program
flow right is a good thing?

next you gonna tell us that it's better to use flags
than to use break and/or exceptions...

</F>




From akuchlin at mems-exchange.org  Fri Dec  1 15:53:57 2000
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 01 Dec 2000 15:53:57 -0500
Subject: Regular expression performance
References: <8FFBDFA3Cpgfiguno@195.159.0.219> <_zoV5.958$Kd1.154774@newsb.telia.net> <8FFCA87ADpgfiguno@195.159.0.219> <E_uV5.1030$O%.182639@newsc.telia.net> <8FFCC4A8Cpgfiguno@195.0.200.200> <3dwvdlkut0.fsf@kronos.cnri.reston.va.us> <8FFDC0C47pgfiguno@195.0.200.200>
Message-ID: <3daeafoobe.fsf@kronos.cnri.reston.va.us>

p.g at figu.no (Per Gummedal) writes:
> Now I see that Fredrik was right that sre has problems with IGNORECASE.
> and that my conclusion, that sre has problems with long strings was wrong.
> Instead alternation (p|l) at the beginning of the pattern makes sre very 
> slow. (a bug ?)

I looked at the SRE code too quickly; it seems to do a faster search
when the pattern starts with a literal character, but doesn't actually
determine a set of possible first characters.  So mos95 is fast, but
(p|l) is not because SRE ends up checking every single character.  pre
would determine that any match must start with 'p' or 'l' and then
search for those two characters.

Adding this would therefore require adding a significant optimization
to SRE's engine.  Make a fun 2.1 project...

--amk



From support at internetdiscovery.com  Thu Dec 14 08:58:48 2000
From: support at internetdiscovery.com (Mike Clarkson)
Date: Thu, 14 Dec 2000 13:58:48 GMT
Subject: Freeze.py and Python 2.0?
References: <h2fu2tsqltgh0ujcj224jo62d2dpnu2hnp@4ax.com> <p6qd7f2k5w8.fsf@informatik.hu-berlin.de>
Message-ID: <3a38872c.8769529@24.0.228.33>

On 08 Dec 2000 21:35:51 +0100, Martin von Loewis
<loewis at informatik.hu-berlin.de> wrote:

>Tim Roberts <timr at probo.com> writes:
>
>> The Python doc says that the Freeze tool is present in the Tools/freeze
>> directory of the Python source distribution.  Does this mean I *MUST* build
>> from source in order to get Freeze?  Is there any path to Freeze available
>> to someone using a binary distribution?
>
>No, you just need the tool at first. Since that generates a C file,
>you also need the header files, and libraries. Whether that could
>actually work very much depends on what your distribution provides you
>with.

But how do you get the header files and libraries to be where
freeze.py expects them?  And how does config.c get generated and
compiled.

 Even compiling 2.0 from scratch from the MS project files just gives
you the DLL files; it seems that there is a missing link to get
freeze.py to work.

Any help appreciated.

Mike.


From malan at club-internet.fr  Tue Dec  5 18:11:11 2000
From: malan at club-internet.fr (Alan Marchand)
Date: Wed, 6 Dec 2000 00:11:11 +0100
Subject: an apply question
Message-ID: <90jsj0$24o$1@front6.grolier.fr>

In Python 2.0:

What am I doing wrong?

#!/bin/env python
#############
MyMap = {
'Key1' : 'func1',
'Key2' : 'func2',
'Key3' : 'func3',
'Key4' : 'func4'
}

def func1 ( ):
    print "Here's func1"

items = MyMap.items( )
for x in range( len(items) ):
    key, func = items[x]
    if key == "Key3":
        apply( func, ( ) )
-------------------------
c:\devpy\python test.py

Traceback (most recent call last):
  File "test.py", line 17, in ?
    apply( func, ( ) )
TypeError: Call of non-function (type string)

Thanks.
Alan.




From root at rainerdeyke.com  Sat Dec 16 12:28:03 2000
From: root at rainerdeyke.com (Rainer Deyke)
Date: Sat, 16 Dec 2000 17:28:03 GMT
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <915jcf0c6o@news2.newsguy.com> <9162p3$7t2$1@animus.fel.iae.nl> <MBWZ5.334$h67.37748@newsread1.prod.itd.earthlink.net> <3A397C74.6C8906E1@cosc.canterbury.ac.nz> <91d5520m6j@news1.newsguy.com> <0FE_5.153194$U46.4997180@news1.sttls1.wa.home.com>
Message-ID: <DCN_5.35587$x6.17178643@news2.rdc2.tx.home.com>

"Greg Jorgensen" <gregj at pobox.com> wrote in message
news:0FE_5.153194$U46.4997180 at news1.sttls1.wa.home.com...
> I worked on a huge C/C++ mutant project a few years ago. One of the
> programmers thought this was clever:
>
>   // buried in some nested #include file
>   #define FOREVER true
>   ...
>   while (FOREVER) { ... }
>
> Not to be outdone, another programmer introduced this abomination:
>
>   #define FOR_EVER for ( ; ; )
>   ..
>   FOR_EVER { ... }

#define ever (;;)
for ever { ... }

:-)


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From claird at starbase.neosoft.com  Thu Dec  7 19:40:03 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 7 Dec 2000 18:40:03 -0600
Subject: Parallel Computing
References: <3A300B1D.223630F3@bioeng.ucsd.edu> <874s0fpzr2.fsf@lisboa.ifm.uni-kiel.de>
Message-ID: <DE6EA2149377C08D.DC33C6D21DF0121E.B740EDCD21EB97DD@lp.airnews.net>

In article <874s0fpzr2.fsf at lisboa.ifm.uni-kiel.de>,
Janko Hauser  <jhauser at ifm.uni-kiel.de> wrote:
>There are modules which wrap parts of MPI and PVM, both are available
>from the starship or look at the vaults for a link to them. If you
>want to have parallel communication between Python programs there are
>many different ways, like Dopy, Pyro, CORBA, SOAP, XML-RPC...
			.
			.
			.
Good answer.  The original questioner wrote
>>I haven't had much luck on finding anything about parallel computing via
>>Python.  Does Python support this?  Or do we just need to write a C
			.
			.
			.
It can hardly be overemphasized that "parallel computing"
means different things to different people, and it'll likely
be best for the original questioner to specify his needs a
bit more narrowly.
-- 

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


From jochen at pc1.uni-duesseldorf.de  Thu Dec  7 05:14:18 2000
From: jochen at pc1.uni-duesseldorf.de (Jochen =?iso-8859-1?q?K=FCpper?=)
Date: 07 Dec 2000 11:14:18 +0100
Subject: parallel port access?
References: <3A268378.4DA1B831@sympatico.ca> <3A272881.6F444750@engcorp.com> <3A2D6761.814ABE8E@bioeng.ucsd.edu> <slrn92su1q.12g.ssthapa@localhost.localdomain>
Message-ID: <udhf4gk079.fsf@box.home.de>

>>>>> Suchandra Thapa wrote on Thu, 07 Dec 2000 08:26:19 GMT:

Suchandra> The parallel port and serial port controls are really
Suchandra> platform specific.  Windows has a standard serial interface
Suchandra> but for the parallel interface, windows 98 allows direct
Suchandra> access but nt requires a kernel driver.  Unix systems
Suchandra> probably have there own oddities ranging from non-pc
Suchandra> machines which may not have a parallel or serial port to
Suchandra> whatever access mechanisms linux/*BSD/HURD have.  The task
Suchandra> would probably come fairly close to writing a new driver
Suchandra> for each platform.

Well, on many unices there is the common idea of a printer beeing a
file:-) This means you have a driver sitting behind it, and simply
writing to that file sends your data to the port.
Direct access to the ports in general requires kernel-level priorities
or root-access and going through ioperm or such...

I don't see a need to write a driver for each platform, only DOS
probably, which doesn't bring one by itself [1].

So I would say you'll need a generic wrapper function for sending data
to the OS, that's it. On Unix I'd say all the problem reduces to find
out wether there is a printer-file and what it is called.

Greetings,
Jochen

[1] Then a parport is such a simple device, for most applications you
really don't need a real driver...
-- 
Einigkeit und Recht und Freiheit                http://www.Jochen-Kuepper.de
    Libert?, ?galit?, Fraternit?                GnuPG key: 44BCCD8E
        Sex, drugs and rock-n-roll


From slhath at home.com  Thu Dec  7 19:23:40 2000
From: slhath at home.com (Scott Hathaway)
Date: Fri, 08 Dec 2000 00:23:40 GMT
Subject: creating an odbc link with python
Message-ID: <gSVX5.55717$yR4.1401853@news1.rdc1.tx.home.com>

Hello Everyone,

I am using python to create an installation program for a web app that we
have at our company.  The web app uses an access database that is hooked up
through odbc.  How can I programatically create the dsn in python (I guess I
need to edit the windows registry)?

Thanks,
Scott




From moshez at zadka.site.co.il  Mon Dec 18 01:16:15 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Mon, 18 Dec 2000 08:16:15 +0200 (IST)
Subject: Error evaluating numbers starting in zero
In-Reply-To: <3A3D07BD.85EFEB3D@wag.caltech.edu>
References: <3A3D07BD.85EFEB3D@wag.caltech.edu>
Message-ID: <20001218061615.4DD81A82E@darjeeling.zadka.site.co.il>

On Sun, 17 Dec 2000, "Richard P. Muller" <rpm at wag.caltech.edu> wrote:

> I have a python checkbook program that evaluates date strings, such as
> "01/31/99". Most dates are fine, but if I have a date such as
> "01/08/99", python runs into an error when it tries to eval("08").

Use int() instead of eval().
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From harisri at my-deja.com  Tue Dec  5 00:59:02 2000
From: harisri at my-deja.com (harisri at my-deja.com)
Date: Tue, 05 Dec 2000 05:59:02 GMT
Subject: Newbie question on 'cmpfiles':
Message-ID: <90i074$d73$1@nnrp1.deja.com>

Hello,

I am a newbie user to python 2.0

Can someone help me by explaining 'cmpfiles' function with a simple
example?

I am trying to compare 2 directories, but not sure about the syntaxof
this function. The "Python Reference Library" did not help me on this.

[What does "common" mean in "cmpfiles(dir1, dir2, common)"?]

Thanks in advance,
Hari.


Sent via Deja.com http://www.deja.com/
Before you buy.


From gregj at pobox.com  Wed Dec 27 03:52:42 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Wed, 27 Dec 2000 08:52:42 GMT
Subject: Just starting out!
References: <92buip$kcp$1@nnrp1.deja.com>
Message-ID: <u5i26.194888$U46.5994443@news1.sttls1.wa.home.com>

"birdman" <robericv at my-deja.com> wrote in message
news:92buip$kcp$1 at nnrp1.deja.com...
> I just started teaching myself to write code and wonder if Python is the
> best way to go. Also, Is there a better book than "Learning Python" by
> Mark Lutz and David Ascher available for someone at my level.

Learning Python assumes some programming experience. Check out the free
tutorials at:

http://www.phrantic.com/scoop/tocpyth.htm

(Java and XML tutorials also available on the same site).

Look at the book "Python Essential Reference" by David Beazley (New Riders).
It contains an excellent reference (for Python 1.5.2) and a terse but good
tutorial. The other Python books I've seen are no better than
freely-available web resources.

Reading and studying code until you understand it, and experimenting with
your own code, teaches programming better than books and tutorials. This
newsgroup makes a good study hall.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From mwh21 at cam.ac.uk  Thu Dec  7 04:57:23 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 07 Dec 2000 09:57:23 +0000
Subject: Bug: Readline getting stuck on Linux and Solaris
References: <lxk8zqwupy2.fsf@ioasun3.epfl.ch> <90m53k$ovh$1@nnrp1.deja.com> <m3u28h48ja.fsf@atrus.jesus.cam.ac.uk> <slrn92u8al.76b.quinn@lira.ugcs.caltech.edu>
Message-ID: <m3r93k4kqk.fsf@atrus.jesus.cam.ac.uk>

quinn at lira.ugcs.caltech.edu (Quinn Dunkan) writes:

> Strange.. readline has *never* worked right for me, on IRIX or
> Linux, on any version of python (1.4, 1.5.2, 2.0), and any version
> of readline (3.0, 4.0, 4.1), if python is compiled with threads.
> IRIX (6.4, 6.5, 6.5.2) hangs, Linux (2.0.36, 2.2.12, 2.2.17) exits.

Well, I have readline 2.2.1, which may make a difference.  Can you
supply a stacktrace?

> I just tried your 'while 1: pass' thing, and the behaviour is the same there
> too.

That really is odd, because readline doesn't have control then and
should be out of the picture.  Does control-C work as expected here if
readline isn't enabled, or if you hit it before the first prompt
appears (this may be tricky).

Cheers,
M.

-- 
  Ya, ya, ya, except ... if I were built out of KSR chips, I'd
  be running at 25 or 50 MHz, and would be wrong about ALMOST
  EVERYTHING almost ALL THE TIME just due to being a computer!
                                              -- Tim Peters, 30 Apr 97


From gregj at pobox.com  Wed Dec 20 01:30:06 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Wed, 20 Dec 2000 06:30:06 GMT
Subject: Need Help sorting alphabetically.
References: <3A3F1A91.A7ED9182@araidesign.com> <slrn93u8qc.5lh.scarblac@pino.selwerd.nl>
Message-ID: <OlY%5.172268$U46.5446765@news1.sttls1.wa.home.com>

"Remco Gerlich" <scarblac at pino.selwerd.nl> wrote in message
news:slrn93u8qc.5lh.scarblac at pino.selwerd.nl...
> Chris Arai <chris at araidesign.com> wrote in comp.lang.python:

> > I would like to sort alphabetically (I'm not sure that is the correct
> > term) so that strings of alpha numerics sort alphabetically instead of
> > by ASCII order.
[snip]
> It seems you want to ignore case and ignore everything outside a-z.
> So make a function that throws out all of the other characters and
> turns the rest into a single case, and then compares those:
>
> def compare_alphabetically(a, b):
>    a = filter(lambda x: x.isalpha(), a.upper())
>    b = filter(lambda x: x.isalpha(), b.upper())
>    return cmp(a,b)

The filter is unnecessary; the upper() and lower() methods of string only
change alphabetic characters. A simpler solution is:

def compare_alpha(a, b):
    return cmp(a.lower(), b.lower())
...
m.sort(compare_alpha)

or use a lambda:

m.sort(lambda a,b: cmp(a.lower(), b.lower()))

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From rwklee at home.com  Mon Dec  4 22:40:17 2000
From: rwklee at home.com (Rick Lee)
Date: Tue, 05 Dec 2000 03:40:17 GMT
Subject: Object editors?
Message-ID: <3A2C63A2.E5F8647B@home.com>

I have heard of object/class browsers, either incorporated in IDE's or
stand-alone.  Is there any object/class editor which allows not only
browsing but also editing of object values?  Either stand-alone, or even
better already has an associated GUI built for it.  (I can use the
Python Interpreter, but I am looking for something more user-friendly
for non-Python programmers to use in an application.)

Thanks.

- Rick Lee



From aahz at panix.com  Sat Dec 30 12:06:49 2000
From: aahz at panix.com (Aahz Maruch)
Date: 30 Dec 2000 09:06:49 -0800
Subject: Queuing mechanism that is poll/select'able ?
References: <txk4rzo9o4j.fsf@vortex.ukshells.co.uk>
Message-ID: <92l4n9$r57$1@panix3.panix.com>

In article <txk4rzo9o4j.fsf at vortex.ukshells.co.uk>,
Chris Ebenezer  <chriseb at pobox.com> wrote:
>
>I'm re-writing some python code to use active objects (i.e multiple
>threads).  I have been looking at using the Queue class as a means to
>send messages between different objects.  
>
>However, the one problem is that since the Queue class isn't based
>around a unix abstraction that is descriptor based, it doesn't support
>select().  Since this is the only mechanism that is relatively
>ubiquitous in unix for the use of polling, this causes problems when
>you try to use Queue's in-conjunction with other 'eventing objects'
>like sockets etc.

What might well be easier than what you're suggesting is to spawn off
threads to handle the external events (sockets); those threads then
interact with the standard Queue object(s).
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

Ninth Virtual Anniversary: Tomorrow!


From scarblac at pino.selwerd.nl  Thu Dec 21 11:12:49 2000
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 21 Dec 2000 16:12:49 GMT
Subject: while true: !!!
References: <3A4024E5.98843CFC@cosc.canterbury.ac.nz> <91qadb$1di$1@plutonium.compulink.co.uk> <3A41539E.DF6977D6@cosc.canterbury.ac.nz> <UIn06.2878$PP2.31080@e420r-atl2.usenetserver.com>
Message-ID: <slrn944aqn.8p5.scarblac@pino.selwerd.nl>

Steve Holden <sholden at holdenweb.com> wrote in comp.lang.python:
> Greg Ewing <greg at cosc.canterbury.ac.nz> wrote in message
> news:3A41539E.DF6977D6 at cosc.canterbury.ac.nz...
> > gbreed at cix.compulink.co.uk wrote:
> > >
> > > Take any type, and if
> > > there's a way of it being nothing, it evaluates false.
> >
> > Hmmm, let's try that out. Is [[]] true or false?
> > It's a list containing a list containing nothing.
> > Since [] is false, a list containing nothing must be
> > nothing. So [[]] is a list containing nothing,
> > which is nothing. Ah, it must be false! :-)
> >
> > Nothing-is-weird-stuff,
> >
> if [[]]:
>     print "Ewing is wrong"
> else:
>     print "Ewing is right"
> 
> Ewing is wrong

I'm sure he is perfectly aware of that, and used it to show that 
Graham (gbreed)'s method of checking if something is false doesn't
work.

-- 
Remco Gerlich


From moshez at zadka.site.co.il  Sun Dec 31 10:57:50 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Sun, 31 Dec 2000 17:57:50 +0200 (IST)
Subject: Need help with string method
In-Reply-To: <92if8v$iu1$1@nnrp1.deja.com>
References: <92if8v$iu1$1@nnrp1.deja.com>
Message-ID: <20001231155750.8AE9EA84F@darjeeling.zadka.site.co.il>

On Fri, 29 Dec 2000 16:48:34 GMT, jvanar1 at my-deja.com wrote:
> I have a pretty dumb question here.  I can't seem to get any string
> methods working from *within* a script.  

If you're running your scripts by double-clicking, perhaps an older
installed version of Python is causing you trouble.
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From barry at digicool.com  Mon Dec  4 13:01:30 2000
From: barry at digicool.com (Barry A. Warsaw)
Date: Mon, 4 Dec 2000 13:01:30 -0500
Subject: Mailing lists are moving
References: <31575A892FF6D1118F5800600846864D5B153D@intrepid>
Message-ID: <14891.56314.769765.702624@anthem.concentric.net>

>>>>> "SB" == Simon Brunning <SBrunning at trisystems.co.uk> writes:

    SB> Is the mailing list going to get all the usenet postings that
    SB> it has missed over the past couple of weeks?

Unsure.  First, we don't have the "Usenet story" worked out quiet yet
on the new server, so python-list and python-announce will be the last
to move.  Second, all those messages that are sitting behind the CNRI
Usenet logjam won't get reposted from the new server anyway.  The best
we can hope for is for the logjam to get unwedged on the CNRI side,
and for the new server to have a much more reliable newsfeed.
    
-Barry



From aahz at panix.com  Tue Dec 19 11:55:54 2000
From: aahz at panix.com (Aahz Maruch)
Date: 19 Dec 2000 08:55:54 -0800
Subject: Is Stackless and uthread.py SMP friendly?
References: <91br9v$qa4$1@nnrp1.deja.com> <91dp4j$5fj$1@news.gte.com> <91dt39$fbk$1@nnrp1.deja.com> <m38zpcwno8.fsf@chinon.cnrs-orleans.fr>
Message-ID: <91o3uq$bjf$1@panix2.panix.com>

In article <m38zpcwno8.fsf at chinon.cnrs-orleans.fr>,
Konrad Hinsen  <hinsen at cnrs-orleans.fr> wrote:
>jbalding at hotmail.com writes:
>>
>> Okay. My next question: Is there a simple way of taking advantage of
>> multiple processors using Python, or is this something that is going to
>
>Unless your code spends most of its time in purpose-written C code,
>threads are not the way to go, so you need multiple independent Python
>processes. 

That's not quite true.  Most of the I/O calls in Python release the
global interpreter lock, so I/O heavy code tends to take advantage of
SMP quite well.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

This is a signature anti-virus. 
Please stop the spread of signature viruses!


From ten.llun at ttam  Sun Dec 17 18:51:45 2000
From: ten.llun at ttam (Matthew Schinckel)
Date: Mon, 18 Dec 2000 10:21:45 +1030
Subject: while true: !!!
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <915jcf0c6o@news2.newsguy.com> <9162p3$7t2$1@animus.fel.iae.nl> <MBWZ5.334$h67.37748@newsread1.prod.itd.earthlink.net> <3A397C74.6C8906E1@cosc.canterbury.ac.nz> <91d5520m6j@news1.newsguy.com> <wvw_5.10652$h67.670217@newsread1.prod.itd.earthlink.net> <slrn93mh41.1nk.scarblac@pino.selwerd.nl>
Message-ID: <3a401d05$1_2@news.chariot.net.au>

In message <slrn93mh41.1nk.scarblac at pino.selwerd.nl>, Remco Gerlich wrote:
> Oldayz <sill at localhost.kitenet.net> wrote in comp.lang.python:
> > Hey, how about if you think while 1: looks ugly
> > you do true =3N 1 somewhere on top and use
> > while true: ? I think it's much more obvious
> > and clean and you don't need to change syntax.
> 
> Yes! And you don't need the ugly "break" statement, you can simply do
> "true =3N 0" and the loop will end! (Remember to set it back to 1 later)
> 
:-)

(I'm the guy who forgets to set it back :-)

In my freshman Pascal course, all those years ago (although, until last year
they still taught the same course), I was really pleased that I came up with
(may be bad syntax, haven't used Pascal since 93 :)

HELL_FREEZES_OVER =3N 0

repeat:
 # do stuff
until HELL_FREEZES_OVER


-- 
Matthew Schinckel
matt at null dot net



From grey at despair.rpglink.com  Sun Dec 31 03:13:05 2000
From: grey at despair.rpglink.com (Steve Lamb)
Date: Sun, 31 Dec 2000 08:13:05 -0000
Subject: why no "do : until"?
References: <92jj3j$ff7$1@nnrp1.deja.com> <slrn94shbe.150.grey@teleute.rpglink.com> <X8s36.2731$Qb7.305339@newsb.telia.net> <slrn94smds.150.grey@teleute.rpglink.com> <FOt36.17$vd1.2224@e420r-atl2.usenetserver.com> <slrn94t1vj.150.grey@teleute.rpglink.com> <3A4E95FD.D6C737A7@engcorp.com>
Message-ID: <slrn94tqkh.150.grey@teleute.rpglink.com>

On Sat, 30 Dec 2000 21:12:13 -0500, Peter Hansen <peter at engcorp.com> wrote:
>I understand the point you are making; I'm just pointing out that the
>opposing views seem well-founded.  Areas where I think you missed the
>point are in the importance of code maintainability, the increased
>complexity with your suggested approach, and the feelings of the "entire
>CS community" on this issue.

    I seriously doubt that initing a variable before use causes either code
maintainability problems or increased complexity.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From rwd at cableinet.co.uk  Tue Dec 19 19:11:01 2000
From: rwd at cableinet.co.uk (Richard Dobson)
Date: Wed, 20 Dec 2000 00:11:01 GMT
Subject: MIDI extensions (WIN32, Mac)?
References: <3A3FACEF.DAF77229@cableinet.co.uk> <3A3FE5C3.76B4@loop.com>
Message-ID: <3A3FF97B.BBF116C3@cableinet.co.uk>

Hi Jesse,

thanks, that sound very useful, and I might well get back to you later
on about it, but it isn't what I'm immediately looking for. I want to be
able to at least write MIDI messages to the MIDI OUT port, and hopefully
to be able to read them from MIDI IN too. Then I could write a simple
MIDI filter in Python which would, for example, receive live note data
from MIDI IN, transpose it by some interval, and ouptut the result.


Richard Dobson

Jesse W wrote:
> 
> Dear Richard,
>         I am not sure if this is what you are talking about, but I have written
> a basic class to print out all the events in a MIDI file; I hope to make
> it into a general method to access and modify MIDI files from Python.
> If this might be any help to you please e-mail me.
>                                                 Jesse Weinstein
> 

-- 
Test your DAW with my Soundcard Attrition Page!
http://www.bath.ac.uk/~masrwd (LU: 3rd July 2000)
CDP: http://www.bath.ac.uk/~masjpf/CDP/CDP.htm (LU: 23rd February 2000)


From tim.one at home.com  Thu Dec 21 19:24:09 2000
From: tim.one at home.com (Tim Peters)
Date: Thu, 21 Dec 2000 19:24:09 -0500
Subject: 2 question from a python newbie!
In-Reply-To: <91tq58$cq6$1@uranium.btinternet.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEAMIFAA.tim.one@home.com>

[G. Willoughby]
> i'm not a total programming newbie just a python newbie  :)
> ill check out that site and them, thar books.

Start by working your way through the tutorial that comes with Python.  It
may be all you need.

Else O'Reilly's "Learning Python" is a very good intro for people coming
from other languages.

> P.S. so python is really used as a scripting language and
> nothing more?

Nope.




From ian at cfmu.eurocontrol.be  Mon Dec 18 04:43:13 2000
From: ian at cfmu.eurocontrol.be (Ian Wild)
Date: Mon, 18 Dec 2000 09:43:13 GMT
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <tcmay-FCA22F.14073713122000@news.got.net> <918v7a02scn@news1.newsguy.com> <tcmay-04E80C.16181313122000@news.got.net> <91a1ed02a4j@news2.newsguy.com> <3A389B1A.4F33FFD4@cfmu.eurocontrol.be> <91ae8s02llf@news2.newsguy.com> <3A38CF3E.8BBD2216@cfmu.eurocontrol.be> <91avrn08ql@news2.newsguy.com> <3A39EAB0.1A0528C3@cfmu.eurocontrol.be> <m1snnplma6.fsf@halfdome.holdit.com>
Message-ID: <3A3DDC9D.A29313E2@cfmu.eurocontrol.be>

"Randal L. Schwartz" wrote:
> 
> >>>>> "Ian" == Ian Wild <ian at cfmu.eurocontrol.be> writes:
> 
> >> > classes exist only by convention [in Perl]
> >>
> >> 'only by convention'!?  Have they abrogated the 'bless'
> >> keyword, now?
> 
> Ian> My understanding is that 'bless' does some internal
> Ian> tweakery of a magic array (@INC).  If you promise not
> Ian> to look too hard, you've got classes.  Anything that
> Ian> needs a promise from me I'll regard as convention, not
> Ian> a language construct.  YM, as they say, MV.

[fuller, and more correct, explanation deleted]

> You dismiss Perl too easily.

I don't remember dismissing Perl /at all/.  In fact,
I brought it in to the discussion as an example of
a working OO language where certain (IMHO silly) rules
have no natural place.


From claird at starbase.neosoft.com  Thu Dec 28 16:52:08 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 28 Dec 2000 15:52:08 -0600
Subject: Erlang (was Re: Good introduction to functional programming with Python?)
References: <mailman.977984412.30211.python-list@python.org> <92f0il01a5u@news1.newsguy.com>
Message-ID: <000D41DE0C39A2BB.19C3F8FEDF5B4548.844D5CB760AAABF1@lp.airnews.net>

In article <92f0il01a5u at news1.newsguy.com>,
Alex Martelli <aleaxit at yahoo.com> wrote:
>"Tim Peters" <tim.one at home.com> wrote in message
			.
			.
			.
>> Fine by me.  I must say, though, that Erlang appears to occupy an extreme
>> corner of the FP universe.
>
>It sure looks unique to me for its emphasis on concurrency (if there
>are other FP's that are similarly concurrency-oriented, I have not seen
>them yet).  But Scheme (albeit impure) and Erlang (albeit 'culturally
			.
			.
			.
Clean <URL: http://www.cs.kun.nl/~clean/ > comes close.
-- 

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


From malan at club-internet.fr  Tue Dec  5 18:43:42 2000
From: malan at club-internet.fr (Alan Marchand)
Date: Wed, 6 Dec 2000 00:43:42 +0100
Subject: an apply question
Message-ID: <007b01c05f15$34dcc750$3eaf2cd5@malanwin2000>

In Python 2.0:

What am I doing wrong?

#!/bin/env python
#############
MyMap = {
'Key1' : 'func1',
'Key2' : 'func2',
'Key3' : 'func3',
'Key4' : 'func4'
}

def func1 ( ):
    print "Here's func1"

items = MyMap.items( )
for x in range( len(items) ):
    key, func = items[x]
    if key == "Key3":
        apply( func, ( ) )
-------------------------
c:\devpy\python test.py

Traceback (most recent call last):
  File "test.py", line 17, in ?
    apply( func, ( ) )
TypeError: Call of non-function (type string)

Thanks.
Alan.




From carlgroh at merchantsystems.com  Wed Dec 20 01:52:41 2000
From: carlgroh at merchantsystems.com (carlgroh at merchantsystems.com)
Date: Wed, 20 Dec 2000 01:52:41 -0500 (EST)
Subject: Don't Get Ripped Off!
Message-ID: <20001220065241.A10BFE97D@mail.python.org>

I'm sure... If my competitors get a hold of this letter,
I might as well move out of the country!

Let me explain.

I'm in the business of setting people up to do
business on the Internet. As you probably know, there's only
a few things that you actually need to accomplish this....

1. Something to sell
2. A working website
3. Traffic to the website
4. A way for people to pay you over the net.

Here's my point. While my competitors are charging anywhere from $695
to a few thousand to get people set up to do business on the net,
my company does it all for -- a one time -- complete price of $195!

Maybe that's the reason we currently have over 180,000 clients
using our services. (We list a large number of them on our website).

The $195 includes everything... Your own Domain (website), Free
website hosting unlimited email accounts, autoresponders, shopping
cart system, online credit card acceptace (processing) system,
Internet merchant account, Free software that creates your website,
Free instructions on getting traffic to your website, and many other
Free tools to neumerous to list in a short email message.

Anyway, If you've been thinking about selling online, but
thought it was too expensive or complicated or too confusing,
you've been paying to much attention to my competitors.

If you're interested and want to find out more, It's very simple to do.
Simply call my office -- toll free -- 888-269-7960 and I'll
be glad to explain everything to you in detail. Just be aware,
that I might be on the phone helping someone else out.
So kindly leave a message and I'll get right back to you.

If you're not interested, just delete this message. 
You'll never hear from me again. I guarantee it!

Merchant Systems



From root at [127.0.0.1]  Mon Dec 18 15:33:26 2000
From: root at [127.0.0.1] (deadmeat)
Date: Mon, 18 Dec 2000 20:33:26 GMT
Subject: while true: !!!
References: <B43D149A9AB2D411971300B0D03D7E8B0FBCBB@natasha.auslabs.avaya.com> <mailman.977115541.28453.python-list@python.org>
Message-ID: <qwu%5.8061$KY1.22030@news1.rivrw1.nsw.optushome.com.au>

> This is precisely where having assigment as an expression (like C/C++)
> is useful :
> while ( (line = file.readline()) != None ) :

yuk.. I'd prefer the Pascal way to C's

while file.readline(line) do
 // whatever

however I don't know if variables passed to Python functions can be altered?

> while ( file.exists_next_line() ) :
> line = file.readline()
> do something with the line

I'd go for that.

however the file reading thing is just one example of while 1: so being able
to avoid it in this case is hardly any breakthough.





From aleaxit at yahoo.com  Thu Dec 14 03:56:11 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 14 Dec 2000 09:56:11 +0100
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <914v6l02ouv@news2.newsguy.com> <Pine.GSO.4.21.0012121027001.14542-100000@y.glue.umd.edu> <UksZ5.3364$O%.525994@newsc.telia.net> <Pine.GSO.4.21.0012121142100.14542-100000@y.glue.umd.edu> <slrn93cotl.80c.scarblac@pino.selwerd.nl> <r0pzoi11lhz.fsf@black.cs.washington.edu> <5jvZ5.3588$Kd1.418770@newsb.telia.net> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com> <kuu28971kp.fsf@lasipalatsi.fi> <3A380E4E.8537A0FB@san.rr.com>
Message-ID: <91a1vr02b02@news2.newsguy.com>

"Darren New" <dnew at san.rr.com> wrote in message
news:3A380E4E.8537A0FB at san.rr.com...
> Erno Kuusela wrote:
> > | Having
> > |  x = 2 + 2
> > |  print x
> > | result in anything except "4" is probably a *really* bad idea.
> > | AFAIK, only FORTH and really old Fortrans allow this.
> >
> > agreed - but this is not something (sane) a mutable integer object
> > would break.
>
> Well, X is a variable, and you can already change the value in X.

This _may_ harbor a slight misconception, coming from other languages,
that (in more complex cases) may interfere with optimal understanding
of Python.

There is no value *IN* x.

A variable *REFERS TO* 'a value' (aka 'an object').

The distinction is not very important for immutable values (objects):
although they're not really 'IN' any of the variables (or other
'slots') that REFER TO them, you may still write perfectly good
Python code even if labouring under a misconception that they are.

But as soon as mutable values (objects) enter the picture, then
thinking of values (objects) as being IN variables (or other slots)
brings you nothing but trouble.  It's a good mental model for,
say, C.  It's a terrible one for Python (or, as it happens, Java)...

It's a minor point, but getting into the habit of saying 'changing
what value X refers to' (when you mean peeling the postIt-like
'tag' that we call X, which can be affixed to different objects
at different times), or equivalently 'rebinding X to another value',
can, I think, help avoid certain annoying 'mental slips'...


Alex





From emile at fenx.com  Sat Dec 30 13:47:28 2000
From: emile at fenx.com (Emile van Sebille)
Date: Sat, 30 Dec 2000 10:47:28 -0800
Subject: Lazy Python usage (hack request)
References: <Pine.LNX.4.10.10012301443330.13756-100000@kev.ioc.ee>
Message-ID: <92lak2$7eais$1@ID-11957.news.dfncis.de>

Hi Pearu,

You've got some interesting questions today. ;-)  You did
ask for a hack, and you don't place a restriction on how
python is invoked, so, yes, I think this can be done.  Using
popen2, you could accept input, pass it through to python,
and display output.  In between, you can validate and play
behind the scenes games to get the session results you ask
for.

This is probably only useful for light interactive sessions,
but I do think it would work, and you wouldn't even need to
import lazy.

--

Emile van Sebille
emile at fenx.com
-------------------


"Pearu Peterson" <pearu at ioc.ee> wrote in message
news:Pine.LNX.4.10.10012301443330.13756-100000 at kev.ioc.ee...
>
> Hi,
>
> Is it possible to achive the following behavior in Python?
>
> >>> import lazy
> >>> a = b
> No name 'b' defined, using default 777
> >>> print a
> 777
>
> Note that changing the python session above is not an
option. (Also 'No'
> for the answer;-)
>
> Just to avoid questions 'Why do you need it?', I would use
it in
> PySymbolic (but not in its Kernel!) for defining new
Symbol instances in
> fly (it's very tedious to define 'b = Symbol("b")',etc all
the time).
>
> Here lazy.py would be something like the following:
> #Begin of lazy.py
>
> import UserDict,sys
>
> class myDict(UserDict.UserDict):
>     def __getitem__(self, key):
>         if not self.data.has_key(key):
>             print 'No name %s defined, using default
777'%(`key`)
>             self.data[key] = 777
>         return self.data[key]
>
> sys.modules['__builtin__'].__dict__ =
myDict(sys.modules['__builtin__'].__dict__)
>
> #EOF lazy.py
>
> When trying to set sys.modules['__builtin__'].__dict__, an
exception
> TypeError: read-only special attribute
> is raised.
>
> Is there anyway to go around that? May be in C level?
>
> Any hints, suggestions, or similar are appreciated.
>
> Thanks,
> Pearu
>




From gerson.kurz at t-online.de  Sun Dec 31 11:34:57 2000
From: gerson.kurz at t-online.de (Gerson Kurz)
Date: Sun, 31 Dec 2000 16:34:57 GMT
Subject: More Questions on Embedding Python
Message-ID: <3a4f5df5.33586715@news.t-online.de>

1. Question: 

The following script, slightly modified from demo/embed.c, works just
as expected.
-----------------------------------------------------------
static PyObject* gped_OutputDebugString(
	PyObject *self, PyObject *args )
{
    char* buffer = 0;
    if( PyArg_ParseTuple(args, "s", &buffer))
        OutputDebugString(buffer);
    return NULL;
}

static PyMethodDef gped_methods[] = {
    {"OutputDebugString", gped_OutputDebugString, 1},
    {NULL, NULL}
};

bool PythonSupport::Initialize()
{
    Py_Initialize();
    Py_SetProgramName("gped");
    PyObject* pThisModule = PyImport_AddModule("gped");
    Py_InitModule("gped", gped_methods);

    // startup code
    PyRun_SimpleString("import gped, sys\n");
    PyRun_SimpleString("class gped_stdout:\n"
                       "\tdef write(self,str):\n"
                       "\t\tgped.OutputDebugString(str)\n"
                       "\t\treturn 1\n" );
    PyRun_SimpleString("sys.stdout = gped_stdout()\n" );
    PyRun_SimpleString("print(\"Hello, World\")\n" );
    PyRun_SimpleString("print sys.modules.keys()\n" );
    return true;
}
-----------------------------------------------------------
I will see both "Hello, World" and the sys keys. However, if I issue 

    PyRun_SimpleString("sys.stderr = gped_stdout()\n" );

right before the Hello World statement, I get a "SystemError", and DO
NOT see the keys. Can somebody please explain ?

2. Question: 

Say I want to use a Python as a configuration script for my editor.
For example, I would like to have a configure.py that contains the
lines

	gped.MainWindowPos = [0,0,1024,800]

How can I query that list ? If I use PyArg_ParseTuples, I must first
have a parent object to begin with. 

3. Happy New Year!



From matt at virtualspectator.com  Thu Dec 21 14:45:50 2000
From: matt at virtualspectator.com (matt)
Date: Fri, 22 Dec 2000 08:45:50 +1300
Subject: How to make a GUID with plain python
In-Reply-To: <cuiae9qkq21.fsf@loki.1012internet.at>
References: <3a413a17.3416402@192.168.0.1> <cuiae9qkq21.fsf@loki.1012internet.at>
Message-ID: <00122208495101.00860@localhost.localdomain>

And the Mac address from a modem would be ?
The Mac address from an IDSN converter would be?

.....in a lot of instances Mac addresses are just not there !!!


On Thu, 21 Dec 2000, Michael Lausch wrote:
> gogo at bluedynamics.com (Georg Gogo. BERNHARD) writes:
> 
> > Hello!
> > 
> > I am looking for a way to create a global unique identifier / UUID
> > with python. I found a way to do this with ActivePython in Windows,
> > but I need a solution for other platforms, too. (In ActivePython one
> > can use >>> pythoncom.CreateGuid <<< to get a global unique
> > identifier.)
> > 
> > The GUIDs use the MAC adress of the machine they are created at in
> > their algorithm. I start to think that this can not be done in python,
> > but I still hope that I am wrong in that point.
> 
> 
> use os.popen('/sbin/ifconfig -a','r') and regexp match the output for
> mac addresses.
> 
> -- 
> "Reality is that which, when you stop believing in it, doesn't go away".
>                 -- Philip K. Dick
> -- 
> http://www.python.org/mailman/listinfo/python-list
-- 



From rturpin at my-deja.com  Fri Dec 29 12:01:09 2000
From: rturpin at my-deja.com (rturpin at my-deja.com)
Date: Fri, 29 Dec 2000 17:01:09 GMT
Subject: Python vs. PHP (& Java?)
References: <69BC4C6AE83ED311BC9400805FA7B14AF84D2C@il93exp01.css.mot.com> <mailman.978059571.14515.python-list@python.org>
Message-ID: <92ig0g$jgl$1@nnrp1.deja.com>

On Thu, 28 Dec 2000, Kemp Randy-W18971 <Randy.L.Kemp at motorola.com>
wrote:
>> Could we argue that an artificial language, such as
>> Esperanto, is better to use around the world then a
>> natural language, like English?

In article <mailman.978059571.14515.python-list at python.org>,
  Moshe Zadka <moshez at zadka.site.co.il> wrote:
> Yes. I think this would be wonderful, except political
> constraints dictate otherwise.

The greatest problem is cultural inertia. Knowledge of
their primary tongue is most people's major intellectual
and social asset. The cost of changing this is high,
especially since language is best learned at a young
age. Few would acquire as great proficiency in a new
tongue, regardless of its technical advantage.

Looking at something where the stakes are much less, the
benefit for each individual more clear, and the new skill
is easily learned, cultural inertia STILL wins out over
better notation. No one claims that the QWERTY keyboard
is technically superior to the Dvorzak. I suspect, though,
that the QWERTY keyboard will remain the major character
input device for the next two or three decades. (Before
thinking that voice recognition will knock it out soon,
see how long your throat lasts speaking ten pages of text.
Twice, with edits.)

Russell


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


From olczyk at interaccess.com  Sat Dec 23 10:01:20 2000
From: olczyk at interaccess.com (Thaddeus L. Olczyk)
Date: Sat, 23 Dec 2000 15:01:20 GMT
Subject: Problem with python mode
References: <3a44bcf1.73260781@nntp.interaccess.com>
Message-ID: <3a45be02.73533281@nntp.interaccess.com>

On Sat, 23 Dec 2000 14:59:06 GMT, olczyk at interaccess.com (Thaddeus L.
Olczyk) wrote:

>When i try loading pdb.el in NTemacs I keep getting an error--
>Symbol's function defiition void: toolbar-make-button-list.
>Any idea what the problem is?
>TIA
OOPS. Sorry the version of NTEmacs is 20.7 pdb.el is 1.6.


From gregj at pobox.com  Thu Dec 14 02:56:59 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Thu, 14 Dec 2000 07:56:59 GMT
Subject: xreadlines (was Re: while true: !!!)
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <slrn93g37o.9tj.jepler@potty.housenet>
Message-ID: <f3%Z5.146409$U46.4768601@news1.sttls1.wa.home.com>

"Jeff Epler" <jepler at inetnebr.com> wrote in message
news:slrn93g37o.9tj.jepler at potty.housenet...
> It seems that the most common place that people write the 'while 1:
> ... if test: berak' idiom is when iterating over lines from a file
> where the file is not known to fit in memory.
> ...
> I take this cue to suggest the idea of an 'xreadlines' object.  I have
> implemented 'xreadlines' in Python (where it was slower than all other
> solutions, probably due to method call overhead) and in C (where it
> holds its own with the fastest idiom I know,

There is the fileinput module, which also solves the "ugly infinite loop"
problem that seems to bother some people:

import fileinput

for line in fileinput.input(filename):
    print line

The fileinput module can iterate over a single file, a list of files, or
stdin. I've used fileinput to go through big lists of files (10,000+ email
messages) and it works great. It doesn't appear to do any buffering
itself--it uses file.readline() to read the files.

--
Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com




From jp at ulgo.koti.com.pl  Sun Dec 24 07:41:01 2000
From: jp at ulgo.koti.com.pl (Jacek =?iso-8859-2?Q?Pop=B3awski?=)
Date: 24 Dec 2000 12:41:01 GMT
Subject: string split
References: <slrn94b2go.srm.jp@localhost.localdomain> <mailman.977652367.833.python-list@python.org>
Message-ID: <slrn94brmt.mo.jp@localhost.localdomain>

Tino Wildenhain wrote:
>Anyway, using sgmllib or htmllib could be much more convenient.

yes :) using lynx or mozilla will be easy, but I think this kind
of script is great way to learn python :-D
thanks for help!

-- 
They'll take your diamonds, and then give you steel     
You'll be caught in the middle of the madness     
Just lost like them                                 "All The Fools Sailed Away" 
And part of all the pain that they feel                     - Ronnie James Dio


From dale at out-think.NOSPAMco.uk  Mon Dec  4 08:15:33 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Mon, 04 Dec 2000 13:15:33 +0000
Subject: Mutual import?
Message-ID: <r16n2tkqepahf288gmu6tku5d2sctb72o0@4ax.com>

Is it considered safe and acceptable for two modules to import each other or are there likely to be
unwelcome side effects of that?

The problem is that both require access to values and routines in the other.

Thanks
--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants




From ian at cfmu.eurocontrol.be  Fri Dec 15 04:54:32 2000
From: ian at cfmu.eurocontrol.be (Ian Wild)
Date: Fri, 15 Dec 2000 09:54:32 GMT
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <tcmay-FCA22F.14073713122000@news.got.net> <918v7a02scn@news1.newsguy.com> <tcmay-04E80C.16181313122000@news.got.net> <91a1ed02a4j@news2.newsguy.com> <3A389B1A.4F33FFD4@cfmu.eurocontrol.be> <91ae8s02llf@news2.newsguy.com> <3A38CF3E.8BBD2216@cfmu.eurocontrol.be> <91avrn08ql@news2.newsguy.com>
Message-ID: <3A39EAB0.1A0528C3@cfmu.eurocontrol.be>

Alex Martelli wrote:
>
> [go away, write a book, and]
> become the most sought-after consultant in the industry,

Sorry - I'm with Dogbert on "consultants".


[ian:]
> > classes exist only by convention [in Perl]
> 
> 'only by convention'!?  Have they abrogated the 'bless'
> keyword, now?

My understanding is that 'bless' does some internal
tweakery of a magic array (@INC).  If you promise not
to look too hard, you've got classes.  Anything that
needs a promise from me I'll regard as convention, not
a language construct.  YM, as they say, MV.


From MarkH at ActiveState.com  Wed Dec 20 23:59:32 2000
From: MarkH at ActiveState.com (Mark Hammond)
Date: Thu, 21 Dec 2000 04:59:32 GMT
Subject: Events in win32com
References: <Y1g=Op6IfWSoJn1ISFtOyuvqwBb3@4ax.com> <FY0=OlZFLy71DtnK62u5TEPlhOts@4ax.com> <3A40060B.10504@ActiveState.com> <cYxAOmxi9FgNERgBo29zuCl5c2h1@4ax.com> <ik724tc2v3cpcqfbt7i1ookvaona0mk5ql@4ax.com>
Message-ID: <3A418D02.2040302@ActiveState.com>

Paul Moore wrote:

> Is it possible to do anything like this in Python? If not, can I do
> the same thing another way? (I have created ie via
> DispatchWithEvents(), but the HTML document is a property of this. Can
> I rewrap it somehow, say as
>   doc = DispatchWithEvents(ie.document, DocEvents)

This is (should be) as simple as instantiating a sub-class of the makepy 
generated event class, and passing the document.

eg, something like:

mod = EnsureModule(...)

class MyEvents(mod.IDocumentEvents):
   # your methods here....

handler = MyEvents(ie.document)

# handler should start recieving events.

Mark.



From slhath at home.com  Mon Dec 11 22:43:05 2000
From: slhath at home.com (Scott Hathaway)
Date: Tue, 12 Dec 2000 03:43:05 GMT
Subject: odbc link from python
Message-ID: <d9hZ5.64115$yR4.1675406@news1.rdc1.tx.home.com>

I asked how to create an odbc link from python a couple of days ago.  Thanks
to help from Alex Martelli, I have solved my problem.  If anyone else is
interested, here is the code below.

Since I am just beginning with python, I am sure that vast improvements can
be made.  This code adds a MS Access System dsn named "tri2".

import _winreg as w

parentKey =
w.OpenKey(w.HKEY_LOCAL_MACHINE,r'Software\ODBC\ODBC.INI',0,w.KEY_WRITE)
newKey = w.CreateKey(parentKey,'tri2')
w.SetValueEx(newKey,'DBQ',0,w.REG_SZ,r'c:\apache\htdocs\t2\mdb\Trinity.mdb')
w.SetValueEx(newKey,'Driver',0,w.REG_SZ,r'c:\windows\system\odbcjt32.dll')
w.SetValueEx(newKey,'DriverId',0,w.REG_DWORD,25)
w.SetValueEx(newKey,'FIL',0,w.REG_SZ,'MS Access;')
w.SetValueEx(newKey,'SafeTransactions',0,w.REG_DWORD,0)
w.SetValueEx(newKey,'UID',0,w.REG_SZ,'')

parentKey =
w.OpenKey(w.HKEY_LOCAL_MACHINE,r'Software\ODBC\ODBC.INI\tri2',0,w.KEY_WRITE)
newKey = w.CreateKey(parentKey,'Engines')

parentKey =
w.OpenKey(w.HKEY_LOCAL_MACHINE,r'Software\ODBC\ODBC.INI\tri2\Engines',0,w.KE
Y_WRITE)
newKey = w.CreateKey(parentKey,'Jet')
w.SetValueEx(newKey,'ImplicitCommitSync',0,w.REG_SZ,'')
w.SetValueEx(newKey,'MaxBufferSize',0,w.REG_DWORD,2048)
w.SetValueEx(newKey,'PageTimeout',0,w.REG_DWORD,5)
w.SetValueEx(newKey,'Threads',0,w.REG_DWORD,3)
w.SetValueEx(newKey,'UserCommitSync',0,w.REG_SZ,'Yes')

parentKey = w.OpenKey(w.HKEY_LOCAL_MACHINE,r'Software\ODBC\ODBC.INI\ODBC
Data Sources',0,w.KEY_WRITE)
w.SetValueEx(parentKey,'tri2',0,w.REG_SZ,'Microsoft Access Driver (*.mdb)')

Hope this helps someone else!

Scott




From kc5tja at garnet.armored.net  Fri Dec  8 15:48:20 2000
From: kc5tja at garnet.armored.net (Samuel A. Falvo II)
Date: 8 Dec 2000 12:48:20 -0800
Subject: Tkinter newbie question
Message-ID: <slrn932i2t.b96.kc5tja@garnet.armored.net>

I've set up a basic user interface using Tkinter.  One of the controls is a
listbox.  However, I cannot find any documentation on how to read the
current state of the listbox (e.g., what, if any, items are currently
selected).  I know it has to be possible -- otherwise, what's the point of
having the control?  :D

Thanks in advance for any assistance offered.

-- 
KC5TJA/6, DM13, QRP-L #1447 | Official Channel Saint, *Team Amiga*
Samuel A. Falvo II	    |
Oceanside, CA		    |


From joconnor at mail.cybermesa.com  Thu Dec  7 15:51:18 2000
From: joconnor at mail.cybermesa.com (Jay O'Connor)
Date: Thu, 07 Dec 2000 20:51:18 GMT
Subject: Declaring functions and variable name scope
Message-ID: <200012072051.NAA18468@mail.cybermesa.com>

An embedded and charset-unspecified text was scrubbed...
Name: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20001207/614e2a00/attachment.ksh>

From eff at my-deja.com  Wed Dec 20 06:09:27 2000
From: eff at my-deja.com (eff at my-deja.com)
Date: Wed, 20 Dec 2000 11:09:27 GMT
Subject: Need Help sorting alphabetically.
References: <3A3F1A91.A7ED9182@araidesign.com> <slrn93u8qc.5lh.scarblac@pino.selwerd.nl> <OlY%5.172268$U46.5446765@news1.sttls1.wa.home.com>
Message-ID: <91q415$nmm$1@nnrp1.deja.com>

Greg Jorgensen wrote:
> > It seems you want to ignore case and ignore everything outside a-z.
> > So make a function that throws out all of the other characters and
> > turns the rest into a single case, and then compares those:
> >
> > def compare_alphabetically(a, b):
> >    a = filter(lambda x: x.isalpha(), a.upper())
> >    b = filter(lambda x: x.isalpha(), b.upper())
> >    return cmp(a,b)
>
> The filter is unnecessary; the upper() and lower() methods of
> string only change alphabetic characters.

...but they don't "throw out all other characters".  the original
code matches his description, your proposed replacement doesn't.

(note that it's probably more efficient to use the translate
function/method instead of the filter trick, but that's another
story).

</F>


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


From ksaehwa at my-deja.com  Wed Dec 13 21:23:06 2000
From: ksaehwa at my-deja.com (ksaehwa at my-deja.com)
Date: Thu, 14 Dec 2000 02:23:06 GMT
Subject: Call for Papers: LCTES 2001 (ACM SIGPLAN Workshop)
Message-ID: <919au5$p0j$1@nnrp1.deja.com>

========================================================================

                            Call for Papers

                     ACM SIGPLAN 2001 Workshop on
      Languages, Compilers, and Tools for Embedded Systems (LCTES'2001)
                  Snowbird, Utah, June 22 - 23, 2001
                   (in conjunction with PLDI'2001)

                  http://redwood.snu.ac.kr/lctes2001/

As the complexity of the application domains for the embedded systems
increases, new research challenges are encountered for achieving
desired level of performance goals (which could be diverse such as
speed, real time constraints, code size, power, and scalability).
Due to limited resources locally available, embedded systems are
increasingly used in a networked (wired, wireless, as well as mobile)
environment. In addition, advances in hardware reflect into solving
issues using a combination of hardware and software techniques.
The aim of LCTES is to provide a forum for discussing the latest
research related to the above trends and to allow researchers and
developers working on different aspects to get together and synergize.
Original submissions are invited in all areas relevant to this theme.
Appropriate topics include, but are not limited to the following
aspects of embedded systems:

* Programming Languages and Optimization for Embedded Systems
 - real-time and embedded Java
 - exception and interrupt handling for embedded systems
 - memory management and garbage collection
 - compiler optimizations w.r.t. performance, code size, and power
   consumption
 - dynamic optimizations involving code slicing, binary-to-binary
   translation for re-targetting optimized code
 - program analysis (abstract interpretation, compiler transformation,
   static analysis)
 - code selection, register allocation and instruction scheduling for
   irregular architectures
 - languages and compilers for digital signal processors
* System Software for Embedded Systems
 - embedded real-time operating systems
 - distributed embedded systems
 - embedded middleware
 - embedded Internet
 - home networking (HomeRF, Bluetooth, Powerline Communications, etc.)
 - information appliances
* Design and Analysis Tools for Embedded Systems
 - formal specification and verification
 - object-oriented modeling and design
 - integration and testing
 - profiling, measurement, and debugging
 - scheduling and timing analysis for real-time systems
 - development tools and environments
* Hardware/Software Codesign
 - configurable processors (memories, data paths, register files, etc.)
 - processor architecture tradeoffs for code density, size and power
 - network processors and routers
 - wireless hub processors
 - retargetable development tools/compiler design automation
* Standardization
 - inter-operability and systems design
 - legacy issues and re-engineering

Papers should report new research and should not exceed 6000 words
(approximately 10 typeset on 16-point spacing), including figures and
references (11 pages). Short papers that describe existing
implementations or work-in-progress, or outline new problems or
important issues are also welcome. Short papers should not exceed 3000
words (6 pages). All accepted papers will be presented at the workshop
and published in the proceedings, which will be distributed at the
workshop. We anticipate that the proceedings will be published as an
issue of the ACM SIGPLAN Notices.

Electronic submissions must be received by 5:00 PM Eastern Standard
Time,Sunday, February 25, 2001. Interactive submissions through the
workshop's web page are encouraged. Submissions may also be sent as a
single e-mail message to either of the program co-chairs (MIME
attachments are allowed); the message should contain both the filled-
out form (see web page) and the Postscript paper. Electronic papers
should be in Postscript form, which must be interpretable by
Ghostscript. The Postscript must use standard fonts, or include the
necessary fonts, and must be prepared for USLetter (8.5"x11") or A4
page sizes. Authors who cannot meet these requirements should submit 15
hard copies by post to either of the program co-chairs by airmail that
must be received on or before February 25, 2001. These are firm
constraints; submissions not meeting the criteria described above
will not be considered.

========================================================================
Important Dates

* papers due:  Feb. 25, 2001 (5pm EST)
* author notification: Apr. 15, 2001
* final papers due: May. 19, 2001

========================================================================
Co-Chairs

Seongsoo Hong                           Santosh Pande
School of Electrical Engineering        College of Computing
Seoul National University               801 Atlantic Drive
Seoul 151-742                           Georgia Institute of Technology
KOREA                                   Atlanta, GA 30332
                                        USA
phone: +82-2-880-8370                   phone: +1-404-385-2169,
fax: 882-4656                           fax: 385-2295
sshong at redwood.snu.ac.kr                santosh at cc.gatech.edu

========================================================================
Tentative Program Committee

* Guido Araujo, University of Campinas, Brazil
* David August, Princeton University, USA
* Kiyoung Choi, Seoul National University, Korea
* Keith D. Cooper, Rice University, USA
* Srinivas Devadas, MIT, USA
* Helen Gill, NSF/DARPA, USA
* Rajiv Gupta, University of Arizona, USA
* Mary Hall, USC-ISI, USA
* Mahmut Kandemir, Penn State University, USA
* Rainer Leupers, University of Dortmund, Germany
* David Levine, Analog Devices Inc., USA
* Annie Liu, State University of New York, USA
* Sang Lyul Min, Seoul National University, Korea
* Soo-Mook Moon, Seoul National University, Korea
* Frank Mueller, Humboldt University Berlin, Germany
* Tatsuo Nakajima, Waseda University, Japan
* Alex Nicolau. University of California, Irvine, USA
* Carl Von Platen, IAR Systems, Sweden
* J. Ramanujam, Louisiana State University, USA
* Manas Saksena, TimeSys, USA
* Reid Tatge, Texas Instruments, USA
* Gang-Ryung Uh, Lucent Technologies, USA
* David Whalley, Florida State University, USA

=======================================================================
Steering Committee

* Azer Bestavros, Boston University, USA
* Jack Davidson, University of Virginia, USA
* Richard Gerber, University of Maryland, USA
* Rajiv Gupta, University of Arizona, USA
* Annie Liu, Indiana University, USA
* Thomas Marlowe, Seton Hall University, USA
* Sang Lyul, Min Seoul National University, Korea
* Frank Mueller, Humboldt University Berlin, Germany
* Bill Pugh, University of Maryland, USA
* David Whalley, Florida State University, USA
* Reinhard Wilhelm, University of the Saarland, Germany


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


From aahz at panix.com  Sun Dec 10 20:57:24 2000
From: aahz at panix.com (Aahz Maruch)
Date: 10 Dec 2000 17:57:24 -0800
Subject: Programming Habits in Python
References: <cspU5.82459$U46.2617014@news1.sttls1.wa.home.com> <udn1e8k0qi.fsf@box.home.de> <slrn937caf.h4.amk@207-172-146-162.s162.tnt3.ann.va.dialup.rcn.com> <9116s7$5ts$1@nnrp1.deja.com>
Message-ID: <911ca4$aqo$1@panix6.panix.com>

In article <9116s7$5ts$1 at nnrp1.deja.com>,  <lobozc at my-deja.com> wrote:
>
>So, I agree to most programmers nowadays don't need Knuth's recipes.

I think you're missing Andrew's point: most programmers *do* need
Knuth's recipes (and the theory, too!), but packaged in a way that's
more accessible than the Mix language.
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

What if there were no rhetorical questions?  --Aahz


From vsbabu at erols.com  Thu Dec  7 16:51:11 2000
From: vsbabu at erols.com (Satheesh Babu)
Date: Thu, 7 Dec 2000 16:51:11 -0500
Subject: DCOracle Need Help
References: <t2v9fu2pht2d69@corp.supernews.com> <slrn92vfkd.ut.gerhard.nospam@lilith.hqd-internal.de> <t2vm6g9u2qhle6@corp.supernews.com>
Message-ID: <90p0nv$rrj$1@bob.news.rcn.net>

If you want to download compiled version for windows
, you can download it from
http://vsbabu.csoft.net/ZOracleDA-bin.zip

It works for my Zope installation under NT 4.0, with
Oracle 8.1.6 client libraries.

Your mileage on how well you can use it with Python
will vary :-)

"Jose Martin" <joseamartin at venmex.com> wrote in message
news:t2vm6g9u2qhle6 at corp.supernews.com...
> no, i dont
> can you help me with this ?
>
> Gerhard H?ring <gerhard.nospam at bigfoot.de> escribi? en el mensaje de
> noticias slrn92vfkd.ut.gerhard.nospam at lilith.hqd-internal.de...
> > The Buffer module isn't a pure Python module. It's an extension module
> > written in C. Have you read the installation instructions and compiled
the
> C
> > module?
> >
> > Gerhard H?ring
> >
> > On Thu, 7 Dec 2000 10:48:32 -0400, Jose Martin <joseamartin at venmex.com>
> wrote:
> > >I am trying to use DCOracle on Python 2.0 under Windows or Linux
> > >but when i try to make my scritp :
> > >
> > >The first line is
> > >from DCOracle import *
> > >
> > >when i try to run:
> > >Python say me:
> > >
> > >File "d:\python20\DCOracle\__init__.py", line 52, in ?
> > >    import sys, Buffer
> > >ImportError: No module named Buffer
> > >
> > >I was searched my python path for a module called Buffer, but i dont
have
> it
> > >Whats happennig ?
> >
> > --
> > mail:   gerhard <at> bigfoot <dot> de
> > web:    http://highqualdev.com
>
>




From thomas at xs4all.net  Sun Dec 24 06:42:52 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Sun, 24 Dec 2000 12:42:52 +0100
Subject: Some Python 2.1 ideas
In-Reply-To: <3A45D358.CBEC055A@schneider-kamp.de>; from peter@schneider-kamp.de on Sun, Dec 24, 2000 at 11:43:36AM +0100
References: <005f01c06d01$d8114820$74eb0b18@C322162A> <20001223212159.H5515@xs4all.nl> <3A45D358.CBEC055A@schneider-kamp.de>
Message-ID: <20001224124251.N5483@xs4all.nl>

On Sun, Dec 24, 2000 at 11:43:36AM +0100, Peter Schneider-Kamp wrote:

> Can you detail the 'for i in <x> .. <y>' proposal?

It's just syntactic sugar (with an opening for some optimization internally)
for

  for i in range(<x>, <y>)

or, *perhaps*,

  for i in range(<x>, <y>+1)

Guido and usability studies would have to determine which is best :-)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From yusuf.luele at bov.de  Tue Dec 19 05:37:30 2000
From: yusuf.luele at bov.de (Yusuf Lüle)
Date: Tue, 19 Dec 2000 11:37:30 +0100
Subject: problem with pickle.loads
Message-ID: <91ndqv$o7$1@news10.roka.net>

Hello,
I have a problem with serializing an object.
In a python script I create an object (an instance of a class) and call the
pickle.dumps function
(serobj = pickle.dumps(obj)).
Then I put this serialized result (serobj) in an html page in an hidden
input box and sent it  to the next python script.
The second python script now reads the serialzed object from the environment
of the webserver.
(I use os.environ["QUERY_STRING"], parsing and decoding the part I need).
After that I try to unserialize the result with the function obj =
pickle.loads(serobj).
Python throws an error:
 File "second.py", line 58, in ?
    obj= pickle.loads(serobj)
  File "/usr/lib/python1.5/pickle.py", line 830, in loads
    return Unpickler(file).load()
  File "/usr/lib/python1.5/pickle.py", line 495, in load
    dispatch[key](self)
  File "/usr/lib/python1.5/pickle.py", line 612, in load_inst
    klass = self.find_class(module, name)
  File "/usr/lib/python1.5/pickle.py", line 669, in find_class
    raise SystemError, \

Can someone help me?






From tim.one at home.com  Sun Dec 31 15:38:53 2000
From: tim.one at home.com (Tim Peters)
Date: Sun, 31 Dec 2000 15:38:53 -0500
Subject: Interesting behaviour of the assignment
In-Reply-To: <20001231153143.86832A84F@darjeeling.zadka.site.co.il>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEBCIGAA.tim.one@home.com>

[Moshe Zadka]
> Here's an example where I'd rather use "is" then "==":
>
> # module foo.py
> SPAM = 0
> EGGS = 1
>
> def eat(what):
> 	if what is SPAM:
> 		eat_spam()
> 	if what is EGGS:
> 		eat_eggs()
> 	raise FoodError()
>
> # user code:
> import foo
>
> foo.eat(foo.SPAM)
> foo.eat(foo.EGGS)
>
> The semantics are accurate here: foo.eat should not be passed 0.0,
> it should be passed only foo.SPAM or foo.EGGS, so a FoodError
> will alert the programmer to his bug sooner.

Yet because Python stores little ints uniquely, they will not get a
FoodError if they happen to pass in an int 0 or 1 however arrived at (e.g.,
foo.eat(int("0"))).

You can make this robust like so:

class FoodType:
    pass

SPAM = FoodType()
EGGS = FoodType()

Now "is" is bulletproof.  A stranger way:

SPAM = []
EGGS = []

Any mutable object type will do; any immutable object type may or may not
work today, and whichever obtains today may change in the next release.

god-is-one-because-god-is-immutable<wink>-ly y'rs  - tim




From fgranger at teleprosoft.com  Thu Dec  7 09:55:22 2000
From: fgranger at teleprosoft.com (Fran=?ISO-8859-1?B?5w==?=ois Granger)
Date: Thu, 07 Dec 2000 15:55:22 +0100
Subject: Help/example of bit testing
References: <90llq7$8fc@dispatch.concentric.net>
Message-ID: <B6556369.1DF7A%fgranger@teleprosoft.com>

in article 90llq7$8fc at dispatch.concentric.net, Bryan Webb at
bww00 at amdahl.com wrote on 6/12/00 16:25:

> Hi,
> I am reading a file (binary)of 1024 char recs (bitmaps) . I can read the
> file , bit im having trouble finding examples of testing bits. I need to
> look at each bit of the above record.
> Looking for examples and ideas.

There is a summary here, which need improvement.

http://francois.granger.free.fr/MacPython/faq/MacPyBeg-Bitwise.html

quick and dirty

genMasq = [1, 2, 4, 8, 16, 32, 64, 128, 256, 512, 1024, 2048, 4096, 8192,
16384, 32768]

data = 198

for i in range(7):
    if data & genMask[i]:
        print "bit" + str(i) + "is on"
-- 
Fran?ois Granger
fgranger at altern.org



From oivvio at cajal.mbb.ki.se  Tue Dec  5 09:49:46 2000
From: oivvio at cajal.mbb.ki.se (Oivvio Polite)
Date: Tue, 5 Dec 2000 15:49:46 +0100
Subject: debugging python extensions in c
Message-ID: <90iv44$mho$1@news.kth.se>

Hi everyone.

We've just started writing a python extension in c.

I know how to use the python debugger and I've  done some c debugging
before.

But how do I go about to debug pointerstuff and memory leaks in python
extension.

We're writing this for both windows and linux.

regards oivvio





From aleaxit at yahoo.com  Fri Dec 15 07:58:25 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 15 Dec 2000 13:58:25 +0100
Subject: xreadlines (was Re: while true: !!!)
References: <8uthf2$pjh$04$1@news.t-online.com> <8utnqk02rlq@news1.newsguy.com> <8uu44b$4nc$04$1@news.t-online.com> <ur94ctqkw.fsf@ctwd0143.fitlinxx.com> <3A34CEA7.CD9D85AE@yahoo.com> <3A36368C.26AADDDE@yahoo.com> <slrn93g37o.9tj.jepler@potty.housenet> <f3%Z5.146409$U46.4768601@news1.sttls1.wa.home.com> <91a6it02flt@news2.newsguy.com> <slrn93jagg.fl.neelk@alum.mit.edu>
Message-ID: <91d4ii0lmq@news1.newsguy.com>

"Neelakantan Krishnaswami" <neelk at alum.mit.edu> wrote in message
news:slrn93jagg.fl.neelk at alum.mit.edu...
> On Thu, 14 Dec 2000 11:14:38 +0100, Alex Martelli <aleaxit at yahoo.com>
wrote:
> >> stdin. I've used fileinput to go through big lists of files (10,000+
email
> >> messages) and it works great. It doesn't appear to do any buffering
> >> itself--it uses file.readline() to read the files.
> >
> > If this was a performance problem, it could of course also be fixed
> > in a future fileinput version without changing code that uses it (again,
> > in-place-rewriting would probably have to inhibit the optimization,
> > although that isn't entirely clear).
>
> While it's true that fileinput is somewhat slow, but it can easily be
> made faster than the usual while 1: loop everyone uses.(Relatively

Yep, exactly my point -- and the chunking of readlines was exactly
what I had in mind here as the performance-fix... thanks for doing
the actual work, which I lazily skipped!  With a larger buffer and
more streamlined __getitem__ (no error-checking, optimization for
the most-frequent case) I can get to roughly 1/2 of readlines()...:

class LinesOf:
    def __init__(self, file, chunkSize=256*1024):
        self.file = file
        self.chunkSize = chunkSize
        self.start = 0
        self.refill()
    def refill(self):
        self.data = self.file.readlines(self.chunkSize)
    def __getitem__(self, i):
        try: return self.data[i-self.start]
        except IndexError:
            self.start += len(self.data)
            self.refill()
            if not self.data: raise IndexError
            return self.data[i-self.start]

import time

def withReadlines(file):
    start = time.clock()
    i = 0
    bytes = 0
    for line in file.readlines():
        #i+=1
        #bytes+=len(line)
        pass
    stend = time.clock()
    return i, bytes, stend-start

def withLinesOf(file):
    start = time.clock()
    i = 0
    bytes = 0
    for line in LinesOf(file):
        #i+=1
        #bytes+=len(line)
        pass
    stend = time.clock()
    return i, bytes, stend-start

def test(filename):
    file=open(filename)
    print withReadlines(file)
    file.close()

    file=open(filename)
    print withLinesOf(file)
    file.close()

if __name__=='__main__':
    import sys
    try: filename = sys.argv[1]
    except IndexError: filename = 'aaa.py'
    test(filename)

The operations in the for-loops are commented out to ensure
we're not timing them too -- decommenting them helps ensure
that LinesOf is actually working, and gives an idea of the
magnitude of the test:

D:\PySym>python aaa.py \winnt\profiles\martelli\personal\findin~1.htm
(6072, 444748, 0.075308712333910496)
(6072, 444748, 0.14522679691782142)

the 'Finding of Facts' HTML file from Microsoft's antitrust
cause -- over the chunksize, to ensure refilling is exercised.

OK, without the operations in the loop, we have:

D:\PySym>python aaa.py \winnt\profiles\martelli\personal\findin~1.htm
(0, 0, 0.052779039576527305)
(0, 0, 0.10314101285470281)

D:\PySym>python aaa.py \winnt\profiles\martelli\personal\findin~1.htm
(0, 0, 0.052812563380942722)
(0, 0, 0.1092364785925366)

Best to run it twice to ensure against cache effects -- first time
I ran it, LinesOf appeared *FASTER*... because it's run _after_
readlines, so it benefited from OS caching of the file!-).


Alex





From jcm at bigskytel.com  Tue Dec  5 21:30:12 2000
From: jcm at bigskytel.com (David Porter)
Date: Tue, 5 Dec 2000 19:30:12 -0700
Subject: Python Origins
In-Reply-To: <90k3pq$50p$1@nnrp1.deja.com>; from nigeljennison@cs.com on Wed, Dec 06, 2000 at 01:12:30AM +0000
References: <90k3pq$50p$1@nnrp1.deja.com>
Message-ID: <20001205193011.A14957@bigskytel.com>

* Nigel <nigeljennison at cs.com>:
>  I am doing some research and I would be grateful if anybody could give
> me some history on the origins of Python.

A short history straight from Guido:

http://www.pythonlabs.com/doc/essays/foreword.html


David
  



From glenfant at equod.com.nospam  Tue Dec  5 05:01:31 2000
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Tue, 5 Dec 2000 11:01:31 +0100
Subject: Using dictionaries
References: <3a2c9264.503856587@news-server.bigpond.net.au>
Message-ID: <90ie5q$ghp$1@reader1.imaginet.fr>

Hi,

Can't understand why you're doing complicated when something simple can
work.
You don't need to use a dictionnary to do this.
If a simple way like this does not work for you, there's something I did not
get from your post.
--- configuration.py (start)---
...
Val1 = 1
Val2 = 'myserver'
Valn = 500
...
--- configuration.py (end)---
--- main.py (start) ---
...
from configuration import *
...
xyz = Val1 + Valn
...
--- main.py (end) ---

"sorular" <sorular at netscape.net> a ?crit dans le message news:
3a2c9264.503856587 at news-server.bigpond.net.au...
> G'day
>
> I am trying to write a python program where a "configuration.py" will
> have all the user entered values in a dictionary format and a
> "main.py" which will import these and use it as local parameters.
>
> But, there are quite a few values in the dictionary
> ie
> dict = { 'key1':'value1' , ' key2':'value2' , ......'keyN':'value2'}
>
> and in the "main.py" these will be all renamed with the local names
> ie
> Val1 = dictionary.dict['key1']
> Val2 = dictionary.dict['key2']
> ....
> ValN = dictionary.dict['keyN']
> etc....
>
> But overall, I still have a large numbe of values (ValN) to list... !
>
> Q1- Is there a more efficient way to do this ? I thought about using a
> for loop but that didnt work....
>
> Q2 - would it be possible only to list the parameters in first file
> (ie dictionary.py
> key1 = value1
> key2 = value2
> .....
> keyN = valueN
> )
> and inside main.py, do a simple import...
> import dictionary *
> so that key1 can be used inside main.py as a local parameter?
>
> thank
>
> HK



From pbjorn at uswest.net  Thu Dec 28 15:48:35 2000
From: pbjorn at uswest.net (Bjorn Pettersen)
Date: Thu, 28 Dec 2000 13:48:35 -0700
Subject: Some Python 2.1 ideas
References: <3A452881.7686E717@uswest.net> <%vb16.187233$U46.5801214@news1.sttls1.wa.home.com> <3A464D5D.B541C04E@uswest.net> <mailman.977688488.25438.python-list@python.org> <20001225150250.6FA06A84F@darjeeling.zadka.site.co.il> <mailman.977760789.10224.python-list@python.org> <roy-EA8C3B.11361325122000@news.panix.com> <3A479C29.2631F27B@uswest.net> <mailman.977876470.23568.python-list@python.org> <3A4A50E6.11BBE0FD@uswest.net> <mailman.977962871.16241.python-list@python.org> <3A4AB05C.D41F4022@uswest.net> <92ftif$17fc$1@nntp6.u.washington.edu> <3A4B8865.86712FA1@uswest.net> <92g4sj$c62i$1@nntp6.u.washington.edu>
Message-ID: <3A4BA722.9DF7EED6@uswest.net>

Donn Cave wrote:

> Quoth Bjorn Pettersen <pbjorn at uswest.net>:
> | Donn Cave wrote:
> |> Quoth Bjorn Pettersen <pbjorn at uswest.net>:
> |> D-Man wrote:
> |>|> Yeah, it is a problem if you don't know where the file came from.  How
> |>|> about creating a standard text file format?  All files could be opened
> |>|> in binary mode on all platforms, then EOL is handled by each
> |>|> application (or better yet make a standard module that does the
> |>|> implementation correctly once).
> |>|
> |>| Feeling like fighting windmills today, eh <wink>? I think the current wish is to
> |>| do it from the other end, i.e. instead of relying on the text file format to be
> |>| "correct", implement a method on string object that is smart enough to handle all
> |>| text file format line endings intelligently.
> |>
> |> Without the context of the I/O stream, to decide whether '\r' is
> |> data or separator, isn't that string function going to be the opposite
> |> of smart?  How does a string handle '\r' intelligently?
> |
> | By assuming that if it's the last character in the string, it came from a mac file and
> | is the line terminator. Similarly for '\r\n' (dos) and '\n' (unix) when at the end of
> | the string.
> |
> | It might be cute to have a MacTextFile class etc. that ensures that the line endings
> | are consistent, but personally I think that is overkill (and is likely to be very
> | slow).
>
> I guess a function like that would often be the right thing.  But
> it doesn't strike me as what I would have called intelligent handling,
> just aggressive in its assumptions.  In a situation where the application
> calls for an aggressive normalization of the data, it's obviously the
> right thing.  Otherwise, it's useless, where a per I/O stream solution
> might still be useful, and I guess cute too.

I guess we're in violent agreement <wink>.  I provided a patch to implement chomp(), so I'll
leave it to you to implement a  patch for the other solution ;-)

-- bjorn




From loewis at informatik.hu-berlin.de  Sat Dec  9 05:49:18 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 09 Dec 2000 11:49:18 +0100
Subject: The secrets of operator.add
References: <g08zpq51vy.fsf@scumbag.ecs.soton.ac.uk> <90r69r02hmn@news2.newsguy.com> <90rt6p$3s1$1@news.nuri.net>
Message-ID: <p6qvgstamz5.fsf@informatik.hu-berlin.de>

"June Kim" <junaftnoon at nospamplzyahoo.com> writes:

> How come the huge difference in its total running time between the two?

My guess it's because of the lambda function: That will give you a
number of extra functions on the C stack, and in particular another
Python ceval invocation, along with a frame object and everything.

OTOH, operator.add is a builtin, which is invoked from reduce without
further ado.

Regards,
Martin


From amk at mira.erols.com  Wed Dec 20 21:30:45 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 21 Dec 2000 02:30:45 GMT
Subject: re different between US and German versions of NT
References: <mailman.977362144.17768.python-list@python.org>
Message-ID: <slrn942rjn.l1.amk@207-172-111-180.s180.tnt1.ann.va.dialup.rcn.com>

On Wed, 20 Dec 2000 20:35:47 -0500, Darrell Gallion <darrell at dorb.com> wrote:
>Found a re pattern that behaved differently on NT4 (US version) than it did
>on NT4 (German version)
>When I switched to use pre instead of sre it worked the same on both.

sre uses the C isalnum() function/macro at one point; isalnum() is
locale-aware, so presumably it's matching more characters on the
German version.  This could be argued to be a bug, or a feature.  I'm
not sure which it is, but would lean toward "bug", personally.

--amk


From joconnor at mail.cybermesa.com  Tue Dec 12 20:00:58 2000
From: joconnor at mail.cybermesa.com (Jay O'Connor)
Date: Wed, 13 Dec 2000 1:00:58 GMT
Subject: Spinning off time intensive procedures to speed CGI  (fwd)
Message-ID: <200012130100.SAA08673@mail.cybermesa.com>

An embedded and charset-unspecified text was scrubbed...
Name: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20001213/d878f08a/attachment.ksh>

From jerry_spicklemire at my-deja.com  Wed Dec 27 10:32:31 2000
From: jerry_spicklemire at my-deja.com (jerry_spicklemire at my-deja.com)
Date: Wed, 27 Dec 2000 15:32:31 GMT
Subject: Python vs PHP (& Java?)
References: <mailman.977897649.4877.python-list@python.org>
Message-ID: <92d22f$cnv$1@nnrp1.deja.com>

 Ben Ocean observed:

> I?m an intermediate-level programmer that?s new to Python (via Zope)
> and I like it a lot. However, it appears to me that, while Python is
> a good replacement for Perl, it cannot replace much of the
> functionality of  PHP, which can be scripted into HTML (whereas
> Python, it appears, must be called). So, I have to wonder about the
> speed of Python vs. PHP for Internet-based applications. Python has
> the advantage of byte-code and compilation, but it?s server-side. PHP
> might not be as fast on the server, but it doesn?t have to make as
> many trips back and forth between server and  client. Plus, I?m
> curious as to why Python wasn?t written in such a way that it could
> be written directly into HTML documents like PHP.

One thing at a time:

 Python packages for embedding scripts in HTML -

  ASPy,
http://archive.dstc.edu.au/aspy/

  Quixote,
http://starbase.neosoft.com/~claird/comp.lang.python/web_python.html#Qui
xote

  WebWare,
http://webware.sourceforge.net/


Any of these will allow you to do PHP/ASP-ish things with Python. On
the other hand, you've already gotten to know Zope, to some degree,
which is more powerful by far for typical Web development, so I'm a
little confused.

It sounds like you may have a basic misunderstanding of the stateless
nature of HTTP. When a (non-Javascript) script is embedded in HTML, the
processing that it invokes actually happens at the server, before the
client ever knows what's up. That's why you see the results of the
script, rather than the original code, when you view the source of the
HTML page. So, whether a PHP script is called, or a Python script,
it's the server side pre-processor that does the work. Javascript is
(mostly) different, since it is interpretted by the client (browser
with Javascript interpretter embedded). The "mostly" qualifier refers
to the real, but unpopular, option of using server side Javascript.
Yes, you can do it, but if you're working server side anyway, why
choose a language as limiting as Javascript?

I haven't heard of any browsers with PHP interpretters embedded, or a
client side PHP plugin, and until such a beast exists, the fundamentals
are unavoidable. But that's not bad. Client side processing is fraught
with compatibility nightmares, and even Java couldn't make much headway
until it moved decisively toward the server side.

> On a similar subject, someone once informed me that Java was by far
> the best language for scripting communication programs over the
> Internet (like chat, IRC, etc.).

Java is pretty good, has lot's of mindshare, and third party support,
no denying any of that. The Network Effect plays into this scenario, in
that one can feel reasonably comfortable with a Java based system for
any of the above applications.

Still, there are many very easy to use Python modules for such
functions, and if you are truly scripting InterNet Communications,
you're dealing with standard protocols. It doesn't matter what language
either end is implemented in. Java based clients can talk to Python
based servers, and vice versa.

The real question is, if you know both Python and Java, which one will
you use when either one will work. I'm very happy to take the word of
all the former Java hackers on this list who sing Python's praises for
many important reasons, productivity, flexibility, and simplicity being
only the most obvious.


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


From aleaxit at yahoo.com  Tue Dec 19 11:50:20 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 19 Dec 2000 17:50:20 +0100
Subject: while true: !!!
References: <B43D149A9AB2D411971300B0D03D7E8B0FBCBB@natasha.auslabs.avaya.com> <v04220801b663769e1800@[192.168.11.33]> <mailman.977147642.18708.python-list@python.org> <91lclt02qr3@news1.newsguy.com> <slrn93st0m.6c1.qrczak@qrnik.zagroda>
Message-ID: <91o3l501i59@news2.newsguy.com>

"Marcin 'Qrczak' Kowalczyk" <qrczak at knm.org.pl> wrote in message
news:slrn93st0m.6c1.qrczak at qrnik.zagroda...
> Mon, 18 Dec 2000 17:05:51 +0100, Alex Martelli <aleaxit at yahoo.com> pisze:
>
> > I think it would be inconsistent, un-Pythonic, to add a
> > fundamental type for 'conceptual' reasons, without any
> > practical, measurable benefit.  For example, Python does
> > away with the distinction between 'a character' and 'a
> > string of length 1' -- a distinction that is religiously
> > maintained in C, Pascal, Java, etc,
>
> OTOH Python does not equate characters with integers. C, C++ and Java
> do.

Wrong!  C's types are always slightly fuzzy, but that does
NOT apply to C++ or Java; e.g., in C++:

void f(int) {
    cout << 'integer\n';
}

void f(char) {
    cout << 'character\n';
}

There is no 'equate' between the char[acter] and the int[eger]
case: if you call
    f('7')
you will see 'character' on standard output, while if you
call
    f(7)
you will see 'integer'.

A character can be _implicitly *coerced*_ to other types,
but that is very different from 'equating': the distinction
is perfectly sharp, just as, say, between a char* and a
const char*, even if the first can be coerced to the
second just as implicitly.


Python does implicit coercion between number types, to
string in a few contexts (e.g., a print statement), and
between sequence types in a few other contexts -- and
that's about it, I believe.

> The difference is conceptual, not in the representation. I don't
> agree that having a separate boolean type would be inconsistent.

A C++ bool is just as coercible as a char, of course.  The
added value, just like for char, is in overloading and,
exactly, representation (an array of char, or a vector of
bool, can be specialized to consume much less space than
arrays/vectors of int or long would) -- issues just not
pertinent to Python (you can 'import array', but what types
can be packed in what fashion into an array.array is quite
a different matter from what types exist as Python typeobjects).

Java takes a different approach to boolean coercion (not
consistent with the approach to _char_ coercion, which it
roughly shares with C or C++) but basically what it gets
from it is on the same plane -- strictly connected to
compile-time checks and operations of a statically typed
language.  You can overload a method differently for
boolean vs int arguments, you can get compiler errors
for some uses of int which the Java designers wanted to
forbid -- that's about it.

Without static typing, Python would of course get none of
these benefits... while still paying all of the costs (in
a conceptual sphere first of all) that it blissfully AVOIDS
by eschewing unnecessary distinctions (just as it avoids
distinguishing between characters and single-character
strings).


> Scheme is dynamically typed and does have separate true and false
> values. Lisp does not, but equates Booleans with the empty list and
> symbol t rather than 0 and 1. Smalltalk is dynamically typed and has
> separate Booleans.

Smalltalk's Booleans are used for runtime polymorphism --
you send an object an 'if: block' message and it responds
appropriately (executing, or not executing, the block).
In an *everything*-is-an-object mindset, that's great --
one less 'statement', one more *message* whose responding
method you can tweak &c.  Python has a much more traditional
approach to statements, and would draw no benefit from
type(i<j) != type(2+2), as it's not going to start sending
'messages' in lieu of good old if and while statements.

As for Scheme, the Revised(4) report says:

"""
Any Scheme value can be used as a boolean value for the
purpose of a conditional test. As explained in section
section Booleans, all values count as true in such a test
except for #f. This report uses the word "true" to refer
to any Scheme value that counts as true, and the word
"false" to refer to #f.

Note: In some implementations the empty list also counts
as false instead of true.
"""

Revised(5) report introduces a Boolean 'type', but it
still basically follows the same rules except for
removing the mention about empty-list counting as false
'in some implementations'.


If any value can be used as boolean, what's the big
gain in also having an 'alleged' Boolean type which
just duplicates what can be done with any other
object?

Note that Scheme also distinguished (in R5) 'char?'
from 'string?'; as it uses this level of fine
grainedness, I guess it IS consistent in having
these impractical 'boolean?' as well.

Python does NOT draw distinctions unless there is
practical use -- no char/string dichotomy; therefore,
for consistency, no boolean/all-other-which-can-be-
USED-as-boolean-but-'ARE'-not-boolean one either.

Korzybski and Wittgenstein might, I believe, appreciate
Python's pragmatic stance on this -- as it happens,
these two giants in the 20th century's reflections
on language, mind, and reality, were both engineers
(and, in spirit if not in degrees, I like to think
of our BDFL as a fellow engineer too... though I
guess that, actually being a mathematician and a
computer scientist by degree, he may be offended!-).


> I think it is only C and languages drawing conventions from C that
> equate booleans with numbers, with "any non-zero" meaning "true"
> and the canonical "true" value of 1.

I'm not sure what you mean here, but, either way, this
just doesn't ring true.  For example, Java surely draws
a lot from C, yet doesn't apply implicit cohercion to
numbers in a Boolean context or vv; BCPL, which we can
be *SURE* didn't "draw any convention from C" (since it
predated it!), and Forth (been a long time since I last
used it, but...), use numbers as booleans.

But, sure, the ability to *avoid* drawing certain semantic
distinctions at inappropriately-low levels (e.g., language
level, OS-level, etc) -- character versus string of length
1, boolean versus integer, textfile versus datafile, etc --
is anything but universal... that's why some languages,
OS's, etc, are better than others.  It's far too easy, in
this as in many other conceptual tasks, to be entrapped by
what Korzybski denounced as 'the IS of identity', to fail
to grasp the right 'family resemblance' as Wittgenstein
might put it; to think, in other words, that since we
cannot truly say 'a IS b' in the IS-of-identity sense,
then our language/OS/etc MUST distinguish a and b into
different classifications.

The right conceptual simplifications may only be invented
a few times, with others copying them from the inventors
(perhaps Postscript from Forth, C from BCPL, etc).

But the right test to apply is an operational semantic
one... what *concrete advantages* would accrue to Python
programmers if type(2>1) != type(2+1), with the rest of
the language staying basically the same?  (If you're
arguing for a _completely_ different language, e.g. one
based on modeling success/failure as non-data a la
Snobol or Icon, or a statically-typed one, with such
things as overloads &c, then it's better to clarify that;
me, I'm discussing issues related to, at most, incremental
changes in Python, not a complete rework to turn it into
a C++, Icon, Dylan, Scheme, or Smalltalk clone:-).


Alex





From pblanchette at pixelsystems.com  Thu Dec 14 08:42:54 2000
From: pblanchette at pixelsystems.com (Patrick Blanchette)
Date: Thu, 14 Dec 2000 13:42:54 GMT
Subject: win32net.NetWkstaSetInfo - How to use it?
References: <3A378524.1090400@pixelsystems.com> <3A3804B3.6010305@ActiveState.com>
Message-ID: <3A38CE4F.7070903@pixelsystems.com>

     Yes I'm interested in your fix and I'M am using Python 2.0.  Thank you.

   I have another bug with the win32net.  It's with the SetFileSecurity 
function.  It works correctly except when I'm using it inside nested 
try-except clause.  I got a 'No more memory is available for security 
information updates.' message.

   Except these minor problems and some missing functions, your win32 
extension is great and it help me very much

Thanks

Mark Hammond wrote:

> Patrick Blanchette wrote:
> 
>> I,
>>   I'm trying to set the computer name of a WinNT workstation using 
>> the NetWkstaSetInfo and the NetWkstaSetInfofunctions.  When I call 
>> NetWkstaSetInfo, It always returns the errorno 998:Invalid access to 
>> memory location.
> 
> 
> My apologies - but it is a bug in win32net.  I have checked in the 
> fix.  If you are on Python 2.0, I could mail you a new .pyd that 
> should work.
> 
> Mark.
> 



From amk at mira.erols.com  Fri Dec 29 17:46:26 2000
From: amk at mira.erols.com (A.M. Kuchling)
Date: 29 Dec 2000 22:46:26 GMT
Subject: fcgi.py
References: <3a4a865c.1325844015@localhost> <3a4a90ab.1328483109@localhost> <lYy26.91$9Ow6.17498496@news.randori.com>
Message-ID: <slrn94q5ra.mb.amk@207-172-111-214.s214.tnt1.ann.va.dialup.rcn.com>

On Wed, 27 Dec 2000 20:03:10 -0800, Robin Dunn <robin at alldunn.com> wrote:
>On the other hand if you don't mind losing the multiple process capability,
>fcgi.py could probably be tweaked to create its own listening socket instead
>of expecting it to already be there.  It could then be started up

Any plans to work on a fcgi.py port to Windows?  One possibility
mentioned in PEP 222 is adding fcgi.py to the standard library, but
lack of Windows compatibility would be an argument against it.

--amk


From mhagger at alum.mit.edu  Thu Dec 21 20:17:12 2000
From: mhagger at alum.mit.edu (Michael Haggerty)
Date: Fri, 22 Dec 2000 01:17:12 GMT
Subject: mail processing module?
References: <91rcig$rgu$1@nnrp1.deja.com>
Message-ID: <m266kd45m0.fsf@freak.kaiserty.com>

Preston Landers <prestonlanders at my-deja.com> writes:
> I'm trying to find out if there's any interest in a kitchen-sink
> Python module for mail handling.
> 
> This module lets you send mail (using smtplib) and retrieve mail
> from either a POP3 or IMAP4 server.

This could be very handy.

Is there a reason to put sending and retrieving into a single module?
They seem easily factorable.

In fact maybe a better breakdown would be one module that represents
an email message and allows data to be MIMEd into and out of it, one
module for sending such messages via SMTP, one module for retrieving
such messages via POP, and one for retrieving via IMAP.  The
interfaces for the two retrieving modules should be as close as
possible, and ideally the interface to the mail-sending routine should
be generic enough that one could write a module with the same
interface but that sends emails by writing them to, e.g., sendmail or
qmail.

(Some of this functionality might overlap with existing Python
modules.)

Michael

-- 
Michael Haggerty
mhagger at alum.mit.edu


From Blair.Lowe at compeng.net  Thu Dec 21 18:34:02 2000
From: Blair.Lowe at compeng.net (Blair Lowe)
Date: Thu, 21 Dec 2000 16:34:02 -0700
Subject: compiling BeOpen-Python-2.0-1.src.rpm
Message-ID: <v04220809b66843de8ce6@[192.168.11.33]>

Hi,

We are using Stock RedHat 6.2. We want to re-configure the 
BeOpen-Python source code for SSL, and general use.

We were compiling the BeOpen-Python-2.0-1.src.rpm with the command:

rpm -bi BeOpen-Python-2.0.spec

, and got the following error:


gcc -fpic  -I/usr/local/include/xmlparse -g -O2 -Wall 
-Wstrict-prototypes -I./..
/Include -I.. -DHAVE_CONFIG_H -c ./pyexpat.c
./pyexpat.c: In function `newxmlparseobject':
./pyexpat.c:639: warning: implicit declaration of function `XML_ParserCreateNS'
./pyexpat.c:639: warning: assignment makes pointer from integer without a cast
./pyexpat.c: In function `pyxml_SetStartNamespaceDeclHandler':
./pyexpat.c:1024: `XML_SetNamespaceDeclHandler' undeclared (first use 
in this fu
nction)
./pyexpat.c:1024: (Each undeclared identifier is reported only once
./pyexpat.c:1024: for each function it appears in.)
./pyexpat.c: In function `pyxml_SetEndNamespaceDeclHandler':
./pyexpat.c:1032: `XML_SetNamespaceDeclHandler' undeclared (first use 
in this fu
nction)
./pyexpat.c: In function `pyxml_SetStartCdataSection':
./pyexpat.c:1040: `XML_SetCdataSectionHandler' undeclared (first use 
in this fun
ction)
./pyexpat.c: In function `pyxml_SetEndCdataSection':
./pyexpat.c:1048: `XML_SetCdataSectionHandler' undeclared (first use 
in this fun
ction)
./pyexpat.c: At top level:
./pyexpat.c:1077: `XML_SetCommentHandler' undeclared here (not in a function)
./pyexpat.c:1077: initializer element for `handler_info[8].setter' is 
not consta
nt
./pyexpat.c:1089: `XML_SetDefaultHandlerExpand' undeclared here (not 
in a functi
on)
./pyexpat.c:1089: initializer element for `handler_info[12].setter' 
is not const
ant
./pyexpat.c:1092: `XML_SetNotStandaloneHandler' undeclared here (not 
in a functi
on)
./pyexpat.c:1092: initializer element for `handler_info[13].setter' 
is not const
ant
make[1]: *** [pyexpat.o] Error 1
make[1]: Leaving directory `/home/blade/src/BUILD/Python-2.0/Modules'
make: *** [sharedmods] Error 2
Bad exit status from /var/tmp/rpm-tmp.8833 (%build)



Why doesn't this compile cleanly? Does it need a patch for RH6.2? 
Seems like an unlikely thing to do.

Thanks,
Blair Lowe
Computer Engineering Inc. http://www.compeng.net
Phone: 780 499 5687 (9 - 5 MST) Fax:   780 435 0693 (24 Hours)

-- 
http://www.python.org/mailman/listinfo/python-list
Computer Engineering Inc. http://www.compeng.net
Phone: 780 499 5687 (9 - 5 MST) Fax:   780 435 0693 (24 Hours)
All email advice that is provided for free is without warrantee: use 
at your own risk. 



From spenrose at well.com  Thu Dec 14 00:55:24 2000
From: spenrose at well.com (Sam Penrose)
Date: Wed, 13 Dec 2000 21:55:24 -0800
Subject: Spinning off time intensive procedures to speed CGI
References: <916bpn$a6i$1@nnrp1.deja.com>
Message-ID: <spenrose-A548F5.21552413122000@news.dnai.com>

In article <916bpn$a6i$1 at nnrp1.deja.com>, leejor at my-deja.com wrote:

> I have a variety of time consuming proceedures that slow
> the response time of my CGI scripts. These are
> proceedures that do not send any information back to the
> web browser. I want to spin off these proceedures to
> allow the original program to imediately continue.
> Examples of these proceedures include: saving to Log
> files, sending emails, or caching database queries.
> 
> How do I accomplish this? I have played with fork() and
> os.system(), but have not achieved the results I need. In
> my searches of the documentation and online forums, I
> have only found methods that _wait_ for the spun off
> proceedures to return a result.
> 
> I don't want to wait.

Hhmm. I find that large (>10 modules, 1000's of lines of Python)
CGI-based systems perform quite reponsively on Intel hardware
while sending emails, saving file uploads, writing out text files,
etc. At least insofar as "responsive" means "load faster than the
New York Times home page over a T1."

If it's taking you, say, 20 seconds to send an email on more than
$200 worth of server under modest load, something other than
CGI is the problem.


From erno-news at erno.iki.fi  Mon Dec 11 14:52:27 2000
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 11 Dec 2000 21:52:27 +0200
Subject: Multiple Threads & IRC
References: <3A3415E0.1080607@bellsouth.net> <3A34E448.8B42C1BC@stud.cs.uit.no>
Message-ID: <ku8zpm91ms.fsf@lasipalatsi.fi>

In article <3A34E448.8B42C1BC at stud.cs.uit.no>, Andr? Larsen Risnes
<andrer at stud.cs.uit.no> writes:

| RFC 1459 is the document that describes the IRC communication
| protocol (which, btw, not all irc servers comply to :-( ).

well, it is from 1993. newer version is in rfc2810-rfc2813.

  -- erno


From rvprasad at cis.ksu.edu  Sun Dec 10 12:57:08 2000
From: rvprasad at cis.ksu.edu (Venkatesh Prasad Ranganath)
Date: 10 Dec 2000 11:57:08 -0600
Subject: Related to os module
Message-ID: <m3k898yxaj.fsf@boss.dreamsoft.com>

Hi,

Following is the code in question

>>>import os
>>>os.execv("/bin/ls", ["-l"])

The output should be that of executing "ls -l" instead it is of "ls".  "-l" is
being discarded.  What might be wrong?

-- 
Venkatesh Prasad Ranganath


From jkahanpa at pcu.helsinki.fi.invalid  Wed Dec  6 10:24:00 2000
From: jkahanpa at pcu.helsinki.fi.invalid (Jere Kahanpaa)
Date: 6 Dec 2000 15:24:00 GMT
Subject: How can i use a DLL in python ?
References: <t2sjn0aji5dkef@corp.supernews.com>
Message-ID: <90llmg$sto$1@oravannahka.helsinki.fi>

^Socrates^ <joseamartin at venmex.com> wrote:
: Adobe was made a toolkit for Java,C++,and Perl , but currently no version
: for Python.
: I was emailed to them and request the Python Version, but a have no response
: from them.

The fastest way (if you know C/C++ and have a compiler) is using one of the interface generators on 
the C++ layer. Try using SWIG (www.swig.org???). 

You could also use the Java interface from JPython.

Happy hacking

	Jere Kahanp??

--
Lord, make my words as sweet as honey, for one day I may have to eat them


From rburnham at cri-inc.com  Tue Dec 12 17:39:07 2000
From: rburnham at cri-inc.com (Roger Burnham)
Date: Tue, 12 Dec 2000 22:39:07 GMT
Subject: Serial IO for Python 2.0???
References: <9166jh$5hh$1@nnrp1.deja.com>
Message-ID: <3a36a72c.8378818@news-server.socal.rr.com>

On Tue, 12 Dec 2000 21:50:45 GMT, harrc at my-deja.com wrote:

>Does anyone know of a Win32 serial port I/O solution for Python 2.0?
>Both win32comm and the sio module are compiled for Python 1.5.
>

As I am unable to access my starship site, I'll send a wise installer
of the sio module linked to Python 2.0 to anyone who sends me a 
self-addressed-and-stamped email.


Roger Burnham
rburnham at cri-inc.com


From tim.one at home.com  Thu Dec 28 14:48:43 2000
From: tim.one at home.com (Tim Peters)
Date: Thu, 28 Dec 2000 14:48:43 -0500
Subject: Interesting behaviour of the assignment
In-Reply-To: <92f4gk$qk6$1@news.nuri.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEMDIFAA.tim.one@home.com>

[June Kim]
> >>> a=100
> >>> b=100
> >>> a is b
> 0
> >>> a=99
> >>> b=99
> >>> a is b
> 1
>
> Is this an intended/expected result?

Python doesn't define when it will and won't reuse the same object for an
integer literal, or, indeed, for any immutable object.  If you want to
compare immutable objects, use "==".  If you want to get *really* scared
<wink>, consider function f() here:

>>> a = 100
>>> b = 100
>>> a is b
0
>>> a = 99
>>> b = 99
>>> a is b
1
>>> def f():
	a = 100
	b = 100
	return a is b

>>> print f()
1
>>>

> What is the lying structure beneath this?

Internal optimizations, that can and do vary across implementations.  In
2.0, it just so happens that CPython always stores positive ints < 100
uniquely, and reuses other int objects if and only if the literals appear in
the same code block (which is why f() reused 100, but the one-at-a-time
bindings at the interactive prompt didn't -- each chunk of code entered at
the prompt is compiled into its own, new code block).   None of that can be
counted on across releases, though, and if you think you *need* to figure it
out you're confused about something deeper <0.5 wink>.

so-if-it's-not-just-idle-curiosity-tell-us-what-you're-trying-to-
    accomplish-ly y'rs  - tim




From pearu at cens.ioc.ee  Fri Dec 15 14:54:58 2000
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Fri, 15 Dec 2000 21:54:58 +0200
Subject: __repr__ policy problem: should module be prefixed to output?
In-Reply-To: <91dfl0014sa@news1.newsguy.com>
References: <3A3A3A6D.2C3B1A91@sbc.su.se> <91dfl0014sa@news1.newsguy.com>
Message-ID: <Pine.LNX.4.10.10012152140170.10103-100000@kev.ioc.ee>


On Fri, 15 Dec 2000, Alex Martelli wrote:

> "Per Kraulis" <per at sbc.su.se> wrote in message
> news:3A3A3A6D.2C3B1A91 at sbc.su.se...

> > The problem is that Alt 1 is appropriate in the context of "import
> > crng", while Alt 2 is better when having done "from crng import *".
> 
> If it was only the latter, it would not be worth supporting except
> for those very rare modules that are _designed_ to be from-import-*'ed.
> 
> But a more reasonable usage, such as:
> 
>     from themodule import specificthing, anothergoodie, classname
> 
> would also favour repr strings of the form 'classname(avalue)' over
> 'themodule.classname(avalue)'.
> 
> Existing modules seem to do it both ways, so it's hard to
> find specific precedents for guidance.
> 
> One thing favouring the 'undecorated' form is that an eval
> _using the module's dictionary as global-dictionary_ should
> then work anyway (as well as string-manipulations before
> eval, which are of course possible in either case).
> 

There is a solution to this repr string dilemma. For example,
consider foo.py:

# begin of file foo.py
class A:
    def __init__(self,value):
        self.value = value
    def __repr__(self):
        return '%s.%s(%s)'%(self.__module__,
                            self.__class__.__name__,
                            self.value)

exec 'import %s'%__name__   # this does the trick
                            # (could it be done also in C extension
                            #  modules?)

#eof foo.py

Now in python session:

>>> import foo
>>> foo.A(3)
foo.A(3)
>>> eval(repr(foo.A(3)))
foo.A(3)

and in another python session:

>>> from foo import *
>>> A(3)
foo.A(3)
>>> eval(repr(A(3)))
foo.A(3)

And the answer to the original question is Yes.

Regards,
	Pearu



From aleaxit at yahoo.com  Tue Dec 12 06:26:27 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 12 Dec 2000 12:26:27 +0100
Subject: Fault trees structure
References: <D_IY5.49$17.2785@m2newsread.uni2.es> <9128e902uo3@news2.newsguy.com> <z8eZ5.249$YZ4.8868@m2newsread.uni2.es> <yfhZ5.23310$Ia1.4003773@typhoon.nyroc.rr.com>
Message-ID: <91521d02r3g@news2.newsguy.com>

"Darrell" <news at dorb.com> wrote in message
news:yfhZ5.23310$Ia1.4003773 at typhoon.nyroc.rr.com...
> Jumping right to code :)
> Seemed more fun than reading too deeply about fault trees.

Nice general structure, it seems -- I'll just offer a few
observations based on Python 2:

>     def connectInput(self, name):
>         if not hasattr(self, "_input"):
>             self._input={}
>         self._input[name]=None

def connectInput(self, name):
    self.__dict__.setdefault('_input',{})[name] = None

is a possible alternative (also in other cases), but
not very readable.  It would seem best to just set
empty dictionaries in __init__ for _input and _output,
so as to avoid having test for them at every connect,
and possibly having errors lurking elsewhere, too:

>     def signal(self, collection, name, value=1):
>         self._input[name]=value
>         self.logic(collection)

This will cause an exception if self has no _input,
but work just fine if _input does exist (because ANY
connectInput was previously called on self -- be it
on this specific name, or another).  It seems not
likely that this is the intended semantics.  Having
_input set at __init__ will make this just work; if
the intent is that name must be a string for which
self.connectInput was previously called, this can
also easily be checked with an has_key, for example.

Of course, giving NodeBase an __init__ means having
to call it in lieu, or from, specific subclasses'
ones, but that seems OK; right now OrNode and
AndNode sport identical __init__ methods, so why
not factor them out of the subclasses and up into
the NodeBase common baseclass.


>     def logic(self, collection):
>         cc= reduce(lambda x,y: x and y, self._input.values())
>         self.propagate(collection, cc)

If the values are constrained to "boolean" (0 and 1), then
using operator.and_ as the first argument to reduce might
well be faster, and perhaps more readable (similarly for
or-nodes).


Alex





From dale at out-think.NOSPAMco.uk  Wed Dec 13 07:44:09 2000
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Wed, 13 Dec 2000 12:44:09 +0000
Subject: reloading modules question
References: <90scdf$hpe$1@troll.powertech.no> <ouk43t494isd9m8fk92udjjbg2pb6crq7q@4ax.com> <910om3$2vk$1@troll.powertech.no> <kg9a3t0s6ah2c7t3pf94o3vl4ls2hespuc@4ax.com> <916m48$648$1@troll.powertech.no>
Message-ID: <mjre3t8ta33bj9agc7l4qktmgisibtlvif@4ax.com>

"Syver Enstad" <syver at NOSPAMcyberwatcher.com> wrote:

><dale>
>> Starting and stopping the web site has always worked for us. (Windows 2000
>Server)
></dale>
>
>You mean something like net stop w3svc? Because just stopping the default
>web site in inetmgr doesn't work for me on Win 2000 server. Unloading the
>application works however.
>
>
Not the default site, the site that's using your code - unless that *is* the default site.

We have dozens of sites running on a single server, each site can be started and stopped
independently from the MMC.

It might be relevent to note that Applciation Protection is set to High on our sites.
--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants


From slhath at home.com.nospam  Mon Dec 11 11:29:06 2000
From: slhath at home.com.nospam (Scott Hathaway)
Date: Mon, 11 Dec 2000 16:29:06 GMT
Subject: creating an odbc link with python
References: <gSVX5.55717$yR4.1401853@news1.rdc1.tx.home.com> <3A34CD79.F13410CC@infercor.no>
Message-ID: <mh7Z5.6116$bw.472470@news.flash.net>

I have a little more info now.  I basically just need to add a key and some
values to the windows registry.

Thanks,
Scott

"Paul Boddie" <paulb at infercor.no> wrote in message
news:3A34CD79.F13410CC at infercor.no...
> Scott Hathaway wrote:
> >
> > I am using python to create an installation program for a web app that
> > we have at our company.  The web app uses an access database that is
> > hooked up through odbc.  How can I programatically create the dsn in
> > python (I guess I need to edit the windows registry)?
>
> I think that what you are asking is how one can call various functions in
the
> win32 API so that an ODBC data source gets added to the list maintained by
ODBC
> driver manager. Whilst I have no experience in this area, it would be
> interesting to know how one might achieve this. Perhaps it is a case of
calling
> methods in various COM interfaces.
>
> Regards,
>
> Paul




From ee_fhw at stu.ust.hk  Sat Dec 23 00:41:41 2000
From: ee_fhw at stu.ust.hk (Calvin Fong)
Date: Sat, 23 Dec 2000 13:41:41 +0800
Subject: How to read keyboard input
Message-ID: <3A443B15.C7E926A@stu.ust.hk>

Dear all,
    I'm finding a fcuntion to read the input of keyboard
    is it something like getchar, i just get lost.
    anyone know it?
Yours
Calvin



From bobalex at home.com  Sat Dec 23 12:00:25 2000
From: bobalex at home.com (Bob Alexander)
Date: Sat, 23 Dec 2000 09:00:25 -0800
Subject: Some Python 2.1 ideas
Message-ID: <005f01c06d01$d8114820$74eb0b18@C322162A>

Here are some thoughts on possible enhancements to Python. If discussion
shows any positive enthusiasm maybe they could turn into PEPs.

Bob

----
New operator: embedded assignment

As C and Java programmers know, it is sometimes more expressive to embed
assignments inside other expressions. It's my guess that one reason Python
doesn't support this is to make "=" illegal inside expressions, thus
avoiding the "=" vs. "==" errors common in C. A response to that objection
is to use a different token for embedded assignment -- I suggest ":=" since
it's already familiar to many folks as an assignment operator. Then instead
of:

    while 1:
        line = f.readline()
        if not line:
            break
        print line    # yes, this will double space, but see another
suggestion below  :-)

We could code:

    while line := f.readline():
        print line

There are many such cases where "assignment as an operator" can make
programs more compact and expressive.

Is there a downside? The only one I can think of is that assignment targets
would not always appear at the left end of program lines!

In summary:
    - Introduce a new binary operator of low precedence and right
associativity, x := y, which assigns y to x and returns y.
    - "if x = y: ..." still produces an error, but "if "x := y: ..." does
not.


----
Variations on "for"

Here are some possible additional forms of "for" that seem appealing:

Allow slice-like syntax as shorthand for xrange():
    for i in 10:                # equivalent to "for i in xrange(10)"
    for i in 5:10:             # equivalent to "for i in xrange(5, 10)"
    for i in 20:10:-1:       # equivalent to "for i in xrange(20, 10, -1)"

Note that although the functionality is the same as using xrange, there is
no requirement the the implementation actually use xrange objects to provide
these loops over numeric ranges.

Make the "<variable> in" part optional:
    for 10:                     # loops 10 times (the "10" is the xrange
shorthand suggested above)
    for alist:                   # loops as many times as there are elements
in "alist"

Empty "for" loops indefinitely (equivalent to the beloved "while 1"):
    for:
        <code that eventually breaks or raises>

A subthought: Slice notation as shorthand for xrange() need not be
restricted to "for" statements. (I've never really cared for the name
"xrange" for the functionality it provides, anyway. I'd be happy to see an
alternative!)


----
String method to remove trailing newline

Since readline() and its relatives return strings with '\n' attached
(sometimes), it would be convenient to have a built-in string method that
removes one if it exists (seems like I'm often coding little functions to do
this).

    <string>.chomp()    # sorry, borrowed the method name from Perl

Then, combined with above suggestions:

    while line := f.readline();
        print line.chomp()

This method is not much different from <string>.rstrip(), except that it
leaves other trailing spaces intact.





From emile at fenx.com  Fri Dec 29 12:54:45 2000
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 29 Dec 2000 09:54:45 -0800
Subject: Bug or Feature?
Message-ID: <92ij58$74k3p$1@ID-11957.news.dfncis.de>

Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)]
on win32
(and linux current cvs)

This works:

print eval.__doc__

but this doesn't:

print exec.__doc__


--

Emile van Sebille
emile at fenx.com
-------------------






From mballen at erols.com  Sun Dec 31 03:27:13 2000
From: mballen at erols.com (Michael B. Allen)
Date: 31 Dec 2000 08:27:13 GMT
Subject: tftplib ?
References: <Pine.LNX.4.21.0012310129140.9173-100000@saturn.ispol.com> <3A4EDF22.30F29C1@zip.com.au>
Message-ID: <slrn94trvl.b73.mballen@angus.foo.net>

Erik de Castro Lopo wrote:
>> is there by a chance a tftplib.py or something that implements tftp in
>> Python out there?
>
>What tftp and how does it differ from ftp?

tftp is "trival". I believe the ftp module provided with the Python API
will work just fine Gregory. Yes?

Mike

-- 
signature pending


From garryknight at bigfoot.com  Thu Dec 21 13:08:54 2000
From: garryknight at bigfoot.com (Garry Knight)
Date: Thu, 21 Dec 2000 18:08:54 +0000
Subject: while (thread.alive()):
References: <mailman.977283904.26696.python-list@python.org>
Message-ID: <91tgvm$3q8$2@taliesin.netcom.net.uk>

In article <mailman.977283904.26696.python-list at python.org> "Jay
O'Connor" <joconnor at cybermesa.com> wrote:

> I'm continuously amused that people try to take language design
> forward and still drag obsolete ideas along for the ride.

Like having a print command that doesn't, by default, send output
to the printer...  :o)

-- 
Garry Knight
garryknight at bigfoot.com


From sdm7g at virginia.edu  Tue Dec 12 13:27:23 2000
From: sdm7g at virginia.edu (Steven D. Majewski)
Date: Tue, 12 Dec 2000 13:27:23 -0500 (EST)
Subject: Python-like compiled language
In-Reply-To: <3A35835B.AAA92005@bioeng.ucsd.edu>
References: <913jr3$dmv$1@sylvester.vcn.bc.ca> <3A35835B.AAA92005@bioeng.ucsd.edu>
Message-ID: <Pine.NXT.4.21.0012121317260.232-100000@localhost.virginia.edu>


> Alexander Lawhead wrote:
> > 
> > This may seem like an odd question, but what is the most Python-like
> > compiled language?


On Mon, 11 Dec 2000, Curtis Jensen wrote:

> The closest compiled languege to Python is probaby C/C++. 

I would say objective-c is a closer semantic fit to Python than C++. 

Syntactically, it's just like C with a little bit of Smalltalk 
messaging syntax grafted onto it. 

With PyObjC, integration with Python is easier -- you don't need
SWIG or CXX or anything. ( However, at the moment I'm running into
bugs going in the other direction -- i.e. trying to use the Obj-C
proxy to Python object to extend an objective-C class. I'm using
it on the Mac (OSX), so I don't know how stable the gnu Foundation
classes are on linux. ) 


-- Steve Majewski <sdm7g at Virginia.EDU> 






From obou at my-deja.com  Mon Dec  4 03:49:03 2000
From: obou at my-deja.com (obou at my-deja.com)
Date: Mon, 04 Dec 2000 08:49:03 GMT
Subject: SAP R/3 automation with win32com
Message-ID: <90flpu$gpu$1@nnrp1.deja.com>

Hello!

I'm trying to do SAP R/3 automation with win32com using the SAP DCOM
Connector (a tool that presents SAP Business objects as COM Components).
The business objects (BAPI) are kind of objects with parameters and
methods.
Everything works fine, except the function call (the most important
part of course), and it could be because of the way I have to give the
different parameters to the BAPI method. I have a VB example and don't
know if the way I convert the VB function call to python function call
is correct.
Maybe someone who knows both VB and Python could help me.

Here's the Python code:

from win32com.client import Dispatch

# Create a connection object
sapConn = Dispatch('SAP.CCSession.1')

# Insert connection informations
syst = raw_input('System :')
uname = raw_input('Username: ')
pwd = raw_input('Password: ')
client = raw_input('Client: ')
sapConn.PutSessionInfo(syst, uname, pwd, 'EN', client)

# Log in
sapConn.Logon()

# Create a sales order business object
salesOrd = sapConn.CreateInstance('SAP.SalesOrder.1')

# Create the parameters for calling the method CreateFromData
header = salesOrd.DimAs('BapiCreateFromData', 'OrderHeaderIn')
itemsIn = salesOrd.DimAs('BapiCreateFromData', 'OrderItemsIn')
partners = salesOrd.DimAs('BapiCreateFromData', 'OrderPartners')
ret = salesOrd.DimAs('BapiCreateFromData', 'Return')

# Initialize structure with data.
partners.AddNew()
partners.Fields("Partn_Role").Value = 'AG'
partners.Fields("Partn_Numb").Value = '0000001234'
partners.Update()

# Method call. Everything works fine up to this call!!!
salesOrd.BapiCreateFromData(OrderHeaderIn = header,
                            OrderItemsIn = itemsIn,
                            OrderPartners = partners,
                            Return = ret)

In VB this function call looks like this:
salesOrd.BapiCreateFromData OrderHeaderIn:=header,_
                          OrderItemsIn:=itemsIn,_
                          OrderPartners:=partners,_
                          Return:=ret

I tried it with or without the parameter names, with or without the
return value in the parameters list, I always get this exception:

Traceback (most recent call last):
  File "d:\python20\pythonwin\pywin\framework\scriptutils.py", line
301, in RunScript
    exec codeObject in __main__.__dict__
  File "d:\sapconn.py", line 35, in ?
    Return = ret)
  File "<COMObject CreateInstance>", line 2, in BapiCreateFromData
  File "d:\python20\win32com\client\dynamic.py", line 234, in
_ApplyTypes_
    result = apply(self._oleobj_.InvokeTypes, (dispid, LCID, wFlags,
retType, argTypes) + args)
com_error: (-2147352567, 'Exception occurred.', (0, None, "Table
doesn't contain 1 record\015\012(COM_PARAMETER_NO_SINGLE_RECORD): error
code 16388", None, 0, -2147467260), None)

Thank you for your assistance,

Olivier Boudry



Sent via Deja.com http://www.deja.com/
Before you buy.


From dougfort at downright.com  Sat Dec 30 05:23:48 2000
From: dougfort at downright.com (Doug Fort)
Date: Sat, 30 Dec 2000 05:23:48 -0500
Subject: httplib and '100 Continue' after POST
Message-ID: <UIi36.7084$Sl.370436@iad-read.news.verio.net>

I can't figure out how to persuade httplib to handle a response status of
'100 Continue'.  Has anyone accomplished this?

--
Doug Fort
Senior Meat Manager
Downright Software
http://www.dougfort.com/~dougfort





From jmg at ecs.soton.ac.uk  Fri Dec  8 11:11:29 2000
From: jmg at ecs.soton.ac.uk (Jacek Generowicz)
Date: 08 Dec 2000 16:11:29 +0000
Subject: The secrets of operator.add
Message-ID: <g08zpq51vy.fsf@scumbag.ecs.soton.ac.uk>

Where can I find out about the ways in which the functions in the
operator module are optimized, and how to use them to best effect in my
code ?

For example, the following bit of code:

from operator import add
from profile import run
N = 50000
run ('reduce(lambda x,y:x+y, range(N))')
run ('reduce(add, range(N))')

already throws up a surprise for me:

         50001 function calls in 1.600 CPU seconds

   Ordered by: standard name

   ncalls  tottime  percall  cumtime  percall filename:lineno(function)
    49999    0.840    0.000    0.840    0.000 <string>:1(<lambda>)
        1    0.720    0.720    1.560    1.560 <string>:1(?)
        0    0.000             0.000          profile:0(profiler)
        1    0.040    0.040    1.600    1.600 profile:0(reduce(lambda x,y:x+y, range(N)))


         2 function calls in 0.060 CPU seconds

   Ordered by: standard name

   ncalls  tottime  percall  cumtime  percall filename:lineno(function)
        1    0.060    0.060    0.060    0.060 <string>:1(?)
        0    0.000             0.000          profile:0(profiler)
        1    0.000    0.000    0.060    0.060 profile:0(reduce(add, range(N)))


It looks as if multiple calls to operator.add are avoided somehow
. . . or am I failing to understand the profiler's output ?

Is there a way of getting similar advantages in

run ('map(lambda x:x+1, range(N))')  ?

Is there a Python way of currying add? 

Jacek


From pshinners at mediaone.net  Sat Dec 23 13:53:22 2000
From: pshinners at mediaone.net (Pete Shinners)
Date: Sat, 23 Dec 2000 18:53:22 GMT
Subject: Some Python 2.1 ideas
References: <mailman.977590932.19335.python-list@python.org>
Message-ID: <Cw616.23054$qN6.3308205@typhoon.we.rr.com>

"Bob Alexander" <bobalex at home.com> wrote
> String method to remove trailing newline

there is a string method named "strip()" that will
remove all whitespace (newline&linefeeds included)
from a string.

actually, since strings are immutable, it returns a
new string with the appropriate changes made.

there are also "rstrip" and "lstrip" functions which
will remove whitespace from only the end or beginning
of the string.


>>> print 'Example string gotten from readline\n'.strip()
'Example string gotten from readline'






From root at [127.0.0.1]  Sun Dec 24 19:57:15 2000
From: root at [127.0.0.1] (deadmeat)
Date: Mon, 25 Dec 2000 00:57:15 GMT
Subject: os.exec?
Message-ID: <LXw16.8561$KY1.23035@news1.rivrw1.nsw.optushome.com.au>

How to I execute an external program without replacing the Python
interpreter? I want control to return to Python when the external program is
completed. I still want stdio to go to the console not to a pipe if
possible.






From sh at ttsoftware.co.uk  Fri Dec 15 11:38:59 2000
From: sh at ttsoftware.co.uk (Steve Horne)
Date: Fri, 15 Dec 2000 16:38:59 +0000
Subject: Python advocacy
References: <91bln3$lmr$1@nnrp1.deja.com> <91d63k0msl@news1.newsguy.com> <8iak3t4a3lrqcqp9iqftmekdoahuico2b0@4ax.com> <91df1c$j85$1@troll.powertech.no>
Message-ID: <j5ik3tor5ipefh0c8k670fbck4pmatrso2@4ax.com>

On Fri, 15 Dec 2000 16:54:56 +0100, "Syver Enstad"
<syver at NOSPAMcyberwatcher.com> wrote:

>Steve, I think Alex was joking :-)

Today is definitely the day I should have stayed in bed. Either that,
or maybe I should have tried waking up.

Oh well.

-- 
Steve Horne
Home : steve at lurking.demon.co.uk
Work : sh at ttsoftware.co.uk


From nospam at caltech.edu  Sat Dec  9 17:52:28 2000
From: nospam at caltech.edu (Nathaniel Gray)
Date: Sat, 9 Dec 2000 14:52:28 -0800
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net> <3A2EE2AF.AD8A51E2@caltech.edu> <slrn92tu7t.lk.amk@207-172-111-60.s60.tnt1.ann.va.dialup.rcn.com> <tcmay-8BCFAB.23025408122000@news.got.net>
Message-ID: <90udgm$32g@gap.cco.caltech.edu>

Tim May wrote:

> I did some work on simulated annealing while at Intel, which is why I
> commented in the SA thread. 

Wow.  I hadn't even noticed it was the same person.  This one's even worse. 
 Just out of curiosity, why did you bother to take the time to do a google 
search just to castigate the original poster in that thread?

His post:
> Has anybody implemented Simulated Annealing in python?
> 
> I couldn't find it searching dejanews nor python.org. It would save me 
> some time if somebody who had written
> that already would share it;-)) 
> 
> Thanks,
> Jochen

Your response:
> Come on, folks. I find 1210 articles in Google by searching on:
>  
> "simulated annealing" python
>  
> While some are not directly related to "simulated annealing in Python,"
> this is clearly the first place to look.
> Asking us to do a search is nonsense.

He didn't ask you to do a search.  He asked if anybody had solved the 
problem and would be willing to share their code.  He was quite polite, and 
even explicitly stated that he'd already searched dejanews and python.org.
What exactly did you take issue with?  Did he use the wrong search engine 
for your tastes?  

Not every request for information on a newsgroup is laziness, you know.
-- 
Nathaniel Gray
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
                   n8gray <at> caltech <dot> edu
                   California Institute of Technology
                     Computation and Neural Systems
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From wzab at ise.pw.edu.pl  Thu Dec  7 07:46:16 2000
From: wzab at ise.pw.edu.pl (Wojtek Zabolotny)
Date: Thu, 7 Dec 2000 13:46:16 +0100
Subject: Python Numeric (in python 1.5.2) & TNT (Template Numerical Toolkit)
Message-ID: <Pine.LNX.4.21.0012071340290.697-100000@ipebio15.ise.pw.edu.pl>

Hi Python Gurus,

I have a big program for numerical analysis, which uses the TNT 
(http://math.nist.gov/tnt/) classes to deal with matrices.
In fact I use the Fortran_Matrix<double> for my matrices.
Now I need to add the python frontend to it (for interactive
calculations and plotting).
Did anybody tried to tie the TNT matrices to Python Numeric arrays?
I'd like to avoid memory reallocation and data copying when transferring
data between C++ and python parts...

--
			TIA & Regards
			Wojciech Zabolotny
			wzab at ise.pw.edu.pl




From mwh21 at cam.ac.uk  Tue Dec 12 04:29:46 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 12 Dec 2000 09:29:46 +0000
Subject: Inheriting the @ sign from Ruby
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu>
Message-ID: <m3lmtm2dit.fsf@atrus.jesus.cam.ac.uk>

Roy Katz <katz at Glue.umd.edu> writes:

> What faults or strengths do you see with this?

It hasn't got a hope in hell!  The language is incredibly unlikely to
change this much, I think.

Also, I think it looks horrible, but that could just be me.

Cheers,
M.

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


From DanielK at aracnet.com  Tue Dec 19 14:46:26 2000
From: DanielK at aracnet.com (Daniel Klein)
Date: Tue, 19 Dec 2000 11:46:26 -0800
Subject: Converting a list to a string
Message-ID: <sRO%5.168$LU6.63588@typhoon.aracnet.com>

My application needs to be able to convert a list to a string and vice
versa. Taking the first problem, the best I could come up with is

s = ''
for i in range(len(mylist)):
    s += mylist[i]

Of course, I would turn this into a function (or better yet a class method).

Is here a builtin function to do this that I'm missing? Or is this as good
as it can get?

Thanks,
Daniel Klein




From thomas at xs4all.net  Mon Dec  4 12:58:58 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Mon, 4 Dec 2000 18:58:58 +0100
Subject: Python mailing list/usenet bridge.
In-Reply-To: <3A2BCC0A.1ABDE864@tismer.com>; from tismer@tismer.com on Mon, Dec 04, 2000 at 06:53:30PM +0200
References: <8vt9t4$r0f$1@nnrp1.deja.com> <900365$fp6$1@nnrp1.deja.com> <905t5s$6ef$1@nnrp1.deja.com> <200012041005.EAA23178@x56.deja.com> <20001204180354.C4396@xs4all.nl> <3A2BCC0A.1ABDE864@tismer.com>
Message-ID: <20001204185858.E29735@xs4all.nl>

On Mon, Dec 04, 2000 at 06:53:30PM +0200, Christian Tismer wrote:

> What's the best advice for posting:
> To the news, and maybe it doesn't get to the mailing list,
> or vice versa?

Neither gets delivered to the other :P Until the two are rejoined, you'll
have to read in both and post in both. Note that tyrannewsaurus is
read-only, you can't use it to post -- unless you are willing to pay fl.30,-
(US$15, more or less) a month for the priviledge of having an XS4ALL account
=)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From slhath at home.com  Fri Dec 29 16:59:36 2000
From: slhath at home.com (Scott Hathaway)
Date: Fri, 29 Dec 2000 21:59:36 GMT
Subject: automating excel with python ... getting excel to go away
References: <baP26.99236$yR4.2642456@news1.rdc1.tx.home.com> <3A4BDEFD.8000603@ActiveState.com>
Message-ID: <cP736.103690$yR4.2714440@news1.rdc1.tx.home.com>

Thanks!

Scott

"Mark Hammond" <MarkH at ActiveState.com> wrote in message
news:3A4BDEFD.8000603 at ActiveState.com...
> Scott Hathaway wrote:
>
> >  myExcel.Quit
>
> Quit is a method - try:
>
> myExcel.Quit()
>
> Mark.
>




From jochen at pc1.uni-duesseldorf.de  Wed Dec  6 05:42:10 2000
From: jochen at pc1.uni-duesseldorf.de (Jochen =?iso-8859-1?q?K=FCpper?=)
Date: 06 Dec 2000 11:42:10 +0100
Subject: simulated annealing
References: <udn1eluhzo.fsf@yogi.pc1.uni-duesseldorf.de> <roy-9B2D39.19561027112000@news.panix.com> <tcmay-98F57B.17064527112000@news.got.net> <udvgt8xr1t.fsf@yogi.pc1.uni-duesseldorf.de> <tcmay-849C4E.19535003122000@news.got.net>
Message-ID: <udwvddompp.fsf@box.home.de>

>>>>> Tim May wrote on Sun, 03 Dec 2000 19:53:50 -0800:

Tim> In article <udvgt8xr1t.fsf at yogi.pc1.uni-duesseldorf.de>, "Jochen 
Tim> K?pper" <jochen at pc1.uni-duesseldorf.de> wrote:

>> >>>>> Tim May writes:
>> 
>> >> "Jochen K?pper" <jochen at pc1.uni-duesseldorf.de> wrote: 

>> >> Has anybody implemented Simulated Annealing in python?

Ok, here a two pointers to code I found:

http://margaretmorgan.com/wesley/python/anneal.py
http://correio.cc.fc.ul.pt/~jpp/classes/heur-main/node50.html

Both provide basic working implementations. I looked some more into
the second one, so far I haven't found a bug in that SA code, but in
my wrapper:O

Probably there are more people out here using SA without the need to
publish it on the web. If anybody with a more advanced implementation,
probably even ASA or such beasts would be willing to share their code,
I'd be happy;-))
(Tim, I couldn't do a "adaptive simulated annealing" NEAR python
search -- and couldn't find an implementation of ASA or such in python
yet:-))


Tim> Glad you have "found" Google. I had been hearing about how great it was 
Tim> for several months before trying it. 

Me too.

Tim> When I finally did, I realized how much better it was than Alta
Tim> Vista and Metacrawler.  (I used to like Metacrawler because it,
Tim> ostensibly, searches with N different search engines. In
Tim> practice, it doesn't find as much as just Google does.)

It actually searches google! And that is why I was confident it was
the right thing to use metacrawler...

Tim> Sorry if my "Come on, folks" language might have sounded harsh. I just 
Tim> don't the patience to pad my comments out with the typical net.polite 
Tim> bullshit:

Probably you could write what you did, besides the "com on folks";-) 

Greetings,
Jochen
-- 
Einigkeit und Recht und Freiheit                http://www.Jochen-Kuepper.de
    Libert?, ?galit?, Fraternit?                GnuPG key: 44BCCD8E
        Sex, drugs and rock-n-roll


From aleaxit at yahoo.com  Thu Dec 28 04:18:18 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 28 Dec 2000 10:18:18 +0100
Subject: Erlang (was Re: Good introduction to functional programming with Python?)
References: <mailman.977984412.30211.python-list@python.org>
Message-ID: <92f0il01a5u@news1.newsguy.com>

"Tim Peters" <tim.one at home.com> wrote in message
news:mailman.977984412.30211.python-list at python.org...
    [snip]
> I haven't used Erlang at all, so that makes it an excellent battlefield:
no
> unpleasant haggling over facts <wink>.  I'll hazard a guess based on the

Right!

> In the absence of higher-order functions, I can believe programmers don't
> have much trouble reasoning about types without declarations.  They don't

The _scarcity_ (not _absence_:-) of HOF in typical Erlang use may be a
cultural issue (or maybe they just don't buy one all that much in its
typical use cases); and from that may follow the relative lack of
interest in its (existing, but 'add-on') type-inferencing possibilities.

It's not hard to find examples of Erlang HOF's on the web though --
right on http://www.erlang.org/white_paper.html we get a typically
simple one:
"""
As another an example we can write:
  > Adder = fun(N) -> fun(X) -> X + N end end.
  #Fun
  > G = Adder(10).
  #Fun
  > G(5).
  15
"""

No big fuss, and the explicit "fun(...) -> Expr end" syntax with
which, it says, "function objects are introduced"... it *does*
read better than 'lambda', methinks, but this is just a sugary
issue, of course.

Part of the 'cultural' explanation may be that not a little 'typical
FP stuff' was added relatively recently to Erlang, so it's not in
the book; http://www.erlang.org/doc/r7b/pdf/extensions-5.0.1.pdf
summarizes the additions, though, including 'Funs'.  As this doc
says,
"""
Higher Order Functions
Funs encourages us to encapsulate common patterns of design into
functional forms called higher order functions. These functions
not only shortens programs, but also produce clearer programs
because the intended meaning of the program is explicitly rather
than implicitly stated.
"""
and I suspect that the emphasis on 'explicitly rather than
implicitly stated' may charm you enough to convince you to
go easy on the author's obvious difficulty with s-endings
in English verb forms with plural subjects:-).

The typical example of HOF in this doc is:
"""
map(F, [H|T]) -> [F(H)|map(F, T)];
map(F, []) -> [].
We can now express the functions double and add_one in terms
of map as follows:
double(L) -> map(fun(X) -> 2*X end, L).
add_one(L) -> map(fun(X) -> 1 + X end, L).
"""

Actually, of course, 'map', along with several other typical
HOF's, is exported from a 'lists' module.


List comprehensions (aka 'Zermelo-Frankel', aka ZF, ones:-)
are also introduced in the same docs, with rather Haskellish
syntax if I may say so:

> [X || X <- [1,2,a,3,4,b,5,6], integer(X), X > 3].
[4,5,6]

Here, 'explicit is better than implicit' seems to have been
slightly put aside -- each qualifier is either a generator,
if it includes a '<-', or else a filter; Python's choice of
expressing this as:

>>> [X for X in [1,2,'a',3,4,'b',5,6] if type(X)==type(0) if X > 3]
[4,5,6]

looks more explicit to me (not necessarily more readable:-).


> use the static type checker because it's more trouble than it's worth!

Because the HOF's, while present, are not widely used?  Maybe.

> Fine by me.  I must say, though, that Erlang appears to occupy an extreme
> corner of the FP universe.

It sure looks unique to me for its emphasis on concurrency (if there
are other FP's that are similarly concurrency-oriented, I have not seen
them yet).  But Scheme (albeit impure) and Erlang (albeit 'culturally
not highly HOF-prone') together surely make up for a substantial share
of functional-programming use, and just as surely form a solid case for
static-typing NOT being central to FP (just as, say, Smalltalk and
Python together form a solid case for static-typing not being central
to OO, no matter how much Eiffel'ers may rant otherwise:-).


> I *like* the attitude of the Erlang developers.  Looks like a very useful
> language for its target domain.  Even better, according to FAQ 5.7.11
> there's a Python binding (if you really need some higher-order functions
> after all <wink>).

It looks to me as if the Python/Erlang interface is based on an older
(perhaps now-removed) external-functions layer of Erlang, and was only
release at a '0.1' level.  Still, Erlang's current external-functions
approaches, http://www.erlang.org/doc/current/doc/tutorial/part_frame.html,
appear to offer ample scope for such interfacing -- there's a mention
of an IDL compiler, which, as usual, sounds more interesting to me than
'point-to-point' approaches to interlanguage-operation (even when one
of the languages is Python!-).  The barely-mentioned 'IC' IDL compiler
seems to be part of OTP, the Open Telecom Platform framework for Erlang,
so I'm not going to dig into this further at this time (as I don't do
telecom programming:-), but, conceptually, it makes sense...!-)


Alex


>
> if-we-don't-act-fast-i'm-afraid-python-may-lose-the-industrial-
>     control-market-ly y'rs  - tim
>
>




From thomas.heller at ion-tof.com  Wed Dec 13 12:11:28 2000
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Wed, 13 Dec 2000 18:11:28 +0100
Subject: Trivial Distutils problem
References: <Pine.SOL.4.30.0012131525400.9329-100000@mimosa.csv.warwick.ac.uk>
Message-ID: <918ak4$3b44p$1@ID-59885.news.dfncis.de>

"John J. Lee" <phrxy at csv.warwick.ac.uk> wrote in message news:Pine.SOL.4.30.0012131525400.9329-100000 at mimosa.csv.warwick.ac.uk...
>
> With the very basic Distutils script below, I get the following error:
>
> $ python setup.py build
> running build
> running build_py
> error: package directory 'TableIO' does not exist
>
> Why does it think I want to install a package?  It's just one module to go
> in the root package.
>
> I'm on windows (above output pasted from a cygwin window, but it does the
> same in a DOS shell).
>
>
> #! /usr/bin/env python
>
> # Distutils setup.py for TableIO
>
> from distutils.core import setup, Extension
>
> setup(name = "TableIO",
>       version = "1.4",
>       description = "A module for reading ASCII tables into NumPy arrays (and "
>        "lists).",
>       author = "Michael A. Miller",
>       author_email = "miller5 at uiuc.edu",
>       py_modules = ["TableIO.py"]
                       ^^^^^^^^^^
Distutils wnats to install package TableIO with the module py.
Change this into (do NOT specify the .py extension):
>       py_modules = ["TableIO"]

Thomas




From tim.hochberg at ieee.org  Fri Dec 22 20:58:56 2000
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Sat, 23 Dec 2000 01:58:56 GMT
Subject: A Suggestion for Python Dictionary/Class
References: <920fe9$c20$1@hecate.umd.edu>
Message-ID: <AFT06.88460$15.18637154@news1.rdc1.az.home.com>

"William Djaja Tjokroaminata" <billtj at y.glue.umd.edu> >:
>However, there is a fundamental difference with Python class.  The members
> are implemented using dictionary.  The danger with dictionary is, even
> though a key does not exist, we still can make an assignment to it.  When
> we make some typo, this can be hard to debug.  For example:


Several people have suggested using a custom __setattr__ to prevent
modifying the class dictionary. Personally I think this is not normally
worth the hassle, but I think the following is a fun variation on the
__setattr__ anyway. It has the added twist of checking type, although that
could easily be disabled.

-tim
---------------------------------------

import types

class Typed:

    def __setattr__(self, key, value):
        if not hasattr(self.__class__, key):
            raise KeyError("can not set %s" % key)
        if type(value) != getattr(self.__class__, key):
            raise ValueError("%s must be of type %s" % (key,
getattr(self.__class__, key)))
        self.__dict__[key] = value


class MyTypedClass(Typed):

    anInt = types.IntType
    aFloat = types.FloatType
    aString = types.StringType
    aTuple = types.TupleType



m = MyTypedClass()

m.anInt = 5
m.aFloat = 5.0
m.aString = "five"
m.aTuple = (1,2,3,4,5)
try: m.somethingElse = 5
except StandardError, err: print err
try: m.anInt = 5.0
except StandardError, err: print err




From DanielK at aracnet.com  Tue Dec 19 11:36:50 2000
From: DanielK at aracnet.com (Daniel Klein)
Date: Tue, 19 Dec 2000 08:36:50 -0800
Subject: sys.stdout.write()
Message-ID: <r3M%5.159$LU6.61501@typhoon.aracnet.com>

In 'Learning Python' by Mark Lutz and David Ascher, on page 77 it makes
mention of the 'sys.stdout.write' method. However, when I check this out, I
don't see the 'write' method in the list...

>>> import sys
>>> dir(sys.stdout)
['shell', 'softspace', 'tags']

I must be doing something fundamentally wrong. :^(

I guess this brings up the bigger question of self help. Is there an easy or
prescribed way to find out what module (or class) a particular method is
implemented in?

As a side (but related) issue, the current online reference material does
not always indicate all of the exceptions thrown by a method or function.
What is the best way to determine this when writing code?

Thanks,
Daniel Klein






From joconnor at cybermesa.com  Tue Dec 12 14:10:30 2000
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Tue, 12 Dec 2000 19:10:30 +0000
Subject: Smalltalk and Python
Message-ID: <3A367826.262C9206@cybermesa.com>

>Having said that Python is not really a great substitute for Smalltalk

Agreed, Python is not a good tool for Smalltalk in places where
Smalltalk is a good tool, and vice versa.

Python is a good tool for those used to Smalltalk who want something
palatable in places where Smalltalk is not a good tool (and Perl is not
palatable :)

>I may be  setting up for failure, due to Python's mediocre GUI, but I'm having a go at it.

Have you look at the gtk interface for Python?  Looks pretty good.  Or
is that not an option for your platform

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

"God himself plays the bass strings first when He tunes the soul"



From loewis at informatik.hu-berlin.de  Wed Dec 27 20:20:54 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 28 Dec 2000 02:20:54 +0100
Subject: Kanji
References: <92dl3u$to0$1@news.idiom.com>
Message-ID: <p6qlmt1e3yh.fsf@informatik.hu-berlin.de>

"James T. Dennis" <jadestar at idiom.com> writes:

>  I have the necessary fonts installed.  But how can I access 
>  them from Python?

Depends on your system, somewhat. I would recommend to use Tkinter;
make sure you use at least Tk 8.3 and Python 2.0. Then, when you pass
Unicode objects to the Tk functions, Tk should automatically select
the right font (or else you can always explicitly select one). If you
need to read data in other but UTF-8/16 encodings, you'll also need
the JapaneseCodecs package. Check www.python.org/sigs/i18n-sig for
details.

Regards,
Martin



From fredrik at effbot.org  Thu Dec 21 02:10:15 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Thu, 21 Dec 2000 07:10:15 GMT
Subject: re different between US and German versions of NT
References: <mailman.977362144.17768.python-list@python.org> <slrn942rjn.l1.amk@207-172-111-180.s180.tnt1.ann.va.dialup.rcn.com>
Message-ID: <r1i06.769$Qb7.67123@newsb.telia.net>

A.M. Kuchling wrote:
> sre uses the C isalnum() function/macro at one point; isalnum() is
> locale-aware, so presumably it's matching more characters on the
> German version.

SRE uses isalnum only if you're using the LOCALE flag when
you compile the pattern.

if you don't want the pattern to depend on the locale, don't
use the LOCALE flag...

</F>




From fgeiger at datec.at  Sat Dec 30 09:59:51 2000
From: fgeiger at datec.at (Franz GEIGER)
Date: Sat, 30 Dec 2000 15:59:51 +0100
Subject: Komodo
References: <B672A66F.72CC%bobhicks@adelphia.net>
Message-ID: <92ktej$vfa$1@newsreaderg1.core.theplanet.net>

I tried the beta 1. The GUI isn't finished yet (pref's cannot be saved etc.)
and the whole thing is rather slow, which is probably because of the
software's beta stadium. I'm really lookin' forward the 1.0 release version
of this very promising product.

Regards
Franz GEIGER

P.S.: Be careful when installing, don't use a production machine. My
ActivePython only worked well again after deinst. ActiveKomodo, ActivePerl,
and ActivePython and reinst. ActivePython. Before that PythonWin crashed all
the time after the inst. of Komodo.


"Robert L Hicks" <bobhicks at adelphia.net> wrote in message
news:B672A66F.72CC%bobhicks at adelphia.net...
> Has anyone tried Komodo from ActiveState yet? What are you thoughts on it?
I
> see that a Linux beta is out now as well.
>
> Will Komodo for Linux have to have the ActiveState versions of Perl and
> Python to work?
>
> Bob
>




From aahz at panix.com  Tue Dec 26 18:47:36 2000
From: aahz at panix.com (Aahz Maruch)
Date: 26 Dec 2000 15:47:36 -0800
Subject: One Python 2.1 idea
References: <mailman.977771949.17911.python-list@python.org> <92ab9u$dou$1@nnrp1.deja.com> <92ap0i22b6r@news1.newsguy.com> <slrn94ie9k.anf.neelk@alum.mit.edu>
Message-ID: <92bamo$roe$1@panix3.panix.com>

In article <slrn94ie9k.anf.neelk at alum.mit.edu>,
Neelakantan Krishnaswami <neelk at alum.mit.edu> wrote:
>
>I'm not sure this would help very much. Tim posted a link to an
>attempt to add method caches, and reported inconclusive results.  As
>of late last week, I'm not surprised; the Python interpreterseems to
>be spending most of its time in bytecode instruction dispatch. Logic
>suggests that that's where performance enhancements would have the
>most impact.
>
>Fortunately, I don't have to put my money where my mouth is, since
>someone has already done it for me. :) Vladimir Marangozov has
>modified the Python interpreter (as of 1.5) to use the GNU computed
>labels extension and implemented direct threading.

Would making this standard require a Python installation to use a GNU
compiler?  Or is this just a library that can be linked in?
-- 
                      --- Aahz (Copyright 2000 by aahz at pobox.com)

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

Ninth Virtual Anniversary: 5 days and counting


From z3penguin at penguinpowered.com  Tue Dec  5 15:14:37 2000
From: z3penguin at penguinpowered.com (Z 3 Penguin)
Date: Tue, 5 Dec 2000 15:14:37 -0500
Subject: cursors.echo()
References: <kLOW5.73$%c4.27797@typhoon.aracnet.com>
Message-ID: <90jibj$kdn$1@bob.news.rcn.net>

>From an annoucement by Fredrik a couple of weeks ago:

The Console module is a simple console driver, currently available
for Windows 9X/NT/2000.  This module provides cursor-addressable
text output, plus full support for mouse and keyboard input.

The 001114 release comes with source code, prebuilt binaries for
Python 1.5 and 2.0, documentation, and a Python-style license.

Get your copy from:

    http://effbot.org/efflib/console

enjoy /F



From ullrich at math.okstate.edu  Tue Dec 26 11:47:59 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Tue, 26 Dec 2000 16:47:59 GMT
Subject: A Suggestion for Python Colon Syntax
References: <20001224010638.J5515@xs4all.nl> <mailman.977685437.23291.python-list@python.org>
Message-ID: <92ai3u$ise$1@nnrp1.deja.com>

In article <mailman.977685437.23291.python-list at python.org>,
  "Tim Peters" <tim.one at home.com> wrote:
> [Tim]
> > [...] (ditto the whole "while 1" thread
> > that's been going on virtually non-stop for a decade).
>
> [David C.Ullrich]
> > and we'll never be certain whether it was an intentional
> > joke or not. Keen.
>
> [Thomas Wouters]
> > No, I'm certain Tim did not make that horrible, horrible pun by
> > accident. No way could he inflict such amounts of nausea and pain
> > by accident. Just not possible :)
>
> Winkery seems to irk some people so much that I have been
experimenting with
> leaving it out half the time.  It's a Bad Idea, I'm afraid.  Witness
poor
> Greg Ewing, who recently got beat up even in the presence of an
explicit
> ";-)" thingie.  So I resolve to redouble my explicit winking efforts
in
> 2001,

That's too bad. (I mean really: Take for example a little while
ago in another place, when I was a bit exasperated at someone
who seemed to want the group to do _everything_ for him. I
stated that it wasn't possible to write a function in Pascal
that would determine whether a filename matched a filter like
*.this or that?.ext (the point being sort of it _must_ be
impossible, because if it were possible then all he had to do was
do it and his problem was solved.)  If I'd included an explicit
wank it woulda spoiled it, imo. Certainly would have spoiled the
fun of reading the replies from people who thought I was serious...)

> and  those who can detect humor without it will just have to
live with
> the clues

Wait, was that a joke? I'm not certain.

><wink>.

yeah, I guess it was.

DU

> "while-1"-is-a-wonderfully-self-referential-thread-though-ly y'rs  -
tim
>
>

--
Oh, dejanews lets you add a sig - that's useful...


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


From tim.one at home.com  Tue Dec 26 13:59:37 2000
From: tim.one at home.com (Tim Peters)
Date: Tue, 26 Dec 2000 13:59:37 -0500
Subject: One Python 2.1 idea
In-Reply-To: <92ab9u$dou$1@nnrp1.deja.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEIGIFAA.tim.one@home.com>

[Russell Turpin]
> Has there been some discussion about caching? If done
> well, the vast majority of references would reach
> their object's attribute or method for the cost of a
> couple of integer compares and array fetches. This
> should be an order of magnitude faster than string
> lookup in a dictionary. This especially benefits
> method invocation, since that costs a failed lookup
> to the instance dictionary, and a successful lookup
> in the class dictionary. That's almost enough to
> scare me away from methods!

Martin von L?wis plugged away at this and reported on his experience at the
1998 Python Conference.  Go to

  http://www.foretec.com/python/workshops/1998-11/proceedings.html

and look for "Virtual Methods Tables in Python".  The paper is available
online in HTML form, or downloadable in postscript.

The results were inconclusive, and the reasons for why the results were
inconclusive were also muddy <wink>.  Two gotchas:

1. Methods and attributes can be created or destroyed on the fly, and indeed
even the class hierarchy can mutate in arbitrary ways at runtime.  For that
reason, Martin based his experiments on ideas borrowed from a recent
Smalltalk optimization paper (reference [2] in Martin's paper).  It's a
complex scheme.

2. As you say:

> Most work in a Python program involves object methods
> or attributes. (At least, this is the case in my
> programs. ;-) Increasing the speed of reference to
> these should have appreciable benefit, across the
> board.

Because it is so speed-critical, the dict-based lookup has been optimized
repeatedly over the years -- it's not as slow as you might guess at first.
Or, it might be slower, depending on your first guess <wink>.  There
certainly seems to be great potential here, but, alas, so far nothing has
tapped into it convincingly.

the-pursuit-of-speed-is-a-cruel-mistress-ly y'rs  - tim




From dragon at hitech.net.au  Wed Dec 13 19:30:02 2000
From: dragon at hitech.net.au (dragon at hitech.net.au)
Date: Thu, 14 Dec 2000 00:30:02 GMT
Subject: Python for REAL Projects ?
Message-ID: <3a380d1e.409238@bendigo.news.origin.net.au>

Hello Python People !
I have been looking at, and trying out Python (under Win NT) and
although Python SEEMS interesting can it "cut the mustad" on a real
client project ??

Curently, I develop using DELPHI + Interbase SQL ( or Borland BDE
tables) because , in the main, DELPHI offers "mostly" easy to setup
forms and access to database tables ( not perfect but usable).

FORMS are a major feature of DELPHI and very complex forms may be
built very rapidly  with AUTOMATIC code stub attachments for component
events, ie place the component on the form, 'enter' the object
inspector for that component and write code for the desired  event.
Also your code is not destroyed if you decide to change form layout,
etc. Some systems that have a GUI interface as a addon ( Tkinter ?)
are a real mess to maintain code to form relationships, ie change the
form the code falls apart, change the code and the form fails. !!
 ( JAVA is a good (bad!) example of this )

Maintanence of forms calling forms calling forms and returning via the
call sequence is a non issue under Delphi, other systems I have looked
at require considerable
programming effort just to maintain form relationships, ie: the called
form needs to 'remember" the caller for correct return sequence,
DELPHI looks after this important task automatically. 

How does TKinter handle this ?

For example, a current Delphi project , has 52 forms, 55 delphi code
units, and 18 tables and some 15 reports, with many cross
relationships.

DELPHI is not without its problems however !! :-)

Can this type of project be sucessfully tackled with Python + Tkinter
+ MySQL ??



From dsh8290 at rit.edu  Tue Dec 26 18:24:20 2000
From: dsh8290 at rit.edu (D-Man)
Date: Tue, 26 Dec 2000 18:24:20 -0500
Subject: Some Python 2.1 ideas
In-Reply-To: <3A479C29.2631F27B@uswest.net>; from pbjorn@uswest.net on Mon, Dec 25, 2000 at 12:12:41PM -0700
References: <mailman.977590932.19335.python-list@python.org> <Cw616.23054$qN6.3308205@typhoon.we.rr.com> <3A452881.7686E717@uswest.net> <%vb16.187233$U46.5801214@news1.sttls1.wa.home.com> <3A464D5D.B541C04E@uswest.net> <mailman.977688488.25438.python-list@python.org> <20001225150250.6FA06A84F@darjeeling.zadka.site.co.il> <mailman.977760789.10224.python-list@python.org> <roy-EA8C3B.11361325122000@news.panix.com> <3A479C29.2631F27B@uswest.net>
Message-ID: <20001226182420.D8309@westsidecnc.rh.rit.edu>

On Mon, Dec 25, 2000 at 12:12:41PM -0700, Bjorn Pettersen wrote:
> I think this is a good idea by itself and should be implemented for all of
> strip, lstrip, and rstrip, however it doesn't solve the problem we want to
> solve with chomp (removing a line terminator, whatever the line terminator
> happens to be).

Java has a solution to this:

import sys
sys.line_terminator

> 
> -- bjorn
> 

-D




From aleaxit at yahoo.com  Wed Dec 13 07:27:28 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 13 Dec 2000 13:27:28 +0100
Subject: Label-Value (was: Re: Inheriting the @ sign from Ruby)
References: <Pine.GSO.4.21.0012120138090.2874-100000@y.glue.umd.edu> <r0py9xl1k2n.fsf@black.cs.washington.edu> <3A368E11.20C17E92@san.rr.com> <r0pwvd51gga.fsf@black.cs.washington.edu> <9165he$lon$1@panix3.panix.com>
Message-ID: <917pvv02clr@news2.newsguy.com>

"Aahz Maruch" <aahz at panix.com> wrote in message
news:9165he$lon$1 at panix3.panix.com...
> In article <r0pwvd51gga.fsf at black.cs.washington.edu>,
> Daniel Wood  <daniel at cs.washington.edu> wrote:
> >
> >I'm talking about an *object* that is both an int and mutable.
> >
> >class MutableInt:
> > def __init__( self, default ):
> > self.value = default
> > def set( self, val ):
> > self.value = val
> > def get( self ):
> > return self.value
> > # And maybe
> > def __int__( self ):
> > return self.value
> >
> >My question is: am I confused if I think that this might be useful?
> >And if it is useful then what's the most pythonic way to do this.
> >(And if the best thing is a list of one element, why isn't that much
> >more "hack" than "idiom"?)
>
> This is useful and you've just used the most Pythonic mechanism.

Actually, removing the 'maybe' __int__, one gets full generality
and simplicity -- no specific-type-being-held implications.  One
useful variation:

class Mutable:
    def __init__(self, default=None):
        self.value = default
    def set(self, value):
        self.value = value
        return value
    def get(self):
        return self.value


The 'return value' at the end of the set method is intended to
make this the answer to the perennial 'why isn't assignment an
expression in Python and what can I do about it', by the way:

    line = Mutable()

    while line.set(myfile.readline()):
        frobozz(line.get())

or even:

    temp = Mutable()
    if temp.set(complicated(stuff)) > threshold:
        elaborate_processing(temp.get())

as an alternative to the more normal:

    temp = complicated(stuff)
    if temp > threshold:
        elaborate_processing(temp)


No doubt class Mutable might be 'usefully' (?) enhanced by all
sorts of delegations to self.value.  Here, for example, is
a version that is rather 'transparent' with respect to a
lot of operations:

class TransparentMutable:
    def __init__(self, default=None):
        self.__value = default
    def set(self, value):
        self.__value = value
        return value
    def get(self):
        return self.__value
    def __coerce__(self,other):
        return coerce(self.__value,other)
    def __str__(self):
        return str(self.__value)
    def __repr__(self):
        return repr(self.__value)
    def __getattr__(self, name):
        return getattr(self.__value, name)
    def __setattr__(self, name, value):
        if name=='_Mutable__value':
            self.__dict__[name]=value
        else: return setattr(self.__value, name, value)
    def __getitem__(self, index):
        return self.__value[index]
    def __setitem__(self, index, value):
        self.__value[index] = value


For many uses, an instance of TransparentMutable will now be
almost interchangeable with whatever its __value is holding
at this time...:

>>> m=TransparentMutable()
>>> m.set("loopy")
'loopy'
>>> m+'plop'
'loopyplop'
>>> 'plop'+m
'ploploopy'
>>> for x in m:
...     print x
...
l
o
o
p
y
>>> m.set(3.1415926)
3.1415926000000001
>>> (4.0/3)*m*(1.1*1.1*1.1)
5.5752796674666678
>>>

note that we don't need to define the various operators, as
__coerce__ comes in handy here -- unless we try to operate
on TWO 'transparent mutables', which rather spoils things:-).


Is it worth pursuing this 'rich', aka 'transparent',
approach?  I don't think it is.  The simple Mutable class,
which makes no attempt whatsoever to pass itself off as
something it isn't, has potentially high value for its
simplicity -- and the need to explicitly call .get to
obtain the value it's currently wrapping is no bad thing,
either (explicit is better than implicit...!).

'TransparentMutable' has only one hypothetical advantage:
you can try to tweak it enough to be able to 'drop' it
into existing code without reworking that code -- maybe.
And then again, maybe there could be some little anomaly
here or there (multiplying two TransparentMutable's rather
than one of those and one true number, for example) which
might cause trouble.

I think refactoring the 'existing code' (perhaps using
the simple Mutable) is going to yield better returns than
hacking on TransparentMutable or any equivalent thereof.

Remember Firmitas, Utilitas, Venustas?  Well, Venustas,
beauty, is in the eye of the beholder.  Mutable, the simple
kind, has excellent Firmitas, solidity -- it's simple enough
that it obviously has no defects.  It does the single small
job it's designed to do: letting you 'set this value aside
for later use' as a part of an expression, and get the value
back again; explicitly for each half of the job; period.

The defect of TransparentMutable (and other hacks to that
effect) is that it tries to put Utilitas, convenience, AHEAD
of Firmitas; it strives so hard to be 'convenient to use',
that it becomes somewhat complicated, and therefore fragile.

Listing "comodit?" before "perpetuit?", as I argued in a
post yesterday on a different thread ('convenience' before
'solidity'; Utilitas before Firmitas), was an ideological
aberration (originally by L.B.Alberti, later by Palladio)
wrt Vitruvius' original formulation.  They didn't really
practice what they preached, IMHO: the Tempio Malatestiano
or the Villa Foscari do not in the least compromise
structural solidity in the quest for convenience or beauty.

Maybe they took solidity _for granted_, because, in their
(Renaissance) times and in their (Architecture) calling,
compromises regarding solidity were simply unthinkable.
Well, we're not so lucky, in the software field, today; the
Firmitas of *by far* most software around is imperfect.

We *must* live by "do the simplest thing that can possibly
work" -- give solidity its proper, foremost place.  One
of the debilitating factor for much current software is
a misplaced emphasis on assumed 'convenience' (funky GUIs,
quirky shortcuts, special cases aplenty) to the detriment
of solidity.  A small but crucial step to reverse this
trend, is to start by putting the order right once more...
the way Vitruvius had it!-)


Alex





From phd at phd.pp.ru  Thu Dec 28 06:22:46 2000
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 28 Dec 2000 14:22:46 +0300 (MSK)
Subject: Is this a BUG?
In-Reply-To: <3A4B0C5E.1E7976D2@wielun.dhs.org>
Message-ID: <Pine.LNX.4.30.0012281422120.572-100000@fep132.fep.ru>

On Thu, 28 Dec 2000, QdlatY wrote:
> print "!",form["index"].value,"!"
> FieldStorage(None, None, [MiniFieldStorage('index', '1')])
> ! 1 !
> ----------------
>  ^ ^
> Why the hell there are spaces? index property has value '1'!
> not ' 1 '?!

   "print" insert spaces on ",". Use sys.stdout.write instead.

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




From ckrill at qvlinc.com  Tue Dec 12 19:40:45 2000
From: ckrill at qvlinc.com (Coy Krill)
Date: Tue, 12 Dec 2000 16:40:45 -0800
Subject: Recommendations for a 2.0 book?
References: <slrn93de23.eum.grey@teleute.rpglink.com>
Message-ID: <3A36C58D.DB6F94CC@qvlinc.com>

Yes, just replace your copy of Beazley and if you're in the Everett, WA area,
Half-Price books just so happened to have gotten a couple of them and had them
at...you guessed it.. half price so you could save a few dollars.

Coy Krill

Steve Lamb wrote:

>     OK, just a little while ago I had dabbled in Python 1.5.2 and was using
> the Beazley book as my reference.  I've since lost the book and since I'm not
> far enough along to remember most of the language (that is months of actual
> use away) I am now pondering a replacement.  What is a good replacement now
> that we're on Python 2.0?  Could I get another copy of the Beazley book and be
> able to switch between 1.5.2 and 2.0 fairly easily?
>
> --
>          Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
>          ICQ: 5107343          | main connection to the switchboard of souls.
> -------------------------------+---------------------------------------------



From aleaxit at yahoo.com  Tue Dec 19 09:12:08 2000
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 19 Dec 2000 15:12:08 +0100
Subject: while true: !!!
References: <B43D149A9AB2D411971300B0D03D7E8B0FBCBB@natasha.auslabs.avaya.com> <mailman.977115541.28453.python-list@python.org> <qwu%5.8061$KY1.22030@news1.rivrw1.nsw.optushome.com.au>
Message-ID: <91nqcg018tj@news2.newsguy.com>

"deadmeat" <root@[127.0.0.1]> wrote in message
news:qwu%5.8061$KY1.22030 at news1.rivrw1.nsw.optushome.com.au...
> > This is precisely where having assigment as an expression (like C/C++)
> > is useful :
> > while ( (line = file.readline()) != None ) :
>
> yuk.. I'd prefer the Pascal way to C's
>
> while file.readline(line) do
>  // whatever
>
> however I don't know if variables passed to Python functions can be
altered?

The variable will refer to the same object, before and after the
call, no matter what the function does.  The object itself may
be changed, if it's a mutable object.

If the object that 'line' refers to is a string (immutable, in
Python), then you surely cannot do what you desire.

> however the file reading thing is just one example of while 1: so being
able
> to avoid it in this case is hardly any breakthough.

Of course.  The general idiom, as Knuth said in 1974, is:

    while 1:

        "several" (0 or more)
        lines
        of
        useful
        code

        if condition: break

        "several" (0 or more)
        other
        interesting
        lines

Unless you're in a special case where either block just
happens to collapse to 0 lines, you can try to hide and
obscure the natural structure of your algorithm (hardly
a clever strategy, that!), but that will not change the
way the algorithm WANTS to be coded.  In Python, it's
not even particularly interesting if the _second_ block
collapses to length 0, as Python has no test-at-end loop,
which is an excellent simplification; only if the FIRST
of the two blocks is zero-length does the
    while 1:
        if condition: break
            etc
_naturally_ become
    while not condition:
        etc


Alex





From grante at visi.com  Wed Dec  6 12:50:50 2000
From: grante at visi.com (Grant Edwards)
Date: Wed, 06 Dec 2000 17:50:50 GMT
Subject: update
References: <3A2D9A08.43F8FBE8@hypermath.org> <8sds2t8hfgq54ad8ka43te25hvgcv6f4ot@4ax.com>
Message-ID: <_%uX5.693$TU6.21717@ptah.visi.com>

In article <8sds2t8hfgq54ad8ka43te25hvgcv6f4ot at 4ax.com>, Dale Strickland-Clark wrote:
>leonrom <leonrom at hypermath.org> wrote:
>
>>I run v.1.5.2, do I have to uninstall it before installing
>>v.2.0
>
>I haven't seen recommendations one way or the other elsewhere
>so here's one I made earlier:
>
>I recommend uninstalling the old version before intalling the
>new one.

I don't know if that's a practical suggestion for users of RH
Linux (and some others).  There are system utilities that
depend on 1.5.2 being installed, and the system will complain
pretty loudly if you try to uninstall 1.5.2.  You can ignore
the complaints and uninstall 1.5.2 despite system dependancies,
but I haven't been brave enough to try it.

Has anybody successfully replaced 1.5.2 with 2.0 on a RH[67]
box?

-- 
Grant Edwards                   grante             Yow!  .. the MYSTERIANS are
                                  at               in here with my CORDUROY
                               visi.com            SOAP DISH!!


From peter at engcorp.com  Fri Dec 15 00:29:34 2000
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 15 Dec 2000 00:29:34 -0500
Subject: Python advocacy
References: <91bln3$lmr$1@nnrp1.deja.com>
Message-ID: <3A39AC3E.FA75C4FF@engcorp.com>

jschmitt at vmlabs.com wrote:
> 
> I just read this good article about Perl advocacy.  I think it applies
> to Python equally.  I found the link mentioned from /. this morning.
> 
> http://www.perl.com/pub/2000/12/advocacy.html?wwwrrr_20001213.txt [snip]

I didn't bother reading the whole thing because I noticed early on (and
confirmed throughout) that the author is oblivious to the huge
investment people put into developing systems in a particular language
(not to mention the investment in *learning* that language) and the fact
that as long as the language stays popular, improves, and continues to
serve their needs well, they get a payoff from that investment.

If the language "dies", that investment is largely lost.

The theory is that by advocating a language one can increase the
likelihood of its survival and improvement.  If true, then advocating a
language directly serves ones own needs.

There is no reason to focus on the fact that people appear to take this
sort of thing personally.  That's probably just human nature, and likely
not even a Bad Thing.  The point is there are good and wholly rationale
reasons for advocacy even in its current style, and you could say that
anyone who does *not* advocate a particular language is contributing to
its demise (the "if you're not part of the solution, you're part of the
problem" theory of life :-).

I think I'll just ignore whatever other, possibly valid points
Mark-Jason Dominus may have made and carry on advocating as I always
have.  He hasn't convinced me.


From moshez at zadka.site.co.il  Mon Dec 25 13:25:20 2000
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Mon, 25 Dec 2000 20:25:20 +0200 (IST)
Subject: classes (was Re: Same again please for OOP)
In-Reply-To: <92708d0g4g@news1.newsguy.com>
References: <92708d0g4g@news1.newsguy.com>, <mailman.977623149.8184.python-list@python.org> <9257mf0sl1@news1.newsguy.com> <roy-E5B925.13324224122000@news.panix.com> <925khi019hs@news1.newsguy.com> <kubsu1t81x.fsf@lasipalatsi.fi>
Message-ID: <20001225182520.C7B2EA84F@darjeeling.zadka.site.co.il>

On Mon, 25 Dec 2000, "Alex Martelli" <aleaxit at yahoo.com> wrote:

> Other languages may well-nigh force one to frame
> things up in terms of accessors/mutators, but that's an
> issue with *those* languages -- not one we should copy
> in Python, which lets us be more direct and simple.

You'd be right, if Python's __getattr__ would be better. There are a couple
of PEPs on the issue (Barry's, which AFAIK was rejected, and Paul's), just
to show many people in the Python's development team are dissatisfied. If
you've ever written a __getattr__ method, you probably know what a pain it
is. In contrast to the rest of Python, where I can code up 200-line application
with something on the order of 2 (easily found) bugs, __getattr__ always
seems to lead to twisted recursion bugs, and all kinds of headaches. 
*Particularily* if you also mix in __setattr__.

I've learned the hard way not to write __getattr__ methods. Currently, none
of my code contains them. True gurus, like Gordon McMillan, can use __getattr__
(e.g., in filecmp.py), to get lazy evaluation of attributes, but I tend to
fear those things.

The "computed attributes" interface needs to be improved befroe Python can
be said to be not one of *those* languages.

-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!



From fredrik at effbot.org  Sun Dec 31 09:22:09 2000
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sun, 31 Dec 2000 14:22:09 GMT
Subject: how to flatten & lump lists
References: <92ne7e$5fj$1@nnrp1.deja.com>
Message-ID: <liH36.2811$Qb7.324901@newsb.telia.net>

Kevin Parks wrote:
> I want to be able to flatten a list and i could have sworn that i saw
> that there was something like flatten() built in someplace but i can't
> find it now anywhere in the docs.

> [1, 2, 3, 4, 5, 6, 7, [6, 5, 4, 3, 2, 1]]
>
> how can i get: [1, 2, 3, 4, 5, 6, 7, 6, 5, 4, 3, 2, 1]?

In this case, the easiest solution is to use extend (insert
sequence at end of list) instead of append.  Just change
your code from

    l.append(foo)   #stick it on the end of the other list
    #we need to flatten it some how in to a single list

to

    l.extend(foo)   #stick it on the end of the other list

:::

To flatten an existing sequence, you can use something
like this:

from types import TupleType, ListType

def flatten(seq):
    res = []
    for item in seq:
        if type(item) in (TupleType, ListType):
            res.extend(flatten(item))
        else:
            res.append(item)
    return res

:::

It's time to throw some hammers [1].  See you all next
year!

</F>

1) http://www.msnbc.com/news/508442.asp




From victor at prodigy.net  Wed Dec 27 20:04:34 2000
From: victor at prodigy.net (Victor Muslin)
Date: Thu, 28 Dec 2000 01:04:34 GMT
Subject: fcgi.py
References: <3a4a865c.1325844015@localhost>
Message-ID: <3a4a90ab.1328483109@localhost>

I just found an old posting that says that socket.fromfd() is not
implemented on Windows platform (though Python documentation does not
say so). Is there a FastCGI solution for Python for WindowsNT???

Thanks.

On Thu, 28 Dec 2000 00:32:19 GMT, victor at prodigy.net (Victor Muslin)
wrote:

>Sorry for newbie questions, but I ran into a couple of problems with
>fcgi.py. I am using Python 1.5.2 on WindowsNT 4.0.
>
>1) I tried to run fcgi.py from the command line (to exercise its
>_test() function) and received the following error message:
>
>	D:/Apps/apache/cgi-bin> fcgi.py
>	Exception exceptions.AttributeError: 'err' in <method
>FCGI.__del__ of FCGI instance at 82a220> ignored
>
>The traceback was:
>
>	Traceback (innermost last):
>	  File "fcgi.py", line 386, in _test
>	    req = Accept()
>	  File "fcgi.py", line 231, in __init__
>	    _startup()
>	  File "fcgi.py", line 367, in _startup
>	    s=socket.fromfd(sys.stdin.fileno(), socket.AF_INET,
>	AttributeError: fromfd
>
>Seems that the socket module for WindowsNT Python 1.5.2 does not have
>fromfd() method.
>
>Here's the proof (at least for my Python installation):
>
>>>> pprint.pprint(dir(socket))
>['AF_INET',
> 'INADDR_ALLHOSTS_GROUP',
> 'INADDR_ANY',
> 'INADDR_BROADCAST',
> 'INADDR_LOOPBACK',
> 'INADDR_MAX_LOCAL_GROUP',
> 'INADDR_NONE',
> 'INADDR_UNSPEC_GROUP',
> 'IPPORT_RESERVED',
> 'IPPORT_USERRESERVED',
> 'IPPROTO_GGP',
> 'IPPROTO_ICMP',
> 'IPPROTO_IDP',
> 'IPPROTO_IGMP',
> 'IPPROTO_IP',
> 'IPPROTO_MAX',
> 'IPPROTO_ND',
> 'IPPROTO_PUP',
> 'IPPROTO_RAW',
> 'IPPROTO_TCP',
> 'IPPROTO_UDP',
> 'IP_ADD_MEMBERSHIP',
> 'IP_DROP_MEMBERSHIP',
> 'IP_MULTICAST_IF',
> 'IP_MULTICAST_LOOP',
> 'IP_MULTICAST_TTL',
> 'IP_OPTIONS',
> 'IP_TOS',
> 'IP_TTL',
> 'MSG_DONTROUTE',
> 'MSG_OOB',
> 'MSG_PEEK',
> 'SOCK_DGRAM',
> 'SOCK_RAW',
> 'SOCK_RDM',
> 'SOCK_SEQPACKET',
> 'SOCK_STREAM',
> 'SOL_SOCKET',
> 'SOMAXCONN',
> 'SO_ACCEPTCONN',
> 'SO_BROADCAST',
> 'SO_DEBUG',
> 'SO_DONTROUTE',
> 'SO_ERROR',
> 'SO_KEEPALIVE',
> 'SO_LINGER',
> 'SO_OOBINLINE',
> 'SO_RCVBUF',
> 'SO_RCVLOWAT',
> 'SO_RCVTIMEO',
> 'SO_REUSEADDR',
> 'SO_SNDBUF',
> 'SO_SNDLOWAT',
> 'SO_SNDTIMEO',
> 'SO_TYPE',
> 'SO_USELOOPBACK',
> 'SocketType',
> '__builtins__',
> '__doc__',
> '__file__',
> '__name__',
> '_fileobject',
> '_realsocketcall',
> '_socketobject',
> 'error',
> 'errorTab',
> 'gethostbyaddr',
> 'gethostbyname',
> 'gethostbyname_ex',
> 'gethostname',
> 'getprotobyname',
> 'getservbyname',
> 'htonl',
> 'htons',
> 'ntohl',
> 'ntohs',
> 'socket']
>>>>
>
>socket.fromfd() method is called in the fcgi._startup() function,
>which is called in the constructor of FCGI object. Since
>fcgi._startup() function is catching only socket.error exceptions
>instead of AttributeError exception the program dumps.
>
>If AttributeError except clause is added immediately after the
>socket.error() in fcgi._startup():
>
>    except AttributeError:
>            global _isFCGI
>            _isFCGI = 0
>            return
>
>then the program works fine.
>
>Is there something I am doing wrong or is there something wrong with
>my Python installation?
>
>2) I was wondering whether it is possible to use fcgi.py based
>programs with Apache using cgi-fcgi brdige. If so, how would one
>configure cgi-fcgi to know to launch a python interpreter with a
>specific program to interpret? If there is an example it would be
>greatly appreciated.
>
>Thanks



From jdamunds at coretec-inc.com  Sat Dec  9 10:55:14 2000
From: jdamunds at coretec-inc.com (James Amundson)
Date: Sat, 09 Dec 2000 10:55:14 -0500
Subject: python 2.0
Message-ID: <3A3255E2.DC54424B@coretec-inc.com>

Does anyone know where I can download a binary
version of python 2.0 for HP-UX 10.20.



Thanks,
James.




From nospam at caltech.edu  Wed Dec  6 20:06:55 2000
From: nospam at caltech.edu (Nathan Gray)
Date: Wed, 06 Dec 2000 17:06:55 -0800
Subject: Python Origins
References: <90k3pq$50p$1@nnrp1.deja.com> <tcmay-A67EAA.18361005122000@news.got.net>
Message-ID: <3A2EE2AF.AD8A51E2@caltech.edu>

What a rude response to a polite and legitimate question.

Tim May wrote:

> If you are doing research, then you should surely understand that the
> large amount of material already on the Web, and findable with standard
> search engines, is a much better source of information than asking
> people to custom-write material for you. Or to do your searches for you.

Ah yes, this is why the best historians rely on web searches exclusively for
their research.  There's no point in _asking_ people for information.  If
information exists, it will be on the internet!  Talking to experts is foolish
and redundant.

> A handful of minutes spent with Google will yield you far more than
> waiting and waiting for people here to write histories for you.

You see, we're a bunch of arrogant twits who wouldn't bend over to help our
own mothers if they tripped and fell.  What made you think that people in
comp.lang.python would answer questions about Python?  We have better things
to discuss.

> Why are you "doing some research" if you are this clueless?

How dare you seek to educate yourself!?  Impudent wretch!

It's really upsetting how often I've seen this sort of response to a polite
question on this newsgroup lately.  Tell me, what's so terrible with asking a
group of experts for pointers when researching a subject?  It's a perfectly
legitimate and highly effective way to learn about something.  There's plenty
of information out there that hasn't been put on a web page, and the stuff
that _has_ been put online is often out of date or inaccurate.  Information
from the experts -- even if it's simply confirmation that good web resources
exist -- can be invaluable if you care in the least about the accuracy of your
information.

If you weren't going to help why did you even bother to reply?  One of the
great merits of the Python community has always been the willingness of people
on this newsgroup to help others with their questions.    Since you're so
savvy why not just help the guy out instead of fueling up the flamethrower?
--
Nathaniel Gray
n8gray <at> caltech <dot> edu
California Institute of Technology
Computation and Neural Systems




From mdtompkins at home.com  Wed Dec  6 12:20:46 2000
From: mdtompkins at home.com (Mark Tompkins)
Date: Wed, 06 Dec 2000 17:20:46 +0000
Subject: win32security: setting process privileges
References: <3A2D547F.E4FF9245@home.com>
Message-ID: <3A2E756E.D62711D4@home.com>

Found it.

This may be of interest to others:

The advanced user rights are as follows:

To Act as Part of the Operating System

SE_TCB_NAME

Bypass Traverse Checking

SE_CHANGE_NOTIFY_NAME

Replace a Process Level Token

SE_ASSIGNPRIMARYTOKEN_NAME


Mark Tompkins wrote:

> Hi,
>
> I'm trying to set process privileges in a thread that needs to call
> win32security.LogonUser.
>
> I need to enable
>
> SE_TCB_NAME
> SE_CHANGE_NOTIFY_NAME
> SE_ASSIGNPRIMARYTOKEN_NAME
>
> When I attempt to assign these privileges to the current thread, only
> one assignment succeeds (SE_CHANGE_NOTIFY_NAME).  The MSDK win32 api
> docs INDICATE that SE_CHANGE_NOTIFY_NAME is assigned by default.  The
> following example is for SE_ASSIGNPRIMARYTOKEN_NAME.
>
> >>>import win32security
> >>>import ntsecuritycon
> >>>import win32con
> >>>import win32api
> >>>
> win32security.LookupPrivilegeValue(None,ntsecuritycon.SE_CHANGE_NOTIFY_NAME))
>
> 7
> >>>
> win32security.LookupPrivilegeValue(None,ntsecuritycon.SE_ASSIGNPRIMARYTOKEN_NAME)
>
> 23
> >>>
> win32security.LookupPrivilegeValue(None,ntsecuritycon.SE_ASSIGNPRIMARYTOKEN_NAME)
>
> 3
> >>> flags=ntsecuritycon.TOKEN_ADJUST_PRIVILEGES |
> ntsecuritycon.TOKEN_QUERY
> >>> flags
> 40
> >>>id=win32security.LookupPrivilegeValue(None,ntsecuritycon.SE_ASSIGNPRIMARYTOKEN_NAME)
>
> >>> id
> 3
> >>> nP=[(id,ntsecuritycon.SE_PRIVILEGE_ENABLED)]
> >>> nP
> [(3, 2)]
> >>>
> htoken=win32security.OpenProcessToken(win32api.GetCurrentProcess(),flags)
>
> >>> htoken
> 268
>
> >>> win32security.AdjustTokenPrivileges(htoken,0,nP)
> Traceback (innermost last):
>   File "<interactive input>", line 1, in ?
> api_error: (1300, 'MyAdjustTokenPrivileges', 'Not all privileges
> referenced are assigned to the caller.')
> >>>
>
> What privileges do I as a user need to be assigned, in order that the
> other two privileges are enabled for me, or are grantable by me to a
> process?  I am currently a member of the administrators group on the
> machine that I am coding on.
>
> thx
>
> Mark



From ngps at madcap.dyndns.org  Thu Dec 28 10:41:40 2000
From: ngps at madcap.dyndns.org (Ng Pheng Siong)
Date: 28 Dec 2000 15:41:40 GMT
Subject: sniffing btw my web browser and a server
References: <92aush$rhc$1@news.nuri.net> <92dgju$60f$1@saltmine.radix.net>
Message-ID: <92fmvk$stt$1@violet.singnet.com.sg>

According to Cary O'Brien <cobrien at Radix.Net>:
> Take a look at "ethereal".  It is a (GTK?) graphical packet sniffer that
> unpacks many protocols.  You can use it to snapshot an HTTP session and use
> the "TCP Session" option to see exactly what is going on.
> 
>    http://www.ethereal.com

Also, take note of the recent security advisories on ethereal.

<surfing over to the web site...>

Oh, ok, they mention this on the front page. ;-)


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



From emile at fenx.com  Fri Dec 22 11:41:01 2000
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 22 Dec 2000 08:41:01 -0800
Subject: Newbie Question
References: <05m64tkqev0aavea80ttjse1sj2gd447hg@4ax.com> <hqs64tc5jjndav3p6nvmu67t7hqaf5go00@4ax.com> <91vuj5$5n7is$1@ID-11957.news.dfncis.de> <42074t8a8e8604sgl77egadao721tk62ik@4ax.com>
Message-ID: <92005r$5e67v$1@ID-11957.news.dfncis.de>

I'm still not sure, as this works as well:

import os

class Class:
    def __init__(self, root):
        self.root = root
    def func(self, root, files):
        print self
        print root
        print files
    def run(self):
        os.path.walk(self.root, Class.func, arg)

root = r'C:\etvwin'
arg = Class(root)

arg.run()

--

Emile van Sebille
emile at fenx.com
-------------------


"Bob W." <removethis_kranki2 at earthlink.net> wrote in message
news:42074t8a8e8604sgl77egadao721tk62ik at 4ax.com...
> Being really new, I am not certain about the next
statement.  However,
> it seemed to me that the difference might be that you are
executing
> the walk() in the global namespace rather than the address
space of
> the class.  Thanks for your input.
>





From boud at rempt.xs4all.nl  Fri Dec  8 01:21:04 2000
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 8 Dec 2000 06:21:04 GMT
Subject: MVC module anyone?
References: <90nnq6$8n0$1@news.rz.uni-karlsruhe.de> <6tSX5.767919$8u4.13380209@news1.rdc1.bc.home.com>
Message-ID: <90pukg$mo4$1@news1.xs4all.nl>

Dethe Elza <delza at antarcti.ca> wrote:

> libglade xml for GTK+ and GNOME (http://www.daa.com.au/~james/gnome/), hard
> to find documentation on the XML format itself
> BML (http://www.alphaworks.ibm.com/tech/bml), for Java/Swing
> I don't know of an xml dialect for describing/creating Tkinter interfaces,
> but there is one in the works for wxWindows.  No URL last time I checked.

PyQt forms can also be described in an XML format with Qt
Designer.  There's already an xml-to-python compiler for it: pyuic:
http://www.thekompany.com/projects/pykde/. And it works both on Windows
and on X11.



-- 

Boudewijn Rempt  | http://www.valdyas.org


From dalke at acm.org  Thu Dec 14 23:49:23 2000
From: dalke at acm.org (Andrew Dalke)
Date: Thu, 14 Dec 2000 21:49:23 -0700
Subject: Smalltalk and Python
References: <3A2FFB8B.2030408@zoologic.com> <915qlh$89fa$1@nntp6.u.washington.edu> <917g5o$qh$1@news.rz.uni-karlsruhe.de> <55ne3t4lqvijv14bt7nj3b53oovmjf9ql0@4ax.com> <917r2b$4n7$1@news.rz.uni-karlsruhe.de> <Pine.BSF.4.10.10012131113170.59599-100000@tcfreenet.org> <mrPZ5.21134$x6.12200962@news2.rdc2.tx.home.com> <86d7ewrqat.wl@archon.local.idaemons.org> <tcmay-19C054.14031113122000@news.got.net> <tcmay-FCA22F.14073713122000@news.got.net> <918v7a02scn@news1.newsguy.com> <tcmay-04E80C.16181313122000@news.got.net> <91a1ed02a4j@news2.newsguy.com> <3A389B1A.4F33FFD4@cfmu.eurocontrol.be> <91ae8s02llf@news2.newsguy.com> <3A397778.2E6D84D3@cosc.canterbury.ac.nz>
Message-ID: <91c7r4$7a0$1@slb1.atl.mindspring.net>

Greg Ewing wrote:
>the TADS (Text Adventure Development System)
>lets you declare objects with multiple
>classes, e.g.
>
>   bed: Furniture, Surface, Scenery
>      ...
...
>This is really just syntactic sugar,
...
>To mention something vaguely Python-relevant:
>if Python had a convenient syntax for declaring
>singletons, it could be a pretty good language for
>Interactive Fiction...

One packet of syntactic sugar, coming up!

def make_a(*classes):
  class make_a:
    pass
  make_a.__bases__ = tuple(classes)
  return make_a()

class Furniture:
     def comfy(self): return 1

class Surface:
    def genus(self): return 0  # Imagine a spherical cow...

class Scenery:
    def description(self): return "boring"

>>> bed = make_a(Furniture, Surface, Scenery)
>>> bed
<__main__.make_a instance at 0x120330628>
>>> bed.description()
'boring'
>>> bed.genus()
0
>>> bed.comfy()
1
>>>

                    Andrew
                    dalke at acm.org






From pblanchette at pixelsystems.com  Tue Dec  5 13:34:26 2000
From: pblanchette at pixelsystems.com (Patrick Blanchette)
Date: Tue, 05 Dec 2000 18:34:26 GMT
Subject: win32security.AdjustTokenPrivileges
Message-ID: <3A2D3521.6010105@pixelsystems.com>

I,
   I'd like to reboot a windows NT station using a Python script.  It is 
supposed to be possible using the Python extensions for Windows.  I try 
the following code but an exception is thrown after the call to 
AdjustTokenPrivileges:
-----------------------------------------------
hProcess = GetCurrentProcess()
processToken = OpenProcessToken(hProcess, TOKEN_ADJUST_PRIVILEGES | 
TOKEN_QUERY)
procPrivValue = LookupPrivilegeValue (None, SE_SHUTDOWN_NAME)
AdjustTokenPrivileges (processToken, 0, (provValue, SE_PRIVILEGE_ENABLED) )
ExitWindowsEx(EWX_REBOOT | EWX_FORCE, 0)
-->
Traceback (innermost last):
  File "<interactive input>", line 1, in ?
TypeError: A TOKEN_PRIVILEGES object must be a tuple of (LARGE_INTEGER, int)
-----------------------------------------------

How should I use the AdjustTokenPrivileges function?

thank



From rturpin at my-deja.com  Sun Dec 31 20:39:36 2000
From: rturpin at my-deja.com (rturpin at my-deja.com)
Date: Mon, 01 Jan 2001 01:39:36 GMT
Subject: Is 100,000 entries big for a dictionary?
References: <mailman.978306855.25759.python-list@python.org>
Message-ID: <92on4o$2ei$1@nnrp1.deja.com>

In article <mailman.978306855.25759.python-list at python.org>,
  "Tim Peters" <tim.one at home.com> wrote:
> On most machines hash values are 32 bits, and are always
> computed as such, and are cached in the dict entry.
> When it's time to grow, the table does double in size,
> but hash values are not recomputed -- it simply uses one
> more of the hash bits that's always been there.  Note
> that entries are reinserted when the table grows --
> that's probably not what you meant by "the usual
> extensible hashing mechanism". ..

Actually, that describes exactly what I meant. This is
now a pretty standard algorithm used in a variety of
DBMS and other applications. Only the entries whose
next bit is 1 need to be reinserted, though that can
cause other entries to move in the lower half if there
were previous collisions.

Having-implemented-this-a-time-or-two'ly yrs,
Russell


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


From tim.one at home.com  Sun Dec 31 18:54:16 2000
From: tim.one at home.com (Tim Peters)
Date: Sun, 31 Dec 2000 18:54:16 -0500
Subject: Is 100,000 entries big for a dictionary?
In-Reply-To: <92obq8$qf5$1@nnrp1.deja.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCAECHIGAA.tim.one@home.com>

[rturpin at my-deja.com]
> Thank you for the comments. I assume it does the usual
> extensible hashing mechanism, extending the hash value
> by one bit, and the table by a factor of two, when it
> is time to grow?

On most machines hash values are 32 bits, and are always computed as such,
and are cached in the dict entry.  When it's time to grow, the table does
double in size, but hash values are not recomputed -- it simply uses one
more of the hash bits that's always been there.  Note that entries are
reinserted when the table grows -- that's probably not what you meant by
"the usual extensible hashing mechanism".  A table may also shrink.

> I'm glad to read that it uses open probing. I think this makes
> more sense for tables that reside in memory, and that have no
> strong granularity.

So did Guido <wink>.

>> if-you-do-have-frequent-collisions-i'll-consider-it-a-bug-ly y'rs

> What hash function does it use?

Each type defines its own hash method (and your classes can too, by
supplying a __hash__ method), so, sorry, there's no succinct way to answer
that (the tuple hash is not the string hash is not the int hash etc).  Note
that Python dicts use equality, not object identity, so there are some odd
constraints on hash functions; for example, it must be that hash(42) ==
hash(42L) == hash(42.0) == hash(complex(42)).

it's-almost-time-for-you-to-read-the-source-code<0.9-wink>-ly
    y'rs  - tim




From gordon at ozcomputers.net  Sun Dec 31 20:59:33 2000
From: gordon at ozcomputers.net (gordon at ozcomputers.net)
Date: Mon, 01 Jan 2001 01:59:33 GMT
Subject: Spambot Fodder, Dont Read  8845
Message-ID: <9wR36.118702$bp4.1740600@monolith.news.easynet.net>

gordon at ozcomputers.net
xxjphtltwvmxvvcosuqwgloxzrqoqjxcfilqoppglzpvhwzvtbwnidvf



From mballen at erols.com  Sun Dec 31 20:55:04 2000
From: mballen at erols.com (Michael B. Allen)
Date: 1 Jan 2001 01:55:04 GMT
Subject: SMB Module
References: <92n1lh02evv@drn.newsguy.com>
Message-ID: <slrn94vpc7.544.mballen@angus.foo.net>

Armin Steinhoff wrote:
>In article <slrn94trot.b73.mballen at angus.foo.net>, mballen at erols.com says...
>>
>>I frequent the Samba mailing list. A couple of days ago Richard Sharp
>>put together an smbclient.so library of client functions for SMB. SMB
>>is Microsoft's native file and print sharing protocol. Someone wrote a
>>Perl module using it. I just thought one of you guys might want to do the
>>same for Python. Seems only natrual since there's a nice little library
>>all packaged up with open(), read(), write(), ...etc all setup easy like.
>
>Makes no sense, because a SMB client mapps normaly a M$ volume to a directory
>tree (UNIX) or a virtual drive (M$) ... I don't see in that case any tasks for
>Python.

I'm not really sure what your talking about Armin. SMB Servers are file
servers like ftp, http, nfs, ...etc. It just happends to be by far the
most popular.

And if you have people are dedicated to MS, as many are, then having this
tool might be crucial to allowing you to work in a UNIX environment if
that's what you prefer. I do.

Mike

-- 
signature pending


From tim.one at home.com  Sun Dec 31 18:25:08 2000
From: tim.one at home.com (Tim Peters)
Date: Sun, 31 Dec 2000 18:25:08 -0500
Subject: how to flatten & lump lists
In-Reply-To: <92ne7e$5fj$1@nnrp1.deja.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCECGIGAA.tim.one@home.com>

[cpsoct at my-deja.com]

/F covered the flatten part, so let's look at the other one:

> ...
> additionally i would like to do the opposite, take a list like
> [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
>
> and have it lump up into subgroups (randomly) and return:
> [[1],[2],[3, 4, 5], [6, 7], [8, 9, 10]]

Well, the problem seems ill-specified.  Try this for a start:

import random

def clump(seq):
    result = []
    i, n = 0, len(seq)
    # invariant:  result already contains a clumping of seq[:i]
    while i < n:    # while seq[i:] still needs to be clumped
        j = random.randint(i+1, n)  # assuming empty clumps no good
        result.append(seq[i:j])
        i = j
    return result

Then, e.g.,

for i in range(5):
    print clump(range(1, 11))

may print

[[1], [2, 3, 4, 5, 6, 7, 8], [9], [10]]
[[1, 2, 3, 4, 5, 6, 7, 8, 9], [10]]
[[1, 2, 3, 4, 5, 6, 7], [8], [9, 10]]
[[1, 2, 3, 4, 5, 6, 7, 8], [9], [10]]
[[1, 2, 3, 4, 5, 6, 7, 8, 9, 10]]

As that suggests, you're much more likely to get "large" clumps at the start
than at the end with this algorithm.  That may or may not suit your needs --
you have to define "random" in a sense appropriate for your application.

Fair warning:  generating random instances of large structured objects
"fairly" can be extremely subtle.  In this case, if you want every possible
way of clumping to be equally likely, there's a pretty simple way:

def clump(seq):
    if not seq:
        return []
    result = [seq[:1]]
    for e in seq[1:]:
        # either extend last clump or start a new one, with
        # equal odds
        if random.random() < 0.5:
            # extend last clump
            result[-1].append(e)
        else:
            # start a new clump
            result.append([e])
    return result

Proof sketch:  let the total number of ways to clump a sequence of length N
be C(N).  C(1) == 1 by inspection.  Suppose N > 1, and consider the first
element E of the sequence.  Either E is in a clump by itself, or it isn't.
If it is by itself, the rest of the clumping is a clump of a list of N-1
elements, and there are C(N-1) ways to do that.  Else E is not by itself, so
there's something other than E in the first piece of the clump.  Removing E
from the first piece of the clump thus again yields a clumping of a list of
N-1 elements, and again there are C(N-1) ways to do that.  So E is an a
clump by itself in exactly half the cases.

BTW, this shows that C(N)==2**(N-1).  A more direct way to see that is to
picture the elements in a line:

   e0 e1 e2 e3

and draw vertical bars at the clump boundaries.  For example,

    [[e0], [e1, e2], [e3]]

corresponds to

    e0|e1 e2|e3

If there are N elements, there are N-1 positions at which we *could* draw a
vertical bar.  At each position, we can either "draw a bar or not", so there
are 2**(N-1) total ways of choosing which bars to draw.  That's very
directly what the "if" test in the algorithm is doing:  choosing to draw a
bar (start a new clump) or not (extend the last clump) with equal odds.

good-programming-reduces-to-drawing-lines<wink>-ly y'rs  - tim




From x.y.f at home.com  Sun Dec 31 21:09:23 2000
From: x.y.f at home.com (ktb)
Date: Sun, 31 Dec 2000 20:09:23 -0600
Subject: appending to top of file
Message-ID: <20001231200923.A25114@debian.local>

	Hi, I'm new to python.  Just installed it and read through some
	of the docs.  My question is, can I append lines to the
	beginning of a file as opposed to the end?  I have a web log
	that I would like to write to the file I pull in with php.  I'm
	running Linux and python version 1.5.2.  I searched though some
	of the back archives but didn't have any luck.
	Thanks,
	kent

-- 
  "In order to make an apple pie from scratch,
      you must first create the universe."  
                 - Carl Sagan



From wware at world.std.com  Sun Dec 31 21:50:20 2000
From: wware at world.std.com (Will Ware)
Date: Mon, 1 Jan 2001 02:50:20 GMT
Subject: Pythonic concurrency (was: Erlang (was Re: Good introduction to 
References: <mailman.977984412.30211.python-list@python.org> <92f0il01a5u@news1.newsguy.com> <86wvckqw4h.fsf@localhost.my.domain> <G6AyoC.CK1@world.std.com> <92A081D417204A91.AF29627FD1BF4835.0C336CAB4734825B@lp.airnews.net> <3A4F8E91.3F29F97C@letterror.com>
Message-ID: <G6GqJw.1oD@world.std.com>

> Cameron Laird wrote:
> > Is this true:  microthreads block on system calls (I/O
> > operations, most notoriously) and on invocations of
> > externally-coded extensions?

Just van Rossum (just at letterror.com) wrote:
> They do.

Yup, microthreads treat Python opcodes as atomic, including system I/O
calls and calls to C extensions. This is because Stackless Python is a
mechanism for tweaking the execution order of the Python virtual machine,
which steps thru Python opcodes. My earlier assertion was sloppy. My bad.
-- 
import string,time,os;print string.join((lambda x:x[:10]+x[8:])(map(
lambda x:string.center("*"*(lambda x:((x<24) ### Seasons Greetings, Will Ware
*(x-3))+3)(x),24),range(1,28, 2))),"\n") ################ wware at world.std.com


From mballen at erols.com  Sun Dec 31 21:17:09 2000
From: mballen at erols.com (Michael B. Allen)
Date: 1 Jan 2001 02:17:09 GMT
Subject: SMB Module
References: <92n1lh02evv@drn.newsguy.com> <3A4F3B45.CB459045@stroeder.com>
Message-ID: <slrn94vqll.544.mballen@angus.foo.net>

Michael Str?der wrote:
>which is not only used for file and printer sharing.

Well, it doesn't look like smbclient.so does anything but basic file
operations at the moment.

>How about password stuff, naming service, locating shares etc.? Just

There's no "password stuff" other than what is necessary to authenticate
with a target service. I believe the plan for it, if it does not do
this already, is to list hosts/shares/files/directories and basic file
operations. I think the idea is that it will be used in some kind of
Network Neighborhood/Explorer like GTK+ app for Linux but I have looked
at the code and the API is very general:

smbc_open, smbc_creat, smbc_read, smbc_write, smbc_close, smbc_unlink,
smbc_rename, smbc_lseek, smbc_stat, smbc_fstat, smbc_chown, smbc_chmod,
smbc_opendir, smbc_closedir, smbc_getdents, smbc_mkdir, smbc_lseekdir

Mike

-- 
signature pending


From dave_lambert at ev1.net  Sun Dec 31 18:27:25 2000
From: dave_lambert at ev1.net (David Lambert)
Date: Sun, 31 Dec 2000 17:27:25 -0600
Subject: Bug? in os.chown()
Message-ID: <3A4FC0DD.530F90E6@ev1.net>

I am running Python 2.0 on Mandrake Linux 7.2 and have come across an
apparent anomaly with os.chown(). Running a process as root, I can
change any file's ownerships apart from symbolic links. No error is
given, but the ownership refuses to be changed. Is this a feature or a
bug? How do I change a symbolic link's owner and group?

Any help would be greatly appreciated.

Dave Lambert.


From tim.one at home.com  Sun Dec 31 19:50:36 2000
From: tim.one at home.com (Tim Peters)
Date: Sun, 31 Dec 2000 19:50:36 -0500
Subject: Accitental hang up
In-Reply-To: <3A4D2522.A45A718@envision.co.il>
Message-ID: <LNBBLJKPBEHFEDALKOLCGECKIGAA.tim.one@home.com>

[Nadav Horesh]
> I occasionally work with very large arrays (>100000 elements):
> sometimes, when I accidentally type a name of an array followed by <CR>
> python hangs while trying to format the array for output.  Is there a
> nice way to abort the process without closing the python session?

It's not hung, but it may take a verrrrrry long time to complete building
the output string.  If Ctrl-C doesn't interrupt it, you're stuck.  PEP 217
proposes to add a display hook for interactive use, which you could fiddle
to special-case arrays (or whatever else you wanted to customize about the
default shell output):

    http://python.sourceforge.net/peps/pep-0217.html




From gregj at pobox.com  Sun Dec 31 23:28:06 2000
From: gregj at pobox.com (Greg Jorgensen)
Date: Mon, 01 Jan 2001 04:28:06 GMT
Subject: appending to top of file
References: <mailman.978315002.31351.python-list@python.org>
Message-ID: <qHT36.213461$U46.6540380@news1.sttls1.wa.home.com>

"ktb" <x.y.f at home.com> wrote in message
news:mailman.978315002.31351.python-list at python.org...
> Hi, I'm new to python.  Just installed it and read through some
> of the docs.  My question is, can I append lines to the
> beginning of a file as opposed to the end?  I have a web log
> that I would like to write to the file I pull in with php.  I'm
> running Linux and python version 1.5.2.  I searched though some
> of the back archives but didn't have any luck.

You can only append to the end of a file. You'll have to make a new file,
writing the new content first, then copying the old content.

A more scalable way to handle applications like you describe is to put each
chunk of content (article, log entry, etc.) into a separate file, or into a
database (like mySQL), and construct the HTML page dynamically. For example
if you had a directory of log entries named by date (log-20001229.txt,
log-20001231.txt, log-20010101.txt, etc.) you could easily get a list of
filenames, sort by date, and output each entry. A database engine will be
faster and easier to manage once you get a moderate number of entries.

--
Greg Jorgensen
PDXperts
Portland, Oregon, USA
gregj at pobox.com




From jessw at loop.com  Sun Dec 31 22:40:03 2000
From: jessw at loop.com (Jesse W)
Date: Sun, 31 Dec 2000 19:40:03 -0800
Subject: Accessing variables created after import in imported files
Message-ID: <3A4FFBCB.76A7@loop.com>

Sorry about the somewhat unclear subject line.  I am not totally sure
of the general form of my problem, so it is hard to write it as a
subject line.  My problem involves a program I am writting to allow
people to play a really fun card game I know, Hot Death Uno.  Insted of
posting the actual code(It's really long, and even snipits would be
pretty complex) I have enclosed code which demostrates the problem, but
is not the actual code I am working with.  If someone want's to see it,
I would be glad to e-mail it to them. 
	The code is in two files:
#file 1
import file2

class Game:
    def __init__(self):
	global TheGame
	TheGame=self 
	#this lets other classes access the current
 	#instence of the game without knowing the name it was
	#created as.
	self.foo='Some random variable'
x=Game()
foo=file2.ThingThatUsesGameClass()

#file 2
import file1

class ThingThatUsesGameClass:
    def __init__(self):
	print TheGame.foo #this should use the instance of the Game 	#class
created in the other file.


		Thank you very much for your time,
					Jesse Weinstein


From jasonic at nomadicsltd.com  Sun Dec 31 19:27:40 2000
From: jasonic at nomadicsltd.com (Jason Cunliffe)
Date: Sun, 31 Dec 2000 19:27:40 -0500
Subject: ML & Python question
References: <t4uqtfi3fv7sf2@corp.supernews.com> <slrn94vfjg.4av.neelk@alum.mit.edu>
Message-ID: <t4vji5j4fnph30@corp.supernews.com>

Neelakantan

wow thanks! This is exactly the sort of overview I was looking for.
Already my friend's advice makes more sense. :-)
And your presentation and encouragment really help..

Ironically, while looking through comp.lang.ml today, I found a thread from
someone asking about Python and ML. Eventutally he got a rasonable answer,
but not without some excessive academic flames and obfuscations. A revealing
discussioon in any event, but a shame he was not so lucky to get the
immediate answer I received from you.

Which confirms by belief that in choosing ones tools, 50% of the question
relates not to virtues of the technology, but rather to the quality and
vibrancy if its community.

FYI the thread where Python entered the discussion picks up at:
http://www.mailgate.org/comp/comp.lang.ml/msg00373.html


- Jason
___________________________________________________________
Jason CUNLIFFE = NOMADICS['Interactive Art and Technology']

Neelakantan Krishnaswami <neelk at alum.mit.edu> wrote in message
news:slrn94vfjg.4av.neelk at alum.mit.edu...

> The very short answer is that Python is a very high level, dynamically
> typed OO language, and that standard ML is a very high level,
> statically typed functional language. Both are IMO incredibly fun to
> program in.
[snip lots of great stuff]




From sandj.williams at gte.net  Sun Dec 31 23:36:26 2000
From: sandj.williams at gte.net (Steve Williams)
Date: Mon, 01 Jan 2001 04:36:26 GMT
Subject: why no "do : until"?
References: <mailman.978264900.28841.python-list@python.org>
Message-ID: <3A500A82.C326D15@gte.net>

Tim Peters wrote:

> To summarize 10 years of fruitless discussion on this issue:  yes,
>
>     while 1:
>         ...
>         if finished:
>             break
>
> is the intended way to spell "do ... until" in Python.
>

[cut...cut...cut]

I adore the

    While 1:
        . . .
    Else:
        . . .

construct.

I use it in all my programs.  It's soooo Pythonic.